computers+are+bad
The Ascent to Sandia Crest 28 September 2025 at 00:00

The Ascent to Sandia Crest

28 September 2025 at 00:00

The Rotary Club will take immediate action on the Ellis ranch loop project. The Rotarians reached this decision at their weekly luncheon, held yesterday at the Albarado hotel.

The club's plan is not merely to give Albuquerque a good, short road to the Ellis ranch... They embrace the building of a seventy-mile scenic loop. ¹

Many Western cities are defined, in part, by their mountains. Those moving from town to town often comment on the disorientation, the disruption, caused by a change in the city's relation to the peaks. If you have ever lived in a place with the mountains on the west, and then a place with the mountains on the east, you will know what I mean. We get used to living in the shadows of mountains.

One of the appeals of mountains, perhaps the source of their mysticism, is inaccessibility. Despite their close proximity to Albuquerque, steep slopes and difficult terrain kept the Sandias a world apart from the city, even to this day. Yet we have always been driven to climb, to ascent to the summit.

Humans climb mountains not only as a matter of individual achievement, but also as a matter of infrastructure. Whether the inaccessibility of mountain peaks is a good thing or a bad thing depends on the observer; and even the most challenging mountain ascents are sometimes developed to the scale of an industrial tourism operation.

And somewhere, in between, are the Sandias. Not technically part of the Rocky Mountains but roughly aligned with them, the Sandias lack a clear peak. Instead, the range is an elongated ridge, making up the entire eastern boundary of the city of Albuquerque and extending some distance further north into the Sandia Pueblo. The highest point is at 10,679', relatively modest for the mountain states---but still one of the most prominent in New Mexico, moreso even than the higher-elevation Mt. Taylor and Wheeler Peak.

Today, the Sandias are a major site for recreation. Tourists reach the upper parts of the mountain by one of two means: Sandia Crest Scenic Highway, which climbs the gentler eastern slope of the mountain to the crest itself; or the Sandia Peak Aerial Tramway, which makes a daring ascent from the edge of the city up the western side. Either provides access to an extensive network of trails, as do numerous points in the foothills and along the scenic highway.

Like bagging any peak, these access routes were hard-won. Their present contours---engineering achievements, twists and turns, scars on the landscape---reflect over one hundred years of ambition and conflict. Some of Albuquerque's most prominent figures left their marks on the mountain, as did everyday Burqueños, President Richard M. Nixon, and the slow grind of state and federal bureaucracy. The lay of the land today tells a story about the changing relationship of the city with the mountains, and more broadly of the American public with the wilderness.

It also explains why so many older documents and maps refer to the highway to the crest as a "loop," when it does no such thing. Here's the clickbait headline: Where does the Sandia Loop Highway go? The answer will surprise you.

Exploration of the Sandias by Spanish expeditions were motivated in large part by rumors of great mineral wealth. Certainly there are minerals in the mountains, and the Pueblo people had told the Spanish expeditions coming up the valley of gold and other precious metals, both in the mountains and in the plains beyond them. With the full perspective of history, it seems that these reports were less grounded in fact and more in a well-founded desire to get the Spanish to move on elsewhere. Supposedly several mines were established, but they left little mark on the landscape and their locations are now forgotten.

The Sandias, while rich in scenery, were not rich in gold. For centuries, they remained mostly undeveloped, besides Pueblo settlements in areas such as Tijeras canyon which were later abandoned in large part due to the increasing encroachment of settlers. Just a small number of people, mostly hunters, called the mountains their home. A few mining operations made their way into the mountains during the 19th century, perhaps repeating the mistakes of the Spanish several hundred years before.

It was these mining camps that brought the Ellis family to Albuquerque. They had run a ranch in the eastern plains of New Mexico, but were driven into the city by lack of water. Patriarch George Ellis found work for a fruit distributor, delivering produce to the mining camps in the mountains. Perhaps missing their days out in the country, George took his time as he traveled the canyons and ridges between camps. His wanderings were rewarded when he found Las Huertas Canyon, a narrow slot with a creek, protection from the wind, and what George Ellis himself called "extraordinary beauty."

The Ellis family moved into the canyon as early as 1893, but in 1905 George Ellis filed a land patent for what was, as an outcome of the Treaty of Guadalupe Hidalgo, frontier up for the taking. Ellis Ranch, as it was called, was indeed extraordinary: one of the most pleasant places in the mountains. But, then, the Ellis family were rather extraordinary as well. Life on the mountainside at 7,600 feet wasn't easy, but they kept cattle, chickens, and a modest farm.

Charlotte Ellis, George's daughter, had not yet a year of education at UNM when they moved out of town and up the slopes on the east side of the ridge. Still, she built a notable career as an amateur botanist. It is because of her efforts that Ellis ranch is often described in terms of its plant life---Charlotte collected specimens of hundreds of plant species, some of them new discoveries, and the ranch is said to have contained examples of every flowering plant known to exist in the mountains.

Plants were not Charlotte's only interest, though. She was apparently a skier, in a time well before skiing was known as a recreational pastime. A photo taken in 1896 shows her in a snow-covered landscape, facing away from the camera. She's looking down the slope, wearing a full dress and a pair of wooden skis. Her pose suggests a certain resolve, the kind you would need to spend the winter high above the nearest towns. It is often captioned as the first photograph of a person skiing in New Mexico. It may be the first photo of a woman skiing at all.

It also foreshadows the tremendous impact that skiing would have on the mountains and on Albuquerque. We don't know exactly what she was thinking, but what we do know about her suggests that she viewed the mountains much as we do today: for their beauty, their diversity, and for their thrill.

George Ellis had perhaps filed his land patent just in time. The Forest Service had existed for some years but was reorganized under its current name, and as part of the Department of the Interior, in 1905. The forests of the Sandia Mountains have a complicated legal history owing to the interaction of Spanish and Mexican land grants with the Treaty of Guadalupe Hidalgo at the conclusion of the Mexican-American war. You could say that the matter of the forest's legal ownership was not entirely settled until 2004.

None of that stopped President Roosevelt declaring the Sandias part of the Manzano Forest Reserves in 1906. Shortly after, the Forest Reserves became a National Forest and a ranger station was under construction near Tijeras. Paul Ellis, the son of the family, was soon hired on as one of the national forest's first staff; the Ellis Ranch one of the few private parcels within the federal boundaries.

The Forest Service has always served multiple purposes: "Land of Many Uses," as Forest Service signs still proclaim. Principle concerns of the forest service included lumber, cattle and sheep grazing, mining, and hunting. These uses were already widespread at the dawn of the modern Forest Service, so in practice, much of the role of early Forest Rangers was to bring forest users under control. Extensive regulations and rules were backed by a small set of rangers of the "Stetson era," rugged types not unlike the Ellises.

What we don't always picture as a part of that era was recreation. The "crunchy granola" brand of wilderness recreation most widely promoted today is indeed a more modern phenomenon, but forests have been viewed as recreational assets since the dawn of forest management---the form of that recreation has just changed, largely in response to new concepts of environmental stewardship that emphasize preservation over "improvement."

In 1913, George Ellis passed away. After a small ceremony, he was buried in a cemetery established next to his ranch. With George gone, the rest of the Ellis family began to scatter, moving into towns, closer to work. This left the question of the ranch.

By the time of George's death, Ellis Ranch had already become renowned for hospitality. The newspapers of the day recounted some of Albuquerque's finest taking vacations at the Ellis Ranch, a trip that was something of an adventure in the 1900s. It took a few hours, in fine conditions, to make the way up a rough trail to the Ellis homestead. One local businessman, known to newspaper readers as "the bee man," took some pride in having gotten a Ford automobile to within a mile of the site. Even so, the journey was well-agreed to be worthwhile. Las Huertas canyon was beautiful, and the Ellises known to offer fresh produce, spring water, and mountain air---a commodity in increasingly high demand as tuberculosis swept the country.

The reputation of the Ellis ranch for taking visitors explains the Albuquerque Tribune's matter-of-fact statement that the Ellis ranch would likely become a tourist resort, despite the lack of known buyers. At that time, the Forest Service was already in an infrastructure-building mood. La Madera, then one of the larger settlements in the east mountains, had telephone service and the Forest Service was extending the line to the ranch.

Shortly after, the buyers emerged, Raymond Stamm (uncle of the Stamm of Bradbury Stamm Construction) and Jack Sheehan of Albuquerque.

These young men propose to fit up the ranch as a mountain resort, the house being splendidly adapted to a mountain club house, with fine fishing, hunting, and camping facilities all about. ²

One challenge for this ambitious pair was, of course, the road: for the ranch to be the resort they imagined it needed an easier means of access. It was estimated that this effort would only require a few hundred dollars, reworking about three miles of trail between the ranch and a better road to Tijeras.

Even by 1913, an improved road to Ellis Ranch was a recurring theme in the newspapers. George Ellis, before his death, had estimated $500 to improve the road between the ranch and Placitas to passable condition. After his death, the estimate on this longer segment reached the thousands.

Some of the problem was, of course, hubris: the Albuquerque boosters of the day could look at three miles of faint road climbing up and down cabins and call it a few days of work. When these projects went underway, they always took longer and cost more than expected. To be fair, though, they were also fighting a living foe. Thunderstorm downpours and winter snow repeatedly washed out the road in the steeper sections. Reading of road improvement efforts, you get repeated deja vu. They are, in fact, announcing plans to fix the same section of road over and over again---it having always washed out the previous year.

The Stamm-Sheehan resort seems to have been popular, even as the road stayed rough. In September of 1914, the faculty of UNM held what we might now call an "offsite" there. Indeed, during 1914 and 1915 it seemed that practically everybody who was anybody was vacationing at Ellis Ranch. All the while, though, they were still struggling to get there... a situation that the Rotarians resolved to correct.

The road to Ellis Ranch has been known by many names and promoted by many people. The Rotary Club was a powerful force in the Albuquerque of 1916, though, and their proposal seems to have set the pattern for those that followed. The Ellis Ranch Loop Road, as they called it, would span 70 miles from Albuquerque to Albuquerque. The route went through the Tijeras pass, turned north to San Antonito ³ and passed Ellis Ranch on the way to La Madera and then Placitas. From there, it went nearly to Bernalillo, meeting the Santa Fe-Albuquerque road that would soon become part of Route 66---until 1937, at least.

A big argument for the loop, besides the fact that it makes the Ellis summer resort more easily accessible, is its scenery.

At the time of the Rotary Club proposal, the region's road map was in some ways familiar, and in some ways rather different. Many of the today's routes were already known but their condition was widely variable. The trip from Albuquerque to Tijeras, which would become the ultimate Route 66 alignment, was already in place. It was shorter in 1913, though, not extending much further east, and far from a highway. There was a rough road north to San Antonito, and there had once been a road further north from there to La Madera, but by 1916 it had been closed where it entered a private land grant.

This is probably the biggest difference from our modern road network: the route through Tijeras Canyon was not used for much other than access to Tijeras, and for much of the era covered in this article the preferred way to negotiate the Sandias and Manzanos was to go around them entirely. This is a good part of the reason for Route 66's original north-south alignment through the Rio Grande Valley, one that added significant mileage but avoided the mountain pass that was slow going in good weather and could be outright impassable in winter.

Similarly, from the north, there was an existing road from Bernalillo through Placitas and into the East Mountains. It was, in fact, the main route used by many of the residents there. It was not an easy route, though, long and prone to washouts.

And while these roads came close to forming a loop, they didn't. A section of about four miles was missing, right in the middle. Ellis Ranch was, of course, somewhere in that in-between.

So, the Rotary Club project was mostly the improvement of existing roads, along with the construction of new road in a couple of areas required to complete the loop around the Sandia range. The Rotarians estimated the cost at $5,000, and the Forest Service offered to pitch in.

Now, we must consider that $5,000 was a significant amount of money in 1916, something around $150,000 today. This inflation calculation probably underestimates the ask, because extensive investment in highway construction was not yet the pillar of American culture that it is today. Still, public enthusiasm for the project was high, and by the end of 1916 Bernalillo County had enacted a tax assessment to to support the road. The state legislature joined on, and with the Forest Service's share the project was fully funded. Construction seems to have began in earnest in 1917, on improvements to the road through Tijeras Pass.

Efforts on the loop road coincided, of course, with work to develop Ellis Ranch as a summer resort. Starting in 1917, regular newspaper ads appear for vacation cabin rentals at Ellis Ranch, with instructions to telephone Hugh Cooper. This High Cooper was Hugh Cooper Jr., son of the Hugh Cooper who was pastor of the First Presbyterian Church. Much like George Ellis, Rev. Cooper was a well-known figure around Albuquerque. Given that he officiated the wedding of at least one of Ellis's children, he was presumably friendly with the Ellises as well. The newspapers are not exactly clear on how it came to happen, but Rev. Cooper Sr would soon become much more involved in the ranch: in 1923, he and his sister went in together to buy it.

"We hope to keep this beauty spot of nature, so near our city, from falling into the hands of sports, who will advertise dance pavilions and entertainments of a sordid nature. If our plans mature we hope to make it a place where parents can go with their children and find recreation of an exalted character." ⁴

Rev. Cooper's motivations were, he assured the newspaper, godly rather than pecuniary. The phrase "exalted recreation" was singled out for the headline and is certainly one to remember as you explore the Sandias. Cooper was not, though, immune to the sort of boosterism that so animated the Rotary Club.

"We hope, with the help of the good people of Albuquerque, to make the famous Ellis Ranch a second Estes park."

The 1910s and '20s were boom years in Albuquerque, with a confluence of the railroad shops, the sawmill, and tuberculosis bringing new residents at a fast clip. The population more than doubled from 1900 to 1920, and with the newcomers came urban development. Subdivisions went under construction west of New Town, a skyscraper rose, and car ownership proliferated. The same period, for many of the same reasons, saw the genesis of American tourism. Trains brought visitors from out of town, and cars brought visitors and residents alike out into the surroundings. Boosters, from the city council to the chamber of commerce to a half dozen clubs, all had plans to make the best of it.

There was, to be sure, a certain aspect of keeping up with the joneses. Colorado was building a strong reputation as a tourist destination, and Denver as a beautiful and prosperous city. Arizona and Nevada had not yet gained populous cities, and west Texas was no less sparse than it is now. For the businessmen of Albuquerque, Colorado was the competition... and Colorado was building summer resorts.

The concept of the "summer resort" is mostly forgotten today, but in the 1920s there was little air conditioning and cities were gaining a reputation as crowded and polluted. The spread of Tuberculosis had brought particular attention to "clean, fresh air," and nothing was quite as clean and fresh as a summer day in the mountains. Estes Park, a town in the northern Colorado rockies, had multiple fine hotels by 1910 and attracted the tourism to match.

Further, in 1910, Denver's own boosters (groups like the Real Estate Exchange and the Denver Motor Club) began to acquire land for the Denver Mountain Parks. The beautiful Genesee Park was under construction in 1920, with a lodge and campground. Daniel Park had opened as a scenic drive (or "Auto View"). Plans for Mount Blue Sky Scenic Byway, the highest paved road in North America, had been finalized and announced to great fanfare.

To the modern conservationist it seems odd that these plans were so focused on roads, but the automobile was new and exciting and in the West had almost singlehandedly popularized the idea of "wilderness recreation." The mountains were no longer a destination only for the hardiest adventurers, they were also for city dwellers, out for a drive and a picnic. The ideal of natural preservation was a long, winding road with regular scenic pullouts. If the road could access a summer resort, all the better. Denver was doing it, and so Albuquerque would too.

Behind the scenes, the wheels of government had been turning. The Forest Service approved construction through the mountains in 1919, which would "make accessible some of the most beautiful mountain scenery in the state, in addition to bringing within easy reach sites for summer homes and temporary camping places." At this point in its history, the Forest Service routinely leased space for the construction of private homes, and a small number were established in the national forest as the road progressed inwards.

Bernalillo County had raised $12,500, the Forest Service committed $25,000, and the state was lobbied for another $12,500. A Forest Service engineer, speaking to the County Commission, noted that the road would provide access to 30 million board feet of lumber. The Forest Service said that construction could be finished by the end of the summer, and in September, the section from Placitas to Ellis Ranch opened to drivers. The road from Ellis Ranch to San Antonito had been surveyed, but work had not yet started.

We should consider again how this historic project relates to the modern roads, because there is a surprise. Today, the primary access from Albuquerque to the eastern slopes of the Sandias is via I-40 to Tijeras, and then NM-14 north through Cedar Crest and to then to San Antonito. There, NM-536 branches to the west, becoming Sandia Crest Scenic Highway.

NM-536 turns into Sandia Crest Scenic Highway, which is actually not a numbered highway at all, at a 90-degree turn by the Balsam Glade picnic area. From the picnic area, a much smaller, rougher dirt road descends to the north: NM-165. Today, NM-165 is a rugged and adventurous route to Placitas, impassable in winter and unfriendly to passenger vehicles.

And yet, in the summer of 1919, it was this route that was made a "first class road" to Ellis Ranch. The former site of Ellis Ranch now has a surprisingly well-paved parking lot off of the dirt NM-165, signed as "Las Huertas Canyon," and a trail continues up the creek through much of what was the summer resort. It took about an hour and a half, the Journal reported, to drive from Albuquerque to Ellis Ranch by this route around the north side. The travel time probably isn't all that much faster today, given the slow going on the dirt road.

In 1920, Bernalillo County ran into some trouble on the project funding. The county's plan to fund much of the project through bonds on the tax assessment was, it turns out, in violation of state law. For accounting reasons, the total sum of the bonds would have to be much lower. The county was effectively out of the project, at least as far as money was concerned.

The Forest Service worked according to its own plans. Construction started on a forest road, what is now NM-536, that would later be home to Tinkertown. In summer of 1922, it reached Sulphur Spring, the first of many recreation sites on its long trip up. Progress was brisk: the newspaper reported that 2.5 miles had been completed, then a few more, then eight. The forest road seems to have gone by different names at different times, and perhaps just depending on who was talking, but as 1922 came to a close the Forest Service named it for the canyon that it followed from San Antonito: Tejano Road.

Tejano Road ended about where the bottom of the ski area is found today, some four miles short of Ellis Ranch. Much of the Ellis Ranch Loop was complete, but it wasn't a loop. Plans had been completed for the final stretch, but it was unclear who would pay for the work---and when. The Journal printed an appeals to finish the project, several times mentioning the importance of a "sky line road."

The sky line road was something of a fad in those days. California and Oregon both had them, and plans were underway in many other parts of the west. A skyline road was what we might now call a scenic byway, but with a particular emphasis on views. Many followed a ridge, right down the center, so that drivers could get a view down both sides. Extensive earth-moving and forestry were sometimes involved, modifying the environment to create these views.

The West had a lead on the idea: it wasn't until 1925 that planning started on the most famous, Skyline Drive of Shenandoah National Park. The 100-mile highway was viewed as a wilderness adventure, but was about as far from our modern conception of wilderness as possible. Extensive clearing, of both trees and residents, were required to manicure its vistas. It was seen as one of the finest examples of American recreation, a model to be followed.

In July of 1923, the Ellis Ranch Loop was considered complete except for the four-mile segment between Tejano Road and Ellis Ranch Road and, oddly enough, the reworking of the road from Placitas to Ellis Ranch---the very one that had been completed in 1922. Road construction in the mountains was a fight against both funding and nature. The road seems to have been minimally improved, to keep the fast schedule and tight budget. During winter, snowfall and then snowmelt would wash parts of the road out. Throughout the 1920s and 1930s, the road was constantly being repaired and reworked. This probably explains why NM-165, once the principle route up the Sandias, is now so minimally maintained: the state and the Forest Service gave up.

During 1924, the Forest Service closed two miles of the gap and had the other two underway. In a turn of luck for the Albuquerque boosters, the Forest Service also offered to pick up work on the road to Placitas. The entire Ellis Ranch Loop project had become a Forest Service initiative, and they did what they set out to do.

In September 24th of 1924, an Albuquerque man named Arthur Ford set out through Tijeras Canyon with a friend, Mrs. Karskadon. They left Albuquerque at 10 am, finding that the new section connecting Ellis Ranch was not, officially, open; Ford was little discouraged and simply moved the barriers aside. At 11:45 they reached Ellis Ranch. After lunch, they continued northwards, to Placitas, through Bernalillo, and back into town. Ford's odometer measured the trip at 68 miles. The Ellis Ranch Loop was complete.

Imagine a circle approximately 20 miles in diameter.

Then imagine that this circle encloses some of the most beautiful mountain, valley, and mesa scenery in the world.

The city has always existed in tension with the mountains. The Ellis Family, who most famously opened the mountain to visitors, also oversaw a sort of closure to nature. George Ellis was the last person to see a Grizzly Bear in the Sandias. He came across the bear in 1905; he shot it.

From 1916 to 1924, Albuquerque business leaders admired the mountains for their beauty but lamented their inaccessibility. The Sandias could attract visitors from around the nation, they argued, if only you could get to them. Charlotte Ellis would hoof it on foot and on skis, but, then, she was Charlotte.

Then imagine that this circle is bounded by a highway that is traversible every day in the year.

Imagine, within the 70-mile circumference of this circle, near its eastern rim, a cluster of summer houses, supplied with water, light, and other necessaries. Imagine, at various spots within the circle, fine picnic and camping grounds, where people from the hot city may go for a day's outing.

We have always been driven to climb. As soon as it was possible to drive the full loop, to reach Ellis Ranch on and easy road, to take in the "Million Dollar Playground" of the Sandias, the Kiwanis Club formed a committee of business leaders and Forest Service representatives to consider the future of Sandia development.

It was "not only practicable, but highly necessary," they said, to take the next logical step. Ellis Ranch, at 7,500 feet, was only halfway from the city to the crest. A road all the way to the top---from Ellis Ranch to the crest---would complete their vision of the Sandias. "An effort will be made to have work begun as soon as possible."

You may think you are dreaming. And perhaps you are. But some day you will wake up and find the dream come true. ⁵

The Forest Service cleared six miles of steep, tight switchbacks, from the Balsam Glade area just above Ellis Ranch to the crest itself, over 10,500 feet. The New Mexico Highway Department, bolstered by the completion of Route 66, laid a gravel roadbed. Automobiles had become more common, driving a more popular pastime. It didn't take the adventurous Arthur Ford and Mrs. Karskadon to inaugurate the crest spur. On October 10th, 1927, Highway Department officials at the crest counted 110 cars.

Albuquerque had summited the mountain---but the greatest controversy was still to come.

Albuquerque Journal, 1916-08-04 p 6↩
Albuquerque Tribune, 1916-03-23 p 3↩
The names of the towns at this location are a bit confusing historically, so I am simply using San Antonito to refer to the spot that is currently occupied by Sandia Park and San Antonito. Sandia Park appears to be a renaming of a town that was originally called San Antonio (i.e. the big one), likely due to the presence of another San Antonio, New Mexico in Socorro County.↩
Albuquerque Journal, 1923-08-11 p 10↩
Albuquerque Journal, 1923-07-22 p 3↩

computers+are+bad
T-carrier 20 September 2025 at 00:00

T-carrier

computers+are+bad

20 September 2025 at 00:00

Few aspects of commercial telecommunications have quite the allure of the T-carrier. Well, to me, at least, but then I have very specific interests.

T-carrier has this odd, enduring influence on discussion of internet connections. I remember that for years, some video game installers (perhaps those using Gamespy?) used to ask what kind of internet service you had, with T1 as the "highest" option. The Steam Hardware Survey included T1 among the options for a long time. This was odd, in a lot of ways. It set T1 as sort of the "gold standard" in the minds of gamers, but residential internet service over T1 would have been very rare. Besides, even by the standards of the 2000s T1 service was actually pretty slow.

Still, T1 involved a healthy life as an important "common denominator" in internet connectivity. As a regulated telephone service, it was expensive, but available pretty much anywhere. It also provided a very high standard for reliability and latency, beyond many of the last-mile media we use today.

Telephone Carriers

We think of telephone calls as being carried over a pair of wires. In the early days of the telephone system, it wasn't all that far off to imagine a phone call as a single long circuit of two wires that extended all the way from your telephone to the phone you had called. This was the most naive and straightforward version of circuit switching: connections were established by creating a circuit.

The era of this extremely literal form of circuit switching did not last as long as you might think. First, we have to remember that two-wire telephone circuits don't actually work that well. Low install cost and convenience means that they are the norm between a telephone exchange and its local callers, but for long-distance carriage over the phone network, you get far better results by splitting the "talk" and "listen" sides into two separate pairs. This is called a four-wire telephone circuit, and while you will rarely see four-wire service at a customer's premises, almost all connectivity between telephone exchanges (and even in the internals of the telephone exchange itself) has been four-wire since the dawn of long-distance service.

Four-wire circuits only exacerbated an obvious scaling problem: in the long distance network, you have connections called a toll leads between two exchanges. In a very simple case, two towns might have a toll lead between them. For simple four-wire telephone lines, that toll lead needs four wires for each channel. If it has four wires, only one phone call can take place between the towns at a time. If it has eight wires, two telephone calls. This got very expensive very fast, considering that even heavily-built four-crossarm open wire routes might only have a capacity for eight simultaneous calls.

For obvious reasons, research in the telephone industry focused heavily on ways to combine more calls onto fewer wires. Some simple electrical techniques could be used, like phantoms that combined two underlying pairs into a single additional "phantom" pair for a 50% increase in capacity. You could extend this technique to create more channels, with a noticeable loss in quality.

By the 1920s, the Bell System relied on a technique that we would later call frequency division multiplexing (FDM). By modulating a phone call onto a higher-frequency carrier, you can put it over the same wires as other phone calls modulated onto different frequencies. The devices that actually did this combined multiple channels onto a single line, so they were known as channel banks. The actual formats they used over the wire, since they originally consisted of modulation onto a carrier, came to be known themselves as carriers. AT&T identified the carriers they developed with simple sequential letters. In the 1940s, the state of the art was up to J- and K-carrier, which allowed 16 channels on a four-wire circuit (over open wire and multipair cable, respectively). A four-crossarm open-wire circuit, with sixteen pairs, could support 256 unidirectional circuits for 128 channels---or simultaneous phone calls.

FDM carriers reached their apex with the coaxial-cable based L-carrier and and microwave radio TH and TD carriers ¹, which combined channels into groups, groups into supergroups, and supergroups into mastergroups for total capacities that reached into thousands of channels. Such huge FDM groups required very large bandwidths, though, which could not be achieved over copper pairs.

In the 1950s, rapidly developing digital electronics lead to the development of digital carriers. These carriers relied on a technique called Pulse-Code Modulation or PCM. PCM has sort of an odd name due to the history; it dates so far back into the history of digital communications that it was named before the terminology was well-established. "Pulse-code modulation" really just means "quantizing an analog signal to numbers and sending the numbers," which is now intuitive and obvious, but was an exciting new development of the 1940s.

PCM had a lot of potential for carrying phone calls, because the relaxed needs of voice transmission meant that calls could be digitized to fairly low-rate streams (8kHz by 8 bits) and engineers could see that there was a huge variety of possible techniques for combining and transporting digital signals. The best thing, though, was that digital signals could reliably be transmitted as exact copies rather than analog recreations. That meant that PCM telephone calls could pass through a complex network, with many mux/demux steps, without the reduction in quality that analog FDM carriers caused.

Even better, analog channel banks were large systems with a lot of sensitive analog components. They were expensive to build, required controlled environments, and were still subject to drift that required regular maintenance by technicians. Digital technology involved far fewer sensitive analog components, promising cheaper equipment with less maintenance. Digital was clearly the future.

The question, though, was how to best carry these digital signals over the wiring that made up much of the telephone system: copper pairs. In the late 1950s, Bell Laboratories developed T-carrier as the answer.

T-Carrier

T-carrier is a specification for transporting a stream of bits over copper pairs. The plural here is important: because T-carrier supported multiple channels, it was developed as a trunk technology, used for communication between telephone exchanges rather than between a telephone exchange and a customer. So, like other carriers used for trunks, T-carrier was four-wire, requiring two pairs to carry bidirectional signals.

T-carrier operates at 1.544Mbps, and that's about all you can say about it. The logical protocol used over T-carrier, the actual application of those bits, is determined by a separate protocol called the digital signal or DS. You can roughly think of this as a layer model, with DS running on top of T.

Here, we need to address the elephant in the room: the numbers. The numbers follow a convention used throughout AT&T-developed digital standards that is most clearly illustrated with DS. A DS0, by analogy to DS raised to the zeroeth power, is one single telephone channel expressed as a PCM digital signal. Since a telephone call is conveyed as 8-bit samples at 8kHz, a DS0 is 64kbps of data.

A DS1 is a combination of 24 DS0s, for 1.544Mbps.

A DS2 is a combination of 4 DS1s, for 96 channels or 6.312Mbps.

A DS3 is a combination of 7 DS2s, for 672 channels or 44.736Mbps.

Each level of the DS hierarchy is a TDM combination of several instances of the level below. The numbers are kind of odd, though, right? 24, 4, 7, it has the upsetting feeling of a gallon being four quarts each of which is two pints.

The DS system was developed in close parallel with the carriers actually used to convey the signal, so the short explanation for this odd scheme is that a DS1 is the number of DS0s that fit onto a T1 line, and a DS2 is the number of DS1s that fit onto a T2 line. The numbers are thus parallel: DS1 over T1, DS2 over T2, DS3 over T3. The distinction between T and DS is thus not always that important, and the terms do get used interchangeably.

But still, why 24?

Well, it's said that the number of channels on a T1 was just determined empirically. The 64kbps rate of a DS0 was fixed by the 8b x 8kHz digital format. A test T1 installation was built using a typical in-place copper telephone cable, and the number of channels was increased until it no longer functioned reliably. 24 channels was the magic number, the most achieved without unacceptable errors.

T-Carrier Infrastructure

T1 was designed for operation over a "typical" telephone cable trunk. In the 1950s, this meant a twisted-pair telephone cable installed in 6,600 foot sections with a loading coil at the junction of each section. A loading coil was essentially a big inductor hooked up to a telephone line at regular intervals to compensate for the capacitance of the line---long telephone cables, even four-wire, needed loading coils at regular intervals or the higher frequencies of speech would be lost. Loading coils also had disadvantages, though, in that they imposed a pretty sharp maximum frequency cutoff on the line. High-speed digital signaling needed to operate at those high frequencies, so T1 was designed to fit into existing long cables by replacing the loading coils with repeaters.

That means that T1 required a repeater very 6,600 feet. This repeaters were fairly small devices, often enclosed in pressurized cans to keep water out. 6,600 feet might sound pretty frequent, but because of the loading coil (and splice box) requirements trunk lines usually had underground vaults or equipment cabinets at that interval anyway.

Over time, the 6,600 foot interval became increasingly inconvenient. This was especially true as end-users started to order T1 service, requiring that T1 be supported on local loops that were often longer than 6,600 feet. Rather than installing new repeaters out in the field, it became a widespread practice to deliver T1 over a version of DSL called HDSL. HDSL is older and slower than the newer ADSL and VDSL protocols, and requires four wires, but it was fast enough to carry a DS1 signal and could cover a much longer span than traditional T-carrier. HDSL used the voice frequency band and thus could not coexist with voice calls like ADSL or VDSL, but this had the upside that it "fully controlled" the telephone line and could use measures like continuous monitoring (using a mark signal when there was no traffic) to maintain high reliability.

For the era of internet-over-T1, then, it was far more likely that a given customer actually had an HDSL connection that was converted to T1 at the customer premises by a device called a "smart jack." This pattern of the telco providing a regulated T1 service over a different media of their choice, and converting it at the customer premises, is still common today. T1s ordered later on may have actually been delivered via fiber with a similar on-premises media converter, depending on what was most convenient to the telco.

T1 is typically carried over telephone cable with 8P8C modular connectors, much like Ethernet. It requires two pairs, much like the two-line telephone wiring commonly installed in buildings. However, like most digital carriers, T-carrier is more particular about the wiring than conventional telephone. T1 wiring must be twisted-pair, and it is polarity sensitive.

DS1 Protocol

The DS1 protocol defines the carriage of 24 64kbps channels over a T1 interface. This basically amounts to TDM-muxing the 24 channels by looping over them sending one byte at a time, but there are a surprising number of nuances and details.

Early versions of DS1 only actually carried 7 bits for each sample, which was sufficient for a telephone call when companding was used to recover some of the dynamic range. The eighth bit was used for framing. T-carrier is a completely synchronous system, requiring that all of the equipment on the line have perfectly synchronized "frame clocks" to understand what bits belong to which logical channels. The framing mechanism provided a synchronization signal to achieve this perfect coordination. Later improvements in the framing protocol allowed for the use of all eight bits in some or even all of the samples. This gets to be a complicated and tangled story with many caveats, so I am going to leave it out here or this article would get a lot longer and probably contain a bunch more mistakes.

The various combinations of technologies and conventions used at different points get confusing. If you are curious, look into "robbed bit signaling," an artifact of the transition of where framing and control signals in T1 were placed that was, for some reason, a pet topic of one of my college professors. I think we spent more time on robbed-bit signaling than we did on all of MPLS, which is way cooler. Anyway, the point of this is to understand the protocol overhead involved: T1 operates at 1.544Mbps, but at least one 8-bit "sample" must be used for framing purposes, leaving 1.536Mbps of actual payload. The payload may be further reduced by other framing/signaling overhead, depending on exactly how the channel bank is configured. Most of these issue are specific to carrying telephone calls (and their related signaling); "internet service" T1 lines typically used a maximally-efficient configuration.

The Internet

So far we have pretty much only talked about telephone calls, because that's what T-carrier was developed for. By the 1980s, though, the computer industry was producing a lot of new applications for high-speed digital connections. T1 was widely available, and in many cases a tariffed/regulated service, so it was one of the most widely available high-speed data connections. Especially in the very early days of the internet, was often the only option.

Into the 1990s, T1 was one of the dominant options for commercial internet access. It was rarely seen in homes, though, as it was quite expensive. Keep in mind that, from the perspective of the phone company, a T1 line was basically 24 phone lines. They charged accordingly.

To obtain internet service, you would order internet service either from the telco itself or from an independent provider that then ordered a connection from your premises to their point of presence on an open access basis. In this case you were effectively paying two bills, one to the telco for the T1 line and the other to the independent provider for internet connectivity... but the total was still often more affordable than the telco's high rates for internet services.

Because of the popularity of T1 for internet access, routers with T1 interfaces were readily available. Well, really, the wide variety of media used for data connections before Ethernet became such a common standard means that many routers of the era took interchangeable modules, and T1 was one of the modules you could get for them.

In implementation, a T1 line was basically a fast serial line from your equipment to your ISP's equipment. What actually ran over that serial line was up to the ISP, and there were multiple options. The most classical would be frame relay, an X.25-derived protocol mostly associated with ISDN. PPP was also a fairly common option, as with consumer DSL, and more exotic protocols existed for specialized purposes or ISPs that were just a little weird.

When the internet was new, 1.5Mbps T1 was really very fast---the NSFNET backbone was lauded for its speed when it went online as an all-T1 network in 1991. Of course, today, a 1.5Mbps "backbone" internet connection is pretty laughable. Even as the '90s progressed, 1.5Mbps started to feel tight.

One of the things I find odd about the role of T1 in the history of internet access is that the era when a T1 was "blazing fast" was really pretty short. By 2000, when online gaming for example was taking off, both DSL and cable offered significantly better downstream speeds than T1. However, the nature of T1 as a circuit-switched, telephone-engineered TDM protocol made it very reliable and low-latency, properties that most faster internet media performed poorly on (early DSL notoriously so). Multiplayer gaming would likely have been a better experience on T1 than on a DSL connection offering multiples of the theoretical bandwidth.

A faster T1

Of course, there were options to speed up T-carrier. The most obvious is to combine multiple T1 connections, which was indeed a common practice. Later T1 interfaces were often supplied in multiple for that reason. MLPPP is a variant of the PPP protocol intended for combining the bandwidth of multiple links, referred to in the telephone industry as bonding.

But there were also higher levels in the hierarchy. Remember DS2 and DS3? Well, in practice, T2 wasn't really used. It was far more common to bond multiple T1 connections to reach the equivalent speed of a T2. 44.736Mbps T3 did find use, though. The catch is that T3 required specialized cabling (coaxial pairs!) and had a fairly short range, so it was usually not practical between a telephone exchange and a business.

Fortunately, by the time these larger bandwidths were in demand, fiber optic technology had become well-established. The telephone industry primarily used SONET, a fiber media over which Synchronous Digital Hierarchy (SDH) channels were carried. SONET comes in formats identified by OC (Optical Carrier) numbers in a way very similar to T-carrier numbers. An OC-1 is 51.840Mbps, already faster than T3/DS3. So, in practice, DS3 service was pretty much always provided by a media converter from an OC-1 SONET ring. As bandwidth demands further increased, businesses were much more likely to directly order SONET service rather than T-carrier. SONET was available into the multiple Gbps and enjoyed a long life as a popular internet carrier.

Of course, as the internet proliferated, so too did the stack of network media designed specifically for opportunistic packet-switching computer networks. Chief among them was Ethernet. These protocols have now overtaken traditional telephony protocols in most internet applications, so SONET as well is now mostly out of the picture. On the upside, Ethernet networks are generally more cost-effective on a bandwidth basis and allow you to obtain much faster service than you would be able to afford over the telephone network. On the downside, Ethernet networks have no circuit switching or traffic engineering, and so they do not provide the quality of service that the telephone network did. This means more jitter and less predictability of the available bandwidth, an enduring challenge for real-time media applications.

Miscellaneous

One of the confusing things about T1 to modern readers is its interaction with ISDN. ISDN was a later development that introduced a lot more standardization to digital communications over the telephone network, but it incorporated and reused existing digital technology including T-carrier. In the world of ISDN, a "DS0" or single telephone channel is called a basic rate interface (BRI), while the 24-channel T1 bandwidth is called a primary rate interface (PRI). Many T1 connections in the 1990s were actually ISDN PRIs.

The difference between the two is in the fine details: many of the details related to framing and control signals that were, shall we say, "loosey-goosey" with T-carrier are completely standardized under ISDN. An ISDN PRI always consists of 23 bearer channels ("payload" channels) and one control channel, and the framing is always the same. Since there is a dedicated control channel, there's no need to do weird things with the bits in the bearer channels, and so the overhead is standard across all ISDN PRIs.

In practice, the difference between T1 and ISDN PRI was usually just a matter of configuring your router's interface appropriately. Because of the curious details of the regulation and tariff processes, one was sometimes cheaper than the other, and in general the choice of whether to use a "T1" or a "PRI" was often somewhat arbitrary. It's even possible to use some T1 channels in the traditional way and others as ISDN.

While T1 is now mostly forgotten, some parts of its design live on. DSL and T1 have always had a relationship, DSL having originally been developed as basically a "better T-carrier" for ISDN use. In the modern world, DSL pretty much always refers to either ADSL or VDSL, newer protocols designed for consumer service that can coexist with a voice line and provide very high speeds. Many aspects of how DSL works have their roots in T-carrier, including the common use of protocols like ATM (now rare) or PPP (fading out) to encapsulate IP over DSL.

Okay, I realize this article has been sort of dry and factual, but I thought it'd be interesting to share a bit about T-carrier. I think it's something that people my age vaguely remember as important but have never thought about that much. I, personally, am probably just a bit too old to have had much real interaction with T-carrier. When I worked for an MSP for a bit in high school I saw a few customers that still had HDSL-based T1 service, and when I later interned at General Electric we had an OC-3 SONET connection that was in the process of being turned down. Just really catching the trail end... and yet for years later the Steam Hardware Survey was still asking if I had a T1.

Why did T1 get stuck so long in the very specific context of video games? I assume because video game developers frequently had T-carrier connections to their offices and knew that its guaranteed bandwidth provided very good performance for video games. The poor latency of ADSL meant that, despite a theoretical bandwidth several times larger, it was not really a better choice for the specific application of multiplayer games. So the "T1 is god tier" thing hung around for longer than you would have otherwise expected.

I actually don't know why the microwave carriers have multi-letter names starting in T. Something to look into. This convention is older than T-carrier and presumably started with TDX, the experimental microwave carrier that went into use in the late 1940s. I think the naming convention for carriers changed around the mid-century, as T1 is often said to stand for "transmission system one" which is consistent with later AT&T naming conventions but inconsistent with A-carrier through L-carrier, where the letters didn't stand for anything in particular. On the other hand, it is entirely possible that "T" was just the next letter in the sequence, and it standing for "transmission" was a later invention. You will also see people assert that the "T" stands for "trunk," perhaps evidence that the meaning is made up.↩

computers+are+bad
the video lunchbox 13 September 2025 at 00:00

the video lunchbox

computers+are+bad

13 September 2025 at 00:00

An opening note: would you believe that I have been at this for five years, now? If I planned ahead better, I would have done this on the five-year anniversary, but I missed it. Computers Are Bad is now five years and four months old.

When I originally launched CAB, it was my second attempt at keeping up a blog. The first, which I had called 12 Bit Word, went nowhere and I stopped keeping it up. One of the reasons, I figured, is that I had put too much effort into it. CAB was a very low-effort affair, which was perhaps best exemplified by the website itself. It was monospace and 80 characters wide, a decision that I found funny (in a shitposty way) and generated constant complaints. To be fair, if you didn't like the font, it was "user error:" I only ever specified "monospace" and I can't be blamed that certain platforms default to Courier. But there were problems beyond the appearance; the tool that generated the website was extremely rough and made new features frustrating to implement.

Over the years, I have not invested much (or really any) effort in promoting CAB or even making it presentable. I figured my readership, interested in vintage computing, would probably put up with it anyway. That is at least partially true, and I am not going to put any more effort into promotion, but some things have changed. Over time I have broadened my topics quite a bit, and I now regularly write about things that I would have dropped as "off topic" three or four years ago. Similarly, my readership has broadened, and probably to a set of people that find 80 characters of monospace text less charming.

I think I've also changed my mind in some ways about what is "special" about CAB. One of the things that I really value about it, that I don't think comes across to readers well, is the extent to which it is what I call artisanal internet. It's like something you'd get at the farmer's market. What I mean by this is that CAB is a website generated by a static site generator that I wrote, and a newsletter sent by a mailing list system that I wrote, and you access them by connecting directly to a VM that I administer, on a VM cluster that I administer, on hardware that I own, in a rack that I lease in a data center in downtown Albuquerque, New Mexico. This is a very old-fashioned way of doing things, now, and one of the ironies is that it is a very expensive way of doing things. It would be radically cheaper and easier to use wordpress.com, and it would probably go down less often and definitely go down for reasons that are my fault less often. But I figure people listen to me in part because I don't use wordpress.com, because I have weird and often impractical opinions about how to best contribute to internet culture.

I spent a week on a cruise ship just recently, and took advantage of the great deal of time I had to look at the sea to also get some work done. Strategically, I decided, I want to keep the things that are important to me (doing everything myself) and move on from the things that are not so important (the website looking, objectively, bad). So this is all a long-winded announcement that I am launching, with this post, a complete rewrite of the site generator and partial rewrite of the mailing list manager.

This comes with several benefits to you. First, computer.rip is now much more readable and, arguably, better looking. Second, it should be generally less buggy (although to be fair I had eliminated most of the problems with the old generator through sheer brute force over the years). Perhaps most importantly, the emails sent to the mailing list are no longer the unrendered Markdown files.

I originally didn't use markup of any kind, so it was natural to just email out the plaintext files. But then I wanted links, and then I wanted pictures, leading me to implement Markdown in generating the webpages... but I just kept emailing out the plaintext files. I strongly considered switching to HTML emails as a solution and mostly finished the effort, but in the end I didn't like it. HTML email is a massive pain in the ass and, I think, distasteful. Instead, I modified a Markdown renderer to create human-readable plaintext output. Things like links and images will still be a little weird in the plaintext emails, but vastly better than they were before.

I expect some problems to surface when I put this all live. It is quite possible that RSS readers will consider the most recent ten posts to all be new again due to a change in how the article IDs are generated. I tried to avoid that happening but, look, I'm only going to put so much time into testing and I've found RSS readers to be surprisingly inconsistent. If anything else goes weird, please let me know.

There has long been a certain connection between the computer industry and the art of animation. The computer, with a frame-oriented raster video output, is intrinsically an animation machine. Animation itself is an exacting, time-consuming process that has always relied on technology to expand the frontier of the possible. Walt Disney, before he was a business magnate, was a technical innovator in animation. He made great advances in cel animation techniques during the 1930s, propelling the Disney Company to fame not only by artistic achievement but also by reducing the cost and time involved in creating feature-length animated films.

Most readers will be familiar with the case of Pixar, a technical division of Lucasfilm that operated primarily as a computer company before its 1986 spinoff under computer executive Steve Jobs---who led the company through a series of creative successes that overshadowed the company's technical work until it was known to most only as a film studio.

Animation is hard. There are several techniques, but most ultimately come down to an animator using experience, judgment, and trial and error to get a series of individually composed frames to combine into fluid motion. Disney worked primarily in cel animation: each element of each frame was hand-drawn, but on independent transparent sheets. Each frame was created by overlaying the sheets like layers in a modern image editor. The use of separate cels made composition and corrections easier, by allowing the animator to move and redraw single elements of the final image, but it still took a great deal of experience to produce a reasonable result.

The biggest challenge was in anticipating how motion would appear. From the era of Disney's first work, problems like registration (consistent positioning of non-moving objects) had been greatly simplified by the use of clear cels and alignment pegs on the animator's desk that held cels in exact registration for tracing. But some things in an animation are supposed to move, I would say that's what makes it animation. There was no simple jig for ensuring that motion would come out smoothly, especially for complex movements like a walking or gesturing character. The animator could flip two cels back and forth, but that was about as good as they could get without dedicating the animation to film.

For much of the mid-century, a typical animation workflow looked like this: a key animator would draw out the key frames in final or near-final quality, establishing the most important moments in the animation, the positions and poses of the characters. The key animator or an assistant would then complete a series of rough pencil sketches for the frames that would need to go in between. These sketches were sent to the photography department for a "pencil test."

In the photography department, a rostrum camera was used: a cinema camera, often 16mm, permanently mounted on an adjustable stand that pointed it down at a flat desk. The rostrom camera looked a bit like a photographic enlarger and worked much the same way, but backwards: the photographer laid out the cels or sketches on the desk, adjusted the position and focus of the camera for the desired framing, and then exposed one frame. This process was repeated, over and over and over, a simple economy that explains the common use of a low 12 FPS frame rate in animation.

Once the pencil test had been photographed, the film went to the lab where it was developed, and then returned to the animation studio where the production team could watch it played on a cinema projector in a viewing room. Ideally, any problems would be identified during this first viewing before the key frames and pencil sketches were sent to the small army of assistant animators. These workers would refine the cels and redraw the pencil sketches in part by tracing, creating the "in between" frames of the final animation. Any needed changes were costly, even when caught at the earliest stage, as it usually took a full day for the photography department to return a new a pencil test (making the pencil test very much analogous to the dailies used in film). What separated the most skilled animators from amateurs, then, was often their ability to visualize the movement of their individual frames by imagination. They wanted to get it right the first time.

Graphics posed a challenge to computers for similar reasons. Even a very basic drawing involves a huge number of line segments, which a computer will need to process individually during rendering. Add properties such as color, consider the practicalities of rasterizing, and then make it all move: just the number of simple arithmetic problems involved in computer graphics becomes enormous. It is not a coincidence that we picture all early computer systems as text-only, although it is a bit unfair. Graphical output is older than many realize, originating with vector-mode CRT displays in the 1950s. Still, early computer graphics were very slow. Vector-mode displays were often paired with high-end scientific computers and you could still watch them draw in real time. Early graphics-intensive computer applications like CAD used specialized ASICs for drawing and yet provided nothing like the interactivity we expect from computers today.

The complexity of computer graphics ran head-first against an intense desire for more capable graphical computers, driven most prominently by the CAD industry. Aerospace and other advanced engineering fields were undergoing huge advancements during the second half of the 20th century. World War II had seen adoption of the jet engine, for example, machines which were extremely powerful but involved complex mathematics and a multitude of 3D parts that made them difficult for a human to reason over. The new field of computer-aided design promised a revolutionary leap in engineering capability, but ironically, the computers were quite bad at drawing. In the first decades, CAD output was still being sent to traditional draftsmen for final drawings. The computers were not only slow, but unskilled at the art of drafting: limitations on the number and complexity of the shapes that computers could render limited them to only very basic drawings, without the extensive annotations that would be needed for manufacturing.

During the 1980s, the "workstation" began to replace the mainframe in engineering applications. Today, "workstation" mostly just identifies PCs that are usually extra big and always extra expensive. Historically, workstations were a different class of machines from PCs that often employed fundamentally different architectures. Many workstations were RISC, an architecture selected for better mathematical performance. They frequently ran UNIX or a derivative, and featured the first examples of what we now call a GPU. Some things don't change: they were also very big, and very expensive.

It was the heady days of the space program and the Concorde, then, that brought us modern computer graphics. The intertied requirements for scientific computing, numerical simulation, and computer graphics that emerged from Cold War aerospace and weapons programs forged a strong bond between high-end computing and graphics. One could perhaps say that the nexus between AI and GPUs today is an extension of this era, although I think it's a bit of a stretch given the text-heavy applications. The echoes of the dawn of computer graphics are much quieter today, but still around. They persist, for example, in the heavy emphasis on computer visualization seen throughout scientific computing but especially in defense-related fields. They persist also in the names of the companies born in that era, names like Silicon Graphics and Mentor Graphics.

The development of video technology, basically the combination of preexisting television technology with new video tape recorders, lead to a lot of optimizations in film. Video was simply not of good enough quality to displace film for editing and distribution, but it was fast and inexpensive. For example, beginning in the 1960s filmmakers began to adopt a system called "video assist." A video camera was coupled to the film camera, either side-by-side with matched lenses or even sharing the same lens via a beam splitter. By running a video tape recorder during filming, the crew could generate something like an "instant daily" and play the tape back on an on-set TV. For the first time, a director could film a scene and then immediately rewatch it. Video assist was a huge step forward, especially in the television industry where it furthered the marriage of film techniques and television techniques for the production of television dramas.

It certainly seems that there should be a similar optimization for animation. It's not easy, though. Video technology was all designed around sequences of frames in a continuous analog signal, not individual images stored discretely. With the practicalities of video cameras and video recorders, it was surprisingly difficult to capture single frames and then play them back to back.

In the 1970s, animators Bruce Lyon and John Lamb developed the Lyon-Lamb Video Animation System (VAS). The original version of the VAS was a large workstation that replaced a rostrum camera with a video camera, monitor, and a custom video tape recorder. Much like the film rostrum camera, the VAS allowed an operator to capture a single frame at a time by composing it on the desk. Unlike the traditional method, the resulting animation could be played back immediately on the included monitor.

The VAS was a major innovation in cel animation, and netted both an Academy Award and an Emmy for technical achievement. While it's difficult to say for sure, it seems like a large portion of the cel-animated features of the '80s had used the VAS for pencil tests. The system was particularly well-suited to rotoscoping, overlaying animation on live-action images. Through a combination of analog mixing techniques and keying, the VAS could directly overlay an animator's work on the video, radically accelerating the process. To demonstrate the capability, John Lamb created a rotoscoped music video for the Tom Waits song "The One That Got Away." The resulting video, titled "Tom Waits for No One," was probably the first rotoscoped music video as well as the first production created with the video rotoscope process. As these landmarks often do, it languished in obscurity until it was quietly uploaded to YouTube in 2006.

The VAS was not without its limitations. It was large, and it was expensive. Even later generations of the system, greatly miniaturized through the use of computerized controls and more modern tape recorders, came in at over $30,000 for a complete system. And the VAS was designed around the traditional rostrom camera workflow, intended for a dedicated operator working at a desk. For many smaller studios the system was out of reach, and for forms of animation that were not amenable to top-down photography on a desk, the VAS wasn't feasible.

There are some forms of animation that are 3D---truly 3D. Disney had produced pseudo-3D scenes by mounting cels under a camera on multiple glass planes, for example, but it was obviously possible to do so in a more complete form by the use of animated sculptures or puppets. Practical challenges seem to have left this kind of animation mostly unexplored until the rise of its greatest producer, Will Vinton.

Vinton grew up in McMinnville, Oregon, but left to study at UC Berkeley. His time in Berkeley left him not only with an architecture degree (although he had studied filmmaking as well), but also a friendship with Bob Gardiner. Gardiner had a prolific and unfortunately short artistic career, in which he embraced many novel media including the hologram. Among his inventions, though, was a novel form of animation using clay: Gardiner was fascinated with sculpting and posing clay figures, and demonstrated the animation potential to Vinton. Vinton, in turn, developed a method of using his student film camera to photograph the clay scenes frame by frame.

Their first full project together, Closed Mondays, took the Academy Award for Best Animated Short Film in 1975. It was notable not only for the moving clay sculptures, but for its camerawork. Vinton had realized that in clay animation, where scenes are composed in real 3D space, the camera can be moved from frame to frame just like the figures. Not long after this project, Vinton and Gardiner split up. Gardiner seems to have been a prolific artist in that way where he could never stick to one thing for very long, and Vinton had a mind towards making a business out of this new animation technology. It was Vinton who christened it Claymation, then a trademark of his new studio.

Vinton returned to his home state and opened Will Vinton Studios in Portland. Vinton Studios released a series of successful animated shorts in the '70s, and picked up work on numerous other projects, contributing for example to the "Wizard of Oz" film sequel "Return to Oz" and the Disney film "Captain EO." By far Vinton Studios most famous contributions to our culture, though, are their advertising projects. Will Vinton Studios brought us the California Raisins, the Noid, and walking, talking M&M's.

Will Vinton Studios struggled with producing claymation at commercial scale. Shooting with film cameras, it took hours to see the result. Claymation scenes were more difficult to rework than cel animation, setting an even larger penalty for reshoots. Most radically, claymation scenes had to be shot on sets, with camera and light rigging. Reshooting sections without continuity errors was as challenging as animating those sections in the first place.

To reduce rework, they used pencil tests: quicker, lower-effort versions of scenes shot to test the lighting, motion, and sound synchronization before photography with a film camera. Their pencil tests were apparently captured on a crude system of customized VCRs, allowing the animator to see the previous frame on a monitor as they composed the next, and then to play back the whole sequence. It was better than working from film, but it was still slow going.

The area from Beaverton to Hillsboro, in Oregon near Portland, is sometimes called "the silicon forest" largely on the influence of Intel and Tektronix. As in the better known silicon valley, these two keystone companies were important not only on their own, but also as the progenitors of dozens of new companies. Tektronix, in particular, had a steady stream of employees leaving to start their own businesses. Among these alumni was Mentor Graphics.

Mentor Graphics was an early player in electronic design automation (EDA), sort of like a field of CAD specialized to electronics. Mentor products assisted not just in the physical design of circuit boards and ICs, but also simulation and validation of their functionality. Among the challenges of EDA are its fundamentally graphical nature: the final outputs of EDA are often images, masks for photolithographic manufacturing processes, and engineers want to see both manufacturing drawings and logical diagrams as they work on complex designs.

When Mentor started out in 1981, EDA was in its infancy and relied mostly on custom hardware. Mentor went a different route, building a suite of software products that ran on Motorola 68000-based workstations from Apollo. The all-software architecture had cost and agility advantages, and Mentor outpaced their competition to become the field's leader.

Corporations want for growth, and by the 1990s Mentor had a commanding position in EDA and went looking for other industries to which their graphics-intensive software could be applied. One route they considered was, apparently, animation: computer animation was starting to take off, and there were very few vendors for not just the animation software but the computer platforms capable of rendering the product. In the end, Mentor shied away: companies like Silicon Graphics and Pixar already had a substantial lead, and animation was an industry that Mentor knew little about. As best I can tell, though, it was this brief investigation of a new market that exposed Mentor engineering managers Howard Mozeico and Arthur Babitz to the animation industry.

I don't know much about their career trajectories in the years shortly after, only that they both decided to leave Mentor for their own reasons. Arthur Babitz went into independent consulting, and found a client reminiscent of his work at Mentor, an established animation studio that was expanding into computer graphics: Will Vinton Studios. Babitz's work at Will Vinton Studios seems to have been largely unrelated to claymation, but it exposed him to the process, and he watched the way they used jury-rigged VCRs and consumer video cameras to preview animations.

Just a couple of years later, Mozeico and Babitz talked about their experience with animation at Mentor, a field in which they were both still interested. Babitz explained the process he had seen at Will Vinton Studios, and his ideas for improving it. Both agreed that they wanted to figure out a sort of retirement enterprise, what we might now call a "lifestyle business": they each wanted to found a company that would keep them busy, but not too busy. The pair incorporated Animation Toolworks, headquartered in Mozeico's Sherwood, Oregon home.

In 1998 Animation Toolworks hit trade shows with the Video Lunchbox. The engineering was mostly by Babitz, the design and marketing by Mozeico, and the manufacturing done on contract by a third party. The device took its name from its form factor, a black crinkle paint box with a handle on top of its barn-roof-shaped lid. It was something like the Lyon Lamb VAS, if it was portable, digital, and relatively inexpensive.

The Lunchbox was essentially a framegrabber, a compact and simplified version of the computer framegrabbers that were coming into use in the animation industry. You plugged a video camera into the input, and a television monitor into the output. You could see the output of the camera, live, on the monitor while you composed a scene. Then, one press of a button captured a single frame and stored it. With a press of another button, you could swap between the stored frame and the live image, helping to compose the next. You could even enable an automatic "flip-flop" mode that alternated the two rapidly, for hands-free adjustment.

Each successive press of the capture button stored another frame to the Lunchbox's memory, and buttons allowed you to play the entire set of stored frames as a loop, or manually step forward or backward through the frames. And that was basically it: there were a couple of other convenience features like an intervalometer (for time lapse) and the ability to record short sections of real-time video, but complete operation of the device was really very simple. That seems to have been one of its great assets. The Lunchbox was much easier to sell after Mozeico gave a brief demonstration and said that that was all there is to it.

To professionals, the Lunchbox was a more convenient, more reliable, and more portable version of the video tape recorder or computer framegrabber systems they were already using for pencil tests. Early customers of Animation Toolworks included Will Vinton Studios alongside other animation giants like Disney, MTV, and Academy Award-winning animator Mark Osborne. Animation Toolworks press quoted animators from these firms commenting on the simplicity and ease of use, saying that it had greatly sped up the animation test process.

In a review for Animation World Magazine, Kellie-Bea Rainey wrote:

In most cases, computers as framegrabbers offer more complications than solutions. Many frustrations stem from the complexity of learning the computer, the software and it's constant upgrades. But one of the things Gary Schwartz likes most about the LunchBox is that the system requires no techno-geeks. "Computers are too complex and the technology upgrades are so frequent that the learning curve keeps you from mastering the tools. It seems that computers are taking the focus off the art. The Video LunchBox has a minimum learning curve with no upgrade manuals. Everything is in the box, just plug it in."

Indeed, the Lunchbox was so simple that it caught on well beyond the context of professional studios. It is remembered most as an educational tool. Disney used the Lunchbox for teaching cel animation in a summer program, but closer to home, the Lunchbox made its way to animation enthusiast and second-grade teacher Carrie Caramella. At Redmond, Oregon's John Tuck Elementary School, Caramella acted as director of a student production team that brought their short film "The Polka Dot Day" to the Northwest Film Center's Young People's Film and Video Festival. During the early 2000s, after-school and summer animation programs proliferated, many using claymation, and almost all using the Video Lunchbox.

At $3,500, the Video Lunchbox was not exactly cheap. It cost more than some of the more affordable computer-based options, but it was so much easier to use, and so much more durable, that it was very much at home in a classroom. Caramella:

"By using the lunchbox, we receive instant feedback because the camera acts > as an eye. It is also child-friendly, and you can manipulate the film a lot more."

Caramella championed animation at John Tuck, finding its uses in other topics. A math teacher worked with students to make a short animation of a chicken. In a unit on compound words, Caramella led students in animating their two words together: a sun and a flower dance; the word is "sunflower." Butter and milk, base and ball.

In Lake Oswego, an independent summer program called Earthlight Studios took up the system.

With the lunchbox, Corey's black-and-white drawings spring to life, two catlike animé characters circling each other with broad edged swords. It's the opening seconds of what he envisions will be an action-adventure film.

We can imagine how cringeworthy these student animations must be to their creators today, but early-'00s education was fascinated with multimedia and it seems rare that technology served the instructional role so well.

It was in this context that I crossed ways with the Lunchbox. As a kid, I went to a summer animation program at OMSI---a claymation program, which I hazily remember was sponsored by a Will Vinton Studios employee. In an old industrial building beside the museum, we made crude clay figures and then made them crudely walk around. The museum's inventory of Lunchboxes already showed their age, but they worked, in a way that was so straightforward that I think hardly any time was spent teaching operation of the equipment. It was a far cry from an elementary school film project in which, as I recall, nearly an entire day of class time was burned trying to get video off of a DV camcorder and into iMovie.

Mozeico and Babitz aimed for modest success, and that was exactly what they found. Animation Toolworks got started on so little capital that it turned a profit the first year, and by the second year the two made a comfortable salary---and that was all the company would ever really do. Mozeico and Babitz continued to improve on the concept. In 2000, they launched the Lunchbox Sync, which added an audio recorder and the ability to cue audio clips at specific frame numbers. In 2006, the Lunchbox DV added digital video.

By the mid-2000s, computer multimedia technology had improved by leaps and bounds. Framegrabbers and real-time video capture devices were affordable, and animation software on commodity PCs overtook the Lunchbox on price and features. Still, the ease of use and portability of the Lunchbox was a huge appeal to educators. By 2005 Animation Toolworks was basically an educational technology company, and in the following years computers overtook them in that market as well.

The era of the Lunchbox is over, in more ways than one. A contentious business maneuver by Phil Knight saw Will Vinton pushed out of Will Vinton Studios. He was replaced by Phil Knight's son, Travis Knight, and the studio rebranded to Laika. The company has struggled under its new management, and Laika has not achieved the renaissance of stop-motion that some thought Coraline might bring about. Educational technology has shifted its focus, as a business, to a sort of lightweight version of corporate productivity platforms that is firmly dominated by Google.

Animation Toolworks was still selling the Lunchbox DV as late as 2014, but by 2016 Mozeico and Babitz had fully retired and offered support on existing units only. Mozeico died in 2017, crushed under a tractor on his own vineyard. There are worse ways to go. Arthur Babitz is a Hood River County Commissioner.

Kellie-Bea Rainey:

I took the two-minute tutorial and taped it to the wall. I cleaned off a work table and set up a stage and a character. Then I put my Sharp Slimcam on a tripod... Once the camera was plugged into the LunchBox, I focused it on my animation set-up. Next, I plugged in my monitor.

All the machines were on and all the lights were green, standing by. It's time to hit the red button on the LunchBox and animate!

Yippee! Look Houston, we have an image! That was quick, easy and most of all, painless. I want to do more, and more, and even more.

The next time you hear from me I'll be having fun, teaching my own animation classes and making my own characters come to life. I think Gary Schwartz says it best, "The LunchBox brings the student back to what animation is all about: art, self-esteem, results and creativity."

I think we're all a little nostalgic for the way technology used to be. I know I am. But there is something to be said for a simple device, from a small company, that does a specific thing well. I'm not sure that I have ever, in my life, used a piece of technology that was as immediately compelling as the Video Lunchbox. There are numerous modern alternatives, replete with USB and Bluetooth and iPad apps. Somehow I am confident that none of them are quite as good.

The+Daily+WTF
CodeSOD: Contracting Space 29 September 2025 at 06:30

CodeSOD: Contracting Space

The+Daily+WTF

By: Remy Porter

29 September 2025 at 06:30

A ticket came in marked urgent. When users were entering data in the header field, the spaces they were putting in kept getting mangled. This was in production, and had been in production for sometime.

Mike P picked up the ticket, and was able to track down the problem to a file called Strings.java. Yes, at some point, someone wrote a bunch of string helper functions and jammed them into a package. Of course, many of the functions were re-implementations of existing functions: reinvented wheels, now available in square.

For example, the trim function.

    /**
     * @param str
     * @return The trimmed string, or null if the string is null or an empty string.
     */
    public static String trim(String str) {
        if (str == null) {
            return null;
        }

        String ret = str.trim();

        int len = ret.length();
        char last = '\u0021';    // choose a character that will not be interpreted as whitespace
        char c;
        StringBuffer sb = new StringBuffer();
        for (int i = 0; i < len; i++) {
            c = ret.charAt(i);
            if (c > '\u0020') {
                if (last <= '\u0020') {
                    sb.append(' ');
                }
                sb.append(c);
            }
            last = c;
        }
        ret = sb.toString();

        if ("".equals(ret)) {
            return null;
        } else {
            return ret;
        }
    }

Now, Mike's complaint is that this function could have been replaced with a regular expression. While that would likely be much smaller, regexes are expensive- in performance and frequently in cognitive overhead- and I actually have no objections to people scanning strings.

But let's dig into what we're doing here.

They start with a null check, which sure. Then they trim the string; never a good sign when your homemade trim method calls the built-in.

Then, they iterate across the string, copying characters into a StringBuffer. If the current character is above unicode character 20- the realm of printable characters- and if the last character was a whitespace character, we copy a whitespace character into the output, and then the printable character into the output.

What this function does is simply replace runs of whitespace with single whitespace characters.

"This        string"

becomes

"This string"

Badly I should add. Because there are plenty of whitespace characters which appear above \u0020- like the non-breaking space (\u00A0), and many other control characters. While you might be willing to believe your users will never figure out how to type those, you can't guarantee that they'll never copy/paste them.

For me, however, this function does something far worse than being bad at removing extraneous whitespace. Because it has that check at the end- if I handed it a perfectly good string that is only whitespace, it hands me back a null.

I can see the argument- it's a bad input, so just give me back an objectively bad result. No IsNullOrEmpty check, just a simple null check. But I still hate it- turning an actual value into a null just bothers me, and seems like an easy way to cause problems.

In any case, the root problem with this bug was simply developer invented requirements: the users never wanted stray spaces to be automatically removed in the middle of the string. Trimmed yes, gutted no.

No one tried to use multiple spaces for most of the history of the application, thus no one noticed the problem. No one expected it to not work. Hence the ticket and the panic by users who didn't understand what was going on.

[Advertisement] Picking up NuGet is easy. Getting good at it takes time. Download our guide to learn the best practice of NuGet for the Enterprise.

The+Daily+WTF
Error'd: Pickup Sticklers 26 September 2025 at 06:30

Error'd: Pickup Sticklers

The+Daily+WTF

By: Lyle Seaman

26 September 2025 at 06:30

An Anonymous quality analyst and audiophile accounted "As a returning customer at napalmrecords.com I was forced to update my Billing Address. Fine. Sure. But what if my *House number* is a very big number? More than 10 "symbols"? Fortunately, 0xDEADBEEF for House number and J****** for First Name both passed validation."

And then he proved it, by screenshot:

Richard P. found a flubstitution failure mocking "I'm always on the lookout for new and interesting Lego sets. I definitely don't have {{product.name}} in my collection!"

"I guess short-named siblings aren't allowed for this security question," pointed out Mark T.

Finally, my favorite category of Error'd -- the security snafu. Tim R. reported this one, saying "Sainsbury/Argos in the UK doesn't want just anybody picking up the item I've ordered online and paid for, so they require not one, not two, but 3 pieces of information when I come to collect it. There's surely no way any interloper could possibly find out all 3, unless they were all sent in the same email obviously." Personally, my threat model for my grocery pickups is pretty permissive, but Tim cares.

[Advertisement] Picking up NuGet is easy. Getting good at it takes time. Download our guide to learn the best practice of NuGet for the Enterprise.

The+Daily+WTF
Coded Smorgasbord: High Strung 25 September 2025 at 06:30

Coded Smorgasbord: High Strung

The+Daily+WTF

By: Remy Porter

25 September 2025 at 06:30

Most languages these days have some variation of "is string null or empty" as a convenience function. Certainly, C#, the language we're looking at today does. Let's look at a few example of how this can go wrong, from different developers.

We start with an example from Jason, which is useless, but not a true WTF:

/// <summary>
/// Does the given string contain any characters?
/// </summary>
/// <param name="strToCheck">String to check</param>
/// <returns>
/// true - String contains some characters.
/// false - String is null or empty.
/// </returns>
public static bool StringValid(string strToCheck)
{
        if ((strToCheck == null) ||
                (strToCheck == string.Empty))
                return false;

        return true;
}

Obviously, a better solution here would be to simply return the boolean expression instead of using a conditional, but equally obvious, the even better solution would be to use the built-in. But as implementations go, this doesn't completely lose the plot. It's bad, it shouldn't exist, but it's barely a WTF. How can we make this worse?

Well, Derek sends us an example line, which is scattered through the codebase.

if (Port==null || "".Equals(Port)) { /* do stuff */}

Yes, it's frequently done as a one-liner, like this, with the do stuff jammed all together. And yes, the variable is frequently different- it's likely the developer responsible saved this bit of code as a snippet so they could easily drop it in anywhere. And they dropped it in everywhere. Any place a string got touched in the code, this pattern reared its head.

I especially like the "".Equals call, which is certainly valid, but inverted from how most people would think about doing the check. It echos Python's string join function, which is invoked on the join character (and not the string being joined), which makes me wonder if that's where this developer started out?

I'll never know.

Finally, let's poke at one from Malfist. We jump over to Java for this one. Malfist saw a function called checkNull and foolishly assumed that it returned a boolean if a string was null.

public static final String checkNull(String str, String defaultStr)
{
    if (str == null)
        return defaultStr ;
    else
        return str.trim() ;
}

No, it's not actually a check. It's a coalesce function. Okay, misleading names aside, what is wrong with it? Well, for my money, the fact that the non-null input string gets trimmed, but the default string does not. With the bonus points that this does nothing to verify that the default string isn't null, which means this could easily still propagate null reference exceptions in unexpected places.

I've said it before, and I'll say it again: strings were a mistake. We should just abolish them. No more text, everybody, we're done.

[Advertisement] Keep the plebs out of prod. Restrict NuGet feed privileges with ProGet. Learn more.

The+Daily+WTF
CodeSOD: Across the 4th Dimension 24 September 2025 at 06:30

CodeSOD: Across the 4th Dimension

The+Daily+WTF

By: Remy Porter

24 September 2025 at 06:30

We're going to start with the code, and then talk about it. You've seen it before, you know the chorus: bad date handling:

C_DATE($1)
C_STRING(7;$0)
C_STRING(3;$currentMonth)
C_STRING(2;$currentDay;$currentYear)
C_INTEGER($month)

$currentDay:=String(Day of($1))
$currentDay:=Change string("00";$currentDay;3-Length($currentDay))
$month:=Month of($1)
Case of

: ($month=1)
$currentMonth:="JAN"

: ($month=2)
$currentMonth:="FEB"

: ($month=3)
$currentMonth:="MAR"

: ($month=4)
$currentMonth:="APR"

: ($month=5)
$currentMonth:="MAY"

: ($month=6)
$currentMonth:="JUN"

: ($month=7)
$currentMonth:="JUL"

: ($month=8)
$currentMonth:="AUG"

: ($month=9)
$currentMonth:="SEP"

: ($month=10)
$currentMonth:="OCT"

: ($month=11)
$currentMonth:="NOV"

: ($month=12)
$currentMonth:="DEC"

End case

$currentYear:=Substring(String(Year of($1));3;2)

$0:=$currentDay+$currentMonth+$currentYear

At this point, most of you are asking "what the hell is that?" Well, that's Brewster's contribution to the site, and be ready to be shocked: the code you're looking at isn't the WTF in this story.

Let's rewind to 1984. Every public space was covered with a thin layer of tobacco tar. The Ground Round restaurant chain would sell children's meals based on the weight of the child and have magicians going from table to table during the meal. And nobody quite figured out exactly how relational databases were going to factor into the future, especially because in 1984, the future was on the desktop, not the big iron "server side".

Thus was born "Silver Surfer", which changed its name to "4th Dimension", or 4D. 4D was an RDBMS, an IDE, and a custom programming language. That language is what you see above. Originally, they developed on Apple hardware, and were almost published directly by Apple, but "other vendors" (like FileMaker) were concerned that Apple having a "brand" database would hurt their businesses, and pressured Apple- who at the time was very dependent on its software vendors to keep its ecosystem viable. In 1993, 4D added a server/client deployment. In 1995, it went cross platform and started working on Windows. By 1997 it supported building web applications.

All in all, 4D seems to always have been a step or two behind. It released a few years after FileMaker, which served a similar niche. It moved to Windows a few years after Access was released. It added web support a few years after tools like Cold Fusion (yes, I know) and PHP (I absolutely know) started to make building data-driven web apps more accessible. It started supporting Service Oriented Architectures in 2004, which is probably as close to "on time" as it ever got for shipping a feature based on market demands.

4D still sees infrequent releases. It supports SQL (as of 2008), and PHP (as of 2010). The company behind it still exists. It still ships, and people- like Brewster- still ship applications using it. Which brings us all the way back around to the terrible date handling code.

4D does have a "date display" function, which formats dates. But it only supports a handful of output formats, at least in the version Brewster is using. Which means if you want DD-MMM-YYYY (24-SEP-2025) you have to build it yourself.

Which is what we see above. The rare case where bad date handling isn't inherently the WTF; the absence of good date handling in the available tooling is.

[Advertisement] Utilize BuildMaster to release your software with confidence, at the pace your business demands. Download today!

The+Daily+WTF
CodeSOD: One Last ID 23 September 2025 at 06:30

CodeSOD: One Last ID

The+Daily+WTF

By: Remy Porter

23 September 2025 at 06:30

Chris's company has an unusual deployment. They had a MySQL database hosted on Cloud Provider A. They hired a web development company, which wanted to host their website on Cloud Provider B. Someone said, "Yeah, this makes sense," and wrote the web dev company a sizable check. They app was built, tested, and released, and everyone was happy.

Everyone was happy until the first bills came in. They expected the data load for the entire month to be in the gigabytes range, based on their userbase and expected workloads. But for some reason, the data transfer was many terabytes, blowing up their operational budget for the year in a single month.

Chris fired up a traffic monitor and saw that, yes, huge piles of data were getting shipped around with every request. Well, not every request. Every insert operation ended up retrieving a huge pile of data. A little more research was able to find the culprit:

SELECT last_insert_id() FROM some_table_name

The last_insert_id function is a useful one- it returns the last autogenerated ID in your transaction. So you can INSERT, and then check what ID was assigned to the inserted record. Great. But the way it's meant to be used is like so: SELECT last_insert_id(). Note the lack of a FROM clause.

By adding the FROM, what the developers were actually saying were "grab all rows from this table, and select the last_insert_id once for each one of them". The value of last_insert_id() just got repeated once for each row, and there were a lot of rows. Many millions. So every time a user inserted a row into most tables, the database sent back a single number, repeated millions and millions of times. Each INSERT operation caused a 30MB reply. And when you have high enough traffic, that adds up quickly.

On a technical level, it was an easy fix. On a practical one, it took six weeks to coordinate with the web dev company and their hosting setup to make the change, test the change, and deploy the change. Two of those weeks were simply spent convincing the company that yes, this was in fact happening, and yes, it was in fact their fault.

[Advertisement] Keep the plebs out of prod. Restrict NuGet feed privileges with ProGet. Learn more.

The+Daily+WTF
CodeSOD: Identify a Nap 22 September 2025 at 06:30

CodeSOD: Identify a Nap

The+Daily+WTF

By: Remy Porter

22 September 2025 at 06:30

Guy picked up a bug ticket. There was a Hiesenbug; sometimes, saving a new entry in the application resulted in a duplicate primary key error, which should never happen.

The error was in the message-bus implementation someone else at the company had inner-platformed together, and it didn't take long to understand why it failed.

/**
 * This generator is used to generate message ids.
 * This implementation merely returns the current timestamp as long.
 *
 * We are, thus, limited to insert 1000 new messages per second.
 * That throughput seems reasonable in regard with the overall
 * processing of a ticket.
 *
 * Might have to re-consider that if needed.
 *
 */
public class IdGenerator implements IdentifierGenerator
{

        long previousId;
       
        @Override
        public synchronized Long generate (SessionImplementor session, Object parent) throws HibernateException {
                long newId = new Date().getTime();
                if (newId == previousId) {
                        try { Thread.sleep(1); } catch (InterruptedException ignore) {}
                        newId = new Date().getTime();
                }
                return newId;
        }
}

This generates IDs based off of the current timestamp. If too many requests come in and we start seeing repeating IDs, we sleep for a second and then try again.

This… this is just an autoincrementing counter with extra steps. Which most, but I suppose not all databases supply natively. It does save you the trouble of storing the current counter value outside of a running program, I guess, but at the cost of having your application take a break when it's under heavier than average load.

One thing you might note is absent here: generate doesn't update previousId. Which does, at least, mean we won't ever sleep for a second. But it also means we're not doing anything to avoid collisions here. But that, as it turns out, isn't really that much of a problem. Why?

Because this application doesn't just run on a single server. It's distributed across a handful of nodes, both for load balancing and resiliency. Which means even if the code properly updated previousId, this still wouldn't prevent collisions across multiple nodes, unless they suddenly start syncing previousId amongst each other.

I guess the fix might be to combine a timestamp with something unique to each machine, like… I don't know… hmmm… maybe the MAC address on one of their network interfaces? Oh! Or maybe you could use a sufficiently large random number, like really large. 128-bits or something. Or, if you're getting really fancy, combine the timestamp with some randomness. I dunno, something like that really sounds like it could get you to some kind of universally unique value.

Then again, since the throughput is well under 1,000 messages per second, you could probably also just let your database handle it, and maybe not generate the IDs in code.

[Advertisement] Keep the plebs out of prod. Restrict NuGet feed privileges with ProGet. Learn more.

The+Daily+WTF
Error'd: You Talkin' to Me? 19 September 2025 at 06:30

Error'd: You Talkin' to Me?

The+Daily+WTF

By: Lyle Seaman

19 September 2025 at 06:30

The Beast In Black is back with a simple but silly factual error on the part of the gateway to all (most) human knowledge.

B.J.H. "The old saying is "if you don't like the weather wait five minutes". Weather.com found a time saver." The trick here is to notice that the "now" temperature is not the same as the headline temperature, also presumably now.

"That's some funny math you got there. Be a shame if it was right," says Jason . "The S3 bucket has 10 files in it. Picking any two (or more) causes the Download button to go disabled with this message when moused over. All I could think of is that this S3 bucket must be in the same universe as https://thedailywtf.com/articles/free-birds " Alas, we are all in the same universe as https://thedailywtf.com/articles/free-birds .

"For others, the markets go up and down, but me, I get real dividends!" gloats my new best friend Mr. TA .

David B. is waiting patiently. "Somewhere in the USPS a package awaits delivery. Either rain, nor snow, nor gloom of night shall prevent the carrier on their appointed rounds. When these rounds will occur are not the USPS's problem." We may not know the day, but we know the hour!

[Advertisement] Keep the plebs out of prod. Restrict NuGet feed privileges with ProGet. Learn more.

The+Daily+WTF
CodeSOD: An Echo In Here in here 18 September 2025 at 06:30

CodeSOD: An Echo In Here in here

The+Daily+WTF

By: Remy Porter

18 September 2025 at 06:30

Tobbi sends us a true confession: they wrote this code.

The code we're about to look at is the kind of code that mixes JavaScript and PHP together, using PHP to generate JavaScript code. That's already a terrible anti-pattern, but Tobbi adds another layer to the whole thing.


if (AJAX)
{
    <?php
        echo "AJAX.open(\"POST\", '/timesheets/v2/rapports/FactBCDetail/getDateDebutPeriode.php', true);";
            
    ?>
    
    AJAX.setRequestHeader("Content-type", "application/x-www-form-urlencoded");
    AJAX.onreadystatechange = callback_getDateDebutPeriode;
    AJAX.send(strPostRequest);
}

if (AJAX2)
{
    <?php
        echo "AJAX2.open(\"POST\", '/timesheets/v2/rapports/FactBCDetail/getDateFinPeriode.php', true);";
    ?>
    AJAX2.setRequestHeader("Content-type", "application/x-www-form-urlencoded");
    AJAX2.onreadystatechange = callback_getDateFinPeriode;
    AJAX2.send(strPostRequest);
}

So, this uses server side code to… output string literals which could have just been written directly into the JavaScript without the PHP step.

"What was I thinking when I wrote that?" Tobbi wonders. Likely, you weren't thinking, Tobbi. Have another cup of coffee, I think you need it.

All in all, this code is pretty harmless, but is a malodorous brain-fart. As for absolution: this is why we have code reviews. Either your org doesn't do them, or it doesn't do them well. Anyone can make this kind of mistake, but only organizational failures get this code merged.

[Advertisement] Keep the plebs out of prod. Restrict NuGet feed privileges with ProGet. Learn more.

The+Daily+WTF
Representative Line: Brace Yourself 17 September 2025 at 06:30

Representative Line: Brace Yourself

The+Daily+WTF

By: Remy Porter

17 September 2025 at 06:30

Today's representative line is almost too short to be a full line. But I haven't got a category for representative characters, so we'll roll with it. First, though, we need the setup.

Brody inherited a massive project for a government organization. It was the kind of code base that had thousands of lines per file, and frequently thousands of lines per function. Almost none of those lines were comments. Almost.

In the middle of one of the shorter functions (closer to 500 lines), Brody found this:

//    }

This was the only comment in the entire file. And it's a beautiful one, because it tells us so much. Specifically, it tells us the developer responsible messed up the brace count (because clearly a long function has loads of braces in it), and discovered their code didn't compile. So they went around commenting out extra braces until they found the offender. Code compiled, and voila- on to the next bug, leaving the comment behind.

Now, I don't know for certain that's why a single closing brace is commented out. But also, I know for certain that's what happened, because I've seen developers do exactly that.

[Advertisement] Utilize BuildMaster to release your software with confidence, at the pace your business demands. Download today!

The+Daily+WTF
Representative Line: Reduced to a Union 16 September 2025 at 06:30

Representative Line: Reduced to a Union

The+Daily+WTF

By: Remy Porter

16 September 2025 at 06:30

The code Clemens M supported worked just fine for ages. And then one day, it broke. It didn't break after a deployment, which implied some other sort of bug. So Clemens dug in, playing the game of "what specific data rows are breaking the UI, and why?"

One of the organizational elements of their system was the idea of "zones". I don't know the specifics of the application as a whole, but we can broadly describe it thus:

The application oversaw the making of widgets. Widgets could be assigned to one or more zones. A finished product requires a set of widgets. Thus, the finished product has a number of zones that's the union of all of the zones of its component widgets.

Which someone decided to handle this way:

zones.reduce((accumulator, currentValue) => accumulator = _.union(currentValue))

So, we reduce across zones (which is an array of arrays, where the innermost arrays contain zone names, like zone-0, zone-1). In each step we union it with… nothing. The LoDash union function expects an array of arrays, and returns an array that's the union of all its inputs. This isn't how that function is meant to be used, but the behavior from this incorrect usage was that accumulator would end up holding the last element in zones. Which actually worked until recently, because until recently no one was splitting products across zones. When all the inputs were in the same zone, grabbing the last one was just fine.

The code had been like this for years. It was only just recently, as the company expanded, that it became problematic. The fix, at least, was easy- drop the reduce and just union correctly.

[Advertisement] ProGet’s got you covered with security and access controls on your NuGet feeds. Learn more.

The+Daily+WTF
CodeSOD: Functionally, a Date 15 September 2025 at 06:30

CodeSOD: Functionally, a Date

The+Daily+WTF

By: Remy Porter

15 September 2025 at 06:30

Dates are messy things, full of complicated edge cases and surprising ways for our assumptions to fail. They lack the pure mathematical beauty of other data types, like integers. But that absence doesn't mean we can't apply the beautiful, concise, and simple tools of functional programming to handling dates.

I mean, you or I could. J Banana's co-worker seems to struggle a bit with it.

/**
 * compare two dates, rounding them to the day
 */
private static int compareDates( LocalDateTime date1, LocalDateTime date2 ) {
    List<BiFunction<LocalDateTime,LocalDateTime,Integer>> criterias = Arrays.asList(
            (d1,d2) -> d1.getYear() - d2.getYear(),
            (d1,d2) -> d1.getMonthValue() - d2.getMonthValue(),
            (d1,d2) -> d1.getDayOfMonth() - d2.getDayOfMonth()
        );
    return criterias.stream()
        .map( f -> f.apply(date1, date2) )
        .filter( r -> r != 0 )
        .findFirst()
        .orElse( 0 );
}

This Java code creates a list containing three Java functions. Each function will take two dates and returns an integer. It then streams that list, applying each function in turn to a pair of dates. It then filters through the list of resulting integers for the first non-zero value, and failing that, returns just zero.

Why three functions? Well, because we have to check the year, the month, and the day. Obviously. The goal here is to return a negative value if date1 preceeds date2, zero if they're equal, and positive if date1 is later. And on this metric… it does work. That it works is what makes me hate it, honestly. This not only shouldn't work, but it should make the compiler so angry that the computer gets up and walks away until you've thought about what you've done.

Our submitter replaced all of this with a simple:

return date1.toLocalDate().compareTo( date2.toLocalDate() );

[Advertisement] ProGet’s got you covered with security and access controls on your NuGet feeds. Learn more.

The+Daily+WTF
Error'd: Free Birds 12 September 2025 at 06:30

Error'd: Free Birds

The+Daily+WTF

By: Lyle Seaman

12 September 2025 at 06:30

"These results are incomprensible," Brian wrote testily. "The developers at SkillCertPro must use math derived from an entirely different universe than ours. I can boast a world record number of answered questions in one hour and fifteen minutes somewhere."

"How I Reached Inbox -1," Maia titled her Tickity Tock. "Apparently I've read my messages so thoroughly that my email client (Mailspring) time traveled into the future and read a message before it was even sent."

... which taught Jason how to use Mailspring to order timely tunes. "Apparently, someone invented a time machine and is able to send us vinyls from the future..."

"Yes, we have no bananas," sang out Peter G. , rapping "... or email addresses or phone numbers, but we're going to block your post just the same (and this is better than the previous result of "Whoops something went wrong", because you'd never be able to tell something had gone wrong without that helpful message)."

Finally, our favorite cardsharp Adam R. might have unsharp eyes but sharp browser skills. "While reading an online bridge magazine, I tried to zoom out a bit but was dismayed to find I couldn't zoom out. Once it zooms in to NaN%, you're stuck there."

[Advertisement] BuildMaster allows you to create a self-service release management platform that allows different teams to manage their applications. Explore how!

The+Daily+WTF
CodeSOD: The Getter Setter Getter 11 September 2025 at 06:30

CodeSOD: The Getter Setter Getter

The+Daily+WTF

By: Remy Porter

11 September 2025 at 06:30

Today's Java snippet comes from Capybara James.

The first sign something was wrong was this:

private Map<String, String> getExtractedDataMap(PayloadDto payload) {
    return setExtractedDataToMap(payload);
}

Java conventions tell us that a get method retrieves a value, and a set method mutates the value. So a getter that calls a setter is… confusing. But neither of these are truly getters nor setters.

setExtractedDataToMap converts the PayloadDto to a Map<String, String>. getExtractedMap just calls that, which is just one extra layer of indirection that nobody needed, but whatever. At its core, this is just two badly named methods where there should be one.

But that distracts from the true WTF in here. Why on Earth are we converting an actual Java object to a Map<String,String>? That is a definite code smell, a sign that someone isn't entirely comfortable with object-oriented programming. You can't even say, "Well, maybe for serialization to JSON or something?" because Java has serializers that just do this transparently. And that's just the purpose of a DTO in the first place- to be a bucket that holds data for easy serialization.

We're left wondering what the point of all of this code is, and we're not alone. James writes:

I found this gem of a code snippet while trying to understand a workflow for data flow documentation purpose. I was not quite sure what the original developer was trying to achieve and at this point I just gave up

[Advertisement] Picking up NuGet is easy. Getting good at it takes time. Download our guide to learn the best practice of NuGet for the Enterprise.

The+Daily+WTF
CodeSOD: Upsert Yours 10 September 2025 at 06:30

CodeSOD: Upsert Yours

The+Daily+WTF

By: Remy Porter

10 September 2025 at 06:30

Henrik H sends us a short snippet, for a relative value of short.

We've all seen this method before, but this is a particularly good version of it:

public class CustomerController
{
    public void MyAction(Customer customer)
    {
        // snip 125 lines

        if (customer.someProperty)
            _customerService.UpsertSomething(customer.Id, 
            customer.Code, customer.Name, customer.Address1, 
            customer.Address2, customer.Zip, customer.City, 
            customer.Country, null, null, null, null, null, 
            null, null, null, null, null, null, null, null, 
            null, false, false, null, null, null, null, null, 
            null, null, null, null, null, null, null, false, 
            false, false, false, true, false, null, null, null,
            false, true, false, true, true, 0, false, false, 
            false, false, customer.TemplateId, false, false, false, 
            false, false, string.Empty, true, false, false, false, 
            false, false, false, false, false, true, false, false, 
            true, false, false, MiscEnum.Standard, false, false, 
            false, true, null, null, null);
        else
            _customerService.UpsertSomething(customer.Id, 
            customer.Code, customer.Name, customer.Address1, 
            customer.Address2, customer.Zip, customer.City, 
            customer.Country, null, null, null, null, null, 
            null, null, null, null, null, null, null, null, 
            null, false, false, null, null, null, null, null, 
            null, null, null, null, null, null, null, false, 
            false, false, false, true, false, null, null, null, 
            false, false, false, true, true, 0, false, false, 
            false, false, customer.TemplateId, false, false, false, 
            false, false, string.Empty, true, false, false, false, 
            false, false, false, false, true, true, false, false, 
            true, false, false, MiscEnum.Standard, false, false, 
            false, true, null, null, null);

        // snip 52 lines
    }
}

Welcome to the world's most annoying "spot the difference" puzzle. I've added line breaks (as each UpsertSomething was all on one line in the original) to help you find it. Here's a hint: it's one of the boolean values. I'm sure that narrows it down for you. It means the original developed didn't need the if/else and instead could have simply passed customer.someProperty as a parameter.

Henrick writes:

While on a simple assignment to help a customer migrate from .NET Framework to .NET core, I encountered this code. The 3 lines are unfortunately pretty representative for the codebase

[Advertisement] BuildMaster allows you to create a self-service release management platform that allows different teams to manage their applications. Explore how!

The+Daily+WTF
Myopic Focus 9 September 2025 at 06:30

Myopic Focus

The+Daily+WTF

By: Ellis Morning

9 September 2025 at 06:30

Chops was a developer for Initrode. Early on a Monday, they were summoned to their manager Gary's office before the caffeine had even hit their brain.

Gary glowered up from his office chair as Chops entered. This wasn't looking good. "We need to talk about the latest commit for Taskmaster."

Taskmaster was a large application that'd been around for decades, far longer than Chops had been an employee. Thousands of internal and external customers relied upon it. Refinements over time had led to remarkable stability, its typical uptime now measured in years. However, just last week, their local installation had unexpectedly suffered a significant crash. Chops had been assigned to troubleshooting and repair.

"What's wrong?" Chops asked.

"Your latest commit decreased the number of unit tests!" Gary replied as if Chops had slashed the tires on his BMW.

Within Taskmaster, some objects that were periodically generated were given a unique ID from a pool. The pool was of limited size and required scanning to find a spare ID. Each time a value was needed, a search began where the last search ended. IDs returned to the pool as objects were destroyed would only be reused when the search wrapped back around to the start.

Chops had discovered a bug in the wrap-around logic that would inevitably produce a crash if Taskmaster ran long enough. They also found that if the number of objects created exceeded the size of the pool, this would trigger an infinite loop.

Rather than attempt to patch any of this, Chops had nuked the whole thing and replaced it with code that assigned each object a universally unique identifier (UUID) from a trusted library UUID generator within its constructor. Gone was the bad code, along with its associated unit tests.

Knowing they would probably only get in a handful of words, Chops wonderered how on earth to explain all this in a way that would appease their manager. "Well—"

"That number must NEVER go down!" Gary snapped.

"But—"

"This is non-negotiable! Roll it back and come up with something better!"

And so Chops had no choice but to remove their solution, put all the janky code back in place, and patch over it with kludge. Every comment left to future engineers contained a tone of apology.

Taskmaster became less stable. Time and expensive developer hours were wasted. Risk to internal and external customers increased. But Gary could rest assured, knowing that his favored metric never faltered on his watch.

[Advertisement] Keep all your packages and Docker containers in one place, scan for vulnerabilities, and control who can access different feeds. ProGet installs in minutes and has a powerful free version with a lot of great features that you can upgrade when ready.Learn more.

IT+Notes
FreeBSD vs. SmartOS: Who's Faster for Jails, Zones, and bhyve VMs? 19 September 2025 at 08:50

FreeBSD vs. SmartOS: Who's Faster for Jails, Zones, and bhyve VMs?

IT+Notes

By: Stefano Marinelli

19 September 2025 at 08:50

Which virtualization host performs better? I put FreeBSD and SmartOS in a head-to-head showdown. The performance of Jails, Zones, and bhyve VMs surprised me, forcing a second round of tests on different hardware to find the real winner.

Pixel Envy
That Secret Service SIM Farm Story Is Bogus 28 September 2025 at 05:01

That Secret Service SIM Farm Story Is Bogus

Pixel Envy

By: Nick Heer

28 September 2025 at 05:01

Robert Graham, clarifying the bad reporting of the big SIM farm bust in New York:

The Secret Service is lying to the press. They know it’s just a normal criminal SIM farm and are hyping it into some sort of national security or espionage threat. We know this because they are using the correct technical terms that demonstrate their understanding of typical SIM farm crimes. The claim that they will likely find other such SIM farms in other cities likewise shows they understand this is a normal criminal activity and not any special national security threat.

One of the things we must always keep in mind is that press releases are written to persuade. That is as true for businesses as it is for various government agencies. In this case, the Secret Service wanted attention, so they exaggerated the threat. And one wonders why public trust in institutions is falling.

⌥ Permalink

Pixel Envy
Google Provides Feedback on the Digital Markets Act 27 September 2025 at 05:27

Google Provides Feedback on the Digital Markets Act

Pixel Envy

By: Nick Heer

27 September 2025 at 05:27

Something I missed in posting about Apple’s critical appraisal of the Digital Markets Act is its timing. Why now? Well, it turns out the European Commission sought feedback beginning in July, and with a deadline of just before midnight on 24 September. That is why it published that statement, and why Google did the same.

Oliver Bethell, Google’s “senior director, competition”, a job title which implies a day spent chuckling to oneself:

Consider the DMA’s impact on Europe’s tourism industry. The DMA requires Google Search to stop showing useful travel results that link directly to airline and hotel sites, and instead show links to intermediary websites that charge for inclusion. This raises prices for consumers, reduces traffic to businesses, and makes it harder for people to quickly find reliable, direct booking information.

Key parts of the European tourism industry have already seen free, direct booking traffic from Google Search plummet by up to 30%. A recent study on the economic impact of the DMA estimates that European businesses across sectors could face revenue losses of up to €114 billion.

The study in question, though published by Copenhagen Business School, was funded by the Computer & Communications Industry Association, a tech industry lobbying firm funded in part by Google. I do not have the background to assess if the paper’s conclusions are well-founded, but it should be noted the low-end of the paper’s estimates was a loss of €8.5 billion, or just 0.05% of total industry revenue (page 45). The same lobbyists also funded a survey (PDF) conducted online by Nextrade Group.

Like Apple, Google clearly wants this law to go away. It might say it “remain[s] committed to complying with the DMA” and that it “appreciate[s] the Commission’s consistent openness to regulatory dialogue”, but nobody is fooled. To its credit, Google posted the full response (PDF) it sent the Commission which, though clearly defensive, has less of a public relations sheen than either of the company’s press releases.

⌥ Permalink

Pixel Envy
U.S. Federal Trade Commission Settles With Amazon Just Two Days Into Trial 26 September 2025 at 04:53

U.S. Federal Trade Commission Settles With Amazon Just Two Days Into Trial

Pixel Envy

By: Nick Heer

26 September 2025 at 04:53

In 2023 Lina Khan, then-chair of the U.S. Federal Trade Commission, sued Amazon over using (PDF) “manipulative, coercive, or deceptive user-interface designs known as ‘dark patterns’ to trick consumers into enrolling in automatically-renewing Prime subscriptions” and “knowingly complicat[ing] the cancellation process”. Some people thought this case was a long-shot, or attempted to use Khan’s scholarship against her.

Earlier this week, the trial began to adjudicate the government’s claims which, in addition to accusing Amazon itself, also involved charges against company executives. It was looking promising for the FTC.

Annie Palmer, CNBC:

The FTC notched an early win in the case last week when U.S. District Court Judge John Chun ruled Amazon and two senior executives violated the Restore Online Shoppers’ Confidence Act by gathering Prime members’ billing information before disclosing the terms of the service.

Chun also said that the two senior Amazon executives would be individually liable if a jury sides with the FTC due to the level of oversight they maintained over the Prime enrollment and cancellation process.

Then, just two days into the trial, the FTC announced it had reached a settlement:

The Federal Trade Commission has secured a historic order with Amazon.com, Inc., as well as Senior Vice President Neil Lindsay and Vice President Jamil Ghani, settling allegations that Amazon enrolled millions of consumers in Prime subscriptions without their consent, and knowingly made it difficult for consumers to cancel. Amazon will be required to pay a $1 billion civil penalty, provide $1.5 billion in refunds back to consumers harmed by their deceptive Prime enrollment practices, and cease unlawful enrollment and cancellation practices for Prime.

As usual for settlements like these, Amazon will admit no wrongdoing. The executives will not face liability, something Adam Kovacevich, head of the Chamber of Progress, a tech industry lobbying group, said today was a “wild … theory” driven by “Khan’s ego”. Nonsense. The judge in the case, after saying Amazon broke the law, gave credence to the concept these executives were personally liable for the harm they were alleged to have caused.

Former FTC commissioner Alvaro Bedoya on X:

Based on my initial read, do the executives need to do anything separate from that? Do they pay any fines? Are they being demoted? Are they subject to extra monitoring? Do they need to admit any guilt whatsoever? The answers, as far as I can tell are no, no, no, no, and no. What’s worse, the order applies to the executives for only three years — seven years less than the company.

Two-and-a-half billion is a lot of dollars in the abstract. CIRP estimates there are 197 million U.S. subscribers to Amazon Prime, which costs anywhere from $7 to $15 per month. For the sake of argument, assume everyone is — on average — on the annual plan of $11.58 per month. It will take barely more than one billing cycle for Amazon to recoup that FTC settlement. The executives previously charged will bear little responsibility for this outcome.

Those million-dollar inauguration “investments”, as CBS News put it, sure are paying off.

⌥ Permalink

Pixel Envy
Privacy Commissioner of Canada Releases Findings of Investigation Into TikTok 26 September 2025 at 04:10

Privacy Commissioner of Canada Releases Findings of Investigation Into TikTok

Pixel Envy

By: Nick Heer

26 September 2025 at 04:10

Catharine Tunney, CBC News:

The immensely popular social media app TikTok has been collecting sensitive information from hundreds of thousands of Canadians under 13 years old, a joint investigation by privacy authorities found.

[…]

The privacy commissioners said TikTok agreed to enhance its age verification and provide up-front notices about its wide-ranging collection of data.

Off the top, the Privacy Commissioner’s report was limited in scope and did not examine “perceived risks to national security” since they were not related to “privacy in the context of commercial activity” and have been adjudicated elsewhere. The results of national security reviews by other agencies have not been published. However, the Commissioner’s review of the company’s privacy practices is still comprehensive for what was in scope.

TikTok detects and removes about 500,000 accounts of Canadian children under 13 annually. Yet even though the company has dedicated significant engineering efforts to estimating users’ ages for advertising and to produce recommendations, it has not developed similar capabilities for restricting minors’ access.

Despite my skepticism of the Commissioner’s efficacy in cases like these, this investigation produced a number of results. TikTok made several changes as the investigation progressed, including restricting ad targeting to minors:

As an additional measure, in its response to the Offices’ Preliminary Report of Investigation, TikTok committed to limit ad targeting for users under 18 in Canada. TikTok informed the Offices that it implemented this change on April 1st, 2025. As a result, advertisers can no longer deliver targeted ads to users under 18, other than according to generic data (such as language and approximate location).

This is a restriction TikTok has in place for some regions, but not everywhere. It is not unique to TikTok, either; Meta and Google targeted minors, and Meta reportedly guessed teens’ emotional state for ad targeting purposes. This industry cannot police itself. All of these companies say they have rules against ad targeting to children and have done so for years, yet all of them have been found to ignore those rules when they are inconvenient.

⌥ Permalink

Pixel Envy
Apple Attempts to Rally Users Against E.U. Digital Markets Act 25 September 2025 at 22:14

Apple Attempts to Rally Users Against E.U. Digital Markets Act

Pixel Envy

By: Nick Heer

25 September 2025 at 22:14

Apple issued a press release criticizing the E.U.’s Digital Markets Act in a curious mix of countries. It published it on its European sites — of course — and in Australia, Canada, New Zealand, and the United States, all English-speaking. It also issued the same press release in Brazil, China, Guinea-Bissau, Indonesia, and Thailand — and a handful of other places — but not in Argentina, India, Japan, Mexico, or Singapore. Why this mix? Why did Apple bother to translate it into Thai but not Japanese? It is a fine mystery. Read into it what you will.

Anyway, you will be amazed to know how Apple now views the DMA:

It’s been more than a year since the Digital Markets Act was implemented. Over that time, it’s become clear that the DMA is leading to a worse experience for Apple users in the EU. It’s exposing them to new risks, and disrupting the simple, seamless way their Apple products work together. And as new technologies come out, our European users’ Apple products will only fall further behind.

[…]

That’s why we’re urging regulators to take a closer look at how the law is affecting the EU citizens who use Apple products every day. We believe our users in Europe deserve the best experience on our technology, at the same standard we provide in the rest of the world — and that’s what we’ll keep fighting to deliver.

It thinks the DMA should disappear.

Its reasoning is not great; Michael Tsai read the company’s feature delays more closely and is not convinced. One of the delayed features is Live Translation, about which I wrote:

This is kind of a funny limitation because fully half the languages Live Translation works with — French, German, and Spanish — are the versions spoken in their respective E.U. countries and not, for example, Canadian French or Chilean Spanish. […]

Because of its launch languages, I think Apple expects this holdup will not last for long.

I did not account for a cynical option: Apple is launching with these languages as leverage.

The way I read Apple’s press release is as a fundamental disagreement between the role each party believes it should play, particularly when it comes to user privacy. Apple seems to believe it is its responsibility to implement technical controls to fulfill its definition of privacy and, if that impacts competition and compatibility, too bad. E.U. regulators seem to believe it has policy protections for user privacy, and that users should get to decide how their private data is shared.

Adam Engst, TidBits:

Apple’s claim of “the same standard we provide in the rest of the world” rings somewhat hollow, given that it often adjusts its technology and services to comply with local laws. The company has made significant concessions to operate in China, doesn’t offer FaceTime in the United Arab Emirates, and removes apps from the still-functional Russian App Store at the Russian government’s request. Apple likely pushed back in less public ways in those countries, but in the EU, this public statement appears aimed at rallying its users and influencing the regulatory conversation.

I know what Engst is saying here, and I agree with the sentiment, but this is a bad group of countries to be lumped in together with. That does not mean the DMA is equal to the kinds of policies that restrict services in these other countries. It remains noteworthy how strict Apple is in restricting DMA-mandated features only to countries where they are required, but you can just change your region to work around the UAE FaceTime block.

⌥ Permalink

Pixel Envy
Australian Opposition Parties Encourage Migration to Forthcoming U.S. Version of TikTok 25 September 2025 at 15:21

Australian Opposition Parties Encourage Migration to Forthcoming U.S. Version of TikTok

Pixel Envy

By: Nick Heer

25 September 2025 at 15:21

Oscar Godsell, Sky News:

The opposition’s shadow finance minister James Paterson has since urged the Australian Labor government to follow suit.

Mr Paterson told Sky News if the US was able to create a “safer version” of TikTok, then Australia should liaise with the Trump administration to become part of that solution.

“It would be an unfortunate thing if there was a safe version of TikTok in the United States, but a version of TikTok in Australia which was still controlled by a foreign authoritarian government,” he said.

I am not sure people in Australia are asking for yet more of the country’s media to be under the thumb of Rupert Murdoch. Then again, I also do not think the world needs more social media platforms controlled by the United States, though that is very clearly the wedge the U.S. government is creating: countries can accept the existing version of TikTok, adopt the new U.S.-approved one, or ban them both. The U.S. spinoff does not resolve user privacy problems and it raises new concerns about the goals of its government-friendly ownership and management.

⌥ Permalink

Pixel Envy
Patreon Will Automatically Enable Audience Irritation Features Next Week 25 September 2025 at 03:53

Patreon Will Automatically Enable Audience Irritation Features Next Week

Pixel Envy

By: Nick Heer

25 September 2025 at 03:53

Do you manage a Patreon page as a “creator”? I do; it is where you can give me five dollars per month to add to my guilt over not finishing my thoughts about Liquid Glass.¹ You do not have to give me five dollars. I feel guilty enough as it is.

Anyway, you might have missed an email Patreon sent today advising you that Autopilot will be switched on beginning October 1 unless you manually turn it off. According to Patreon’s email:

Autopilot is a growth engine that automatically sends your members and potential members strategic, timely offers which encourage them to join, upgrade, or retain your membership — without you having to lift a finger.

As an extremely casual user, I do not love this; I think it is basically spam. I am sympathetic toward those who make their living with Patreon. I turned this off. If you have a Patreon creator page and missed this email, now you know.

And if you are a subscriber to anyone on Patreon and begin receiving begging emails next week, please be gracious. They might not be aware this feature was switched on.

I am most looking forward to reading others’ reviews when I am done, which I have so far avoided so my own piece is not tainted. ↥︎

⌥ Permalink

Pixel Envy
Setting Up a New Apple TV Is Still Not Good 25 September 2025 at 03:35

Setting Up a New Apple TV Is Still Not Good

Pixel Envy

By: Nick Heer

25 September 2025 at 03:35

Tonight, I set up a new Apple TV — well, as “new” as a refurbished 2022-though-still-current-generation model can be — and it was not a good time. I know Apple might be releasing a new model later this year, but any upgrades are probably irrelevant for how I have used my existing ten-year-old model. I do not even have a 4K television.

My older model has some drawbacks. It is pretty slow, and the storage space is pitiful — I think it is the 32 GB model — so it keeps offloading apps. What I wanted to do was get a new one and bump the old Apple TV to my kitchen, where I have a receiver and a set of speakers I have used with Bluetooth, and then I would be able to AirPlay music in all my entertaining spaces. Real simple stuff.

Jason Snell, in a sadly still-relevant Six Colors article:

The setup starts promisingly: You can bring your iPhone near the Apple TV, and it will automatically log your Apple ID in. If you’ve got the One Home Screen feature turned on, all your apps will load and appear in all the right places. It will feel like you’ve done a data transfer.

But it’s all a mirage.

One Home Screen is a nice feature, but it’s not an iCloud backup of your Apple TV, nor is it the Apple TV equivalent of Migration Assistant. It is exactly what its name suggests — a home-screen-syncing feature and nothing more.

I went into this upgrade realizing my wife and I would need to set up all our streaming apps again. (She was cool with it.) That is not great, but at least I had that expectation.

But even the “promising” parts of the setup experience did not work for me. When I brought my iPhone near the new Apple TV, it spun before throwing a mysterious error. After setting it up manually, it thought it was not connected to Wi-Fi — even though it was — and then it tried syncing the home screen. Some of the apps are right, but it has not synced all of them, and none of them are in the correct position.

Then I opened Music on my phone to try and AirPlay to both Apple TVs, only to find it was not listed. It turns out that is a separate step. I had to add it to my Home, which again involved me bringing my iPhone into close proximity and tapping a button. This failed the three times I tried it. So I restarted my Apple TV and my phone, and then Settings told me I needed to complete my Home setup. I guess it worked but somehow did not move to the next step. At last, AirPlay worked — and, frankly, it is pretty great.

I know bugs happen about as often as blog posts complaining about bugs. This thing is basically an appliance, though. I am glad Apple ultimately did not make a car.

⌥ Permalink

Pixel Envy
U.S. Secret Service Busts Giant SIM Farm in New York 23 September 2025 at 23:53

U.S. Secret Service Busts Giant SIM Farm in New York

Pixel Envy

By: Nick Heer

23 September 2025 at 23:53

The U.S. Secret Service:

The U.S. Secret Service dismantled a network of electronic devices located throughout the New York tristate area that were used to conduct multiple telecommunications-related threats directed towards senior U.S. government officials, which represented an imminent threat to the agency’s protective operations.

This protective intelligence investigation led to the discovery of more than 300 co-located SIM servers and 100,000 SIM cards across multiple sites.

That sure is a lot of SIM cards, and a scary-sounding mix of words in the press release:

“[…] telecommunications-related threats directed towards senior U.S. government officials […]”
“[…] these devices could be used to conduct a wide range of telecommunications attacks […]”
“These devices were concentrated within 35 miles of the global meeting of the United Nations General Assembly […]”

Reporters pounced. The New York Times, NBC News, CBS News, and even security publications like the Record seized on dramatic statements like those, and another said by the special agent in a video the Service released: “this network had the potential to […] essentially shut down the cellular network in New York City”. Scary stuff.

When I read the early reports, it sure looked to me like some reporters were getting a little over their skis.

For a start, emphasizing the apparent proximity to the U.N. in New York seems to me like a stretch. A thirty-five mile area around the U.N. looks like this — and that is diameter, not radius. If you cannot see that or this third-party website goes away at some point, that is a circle encompassing just about the entire island of Manhattan, going deep into Brooklyn and Queens, stretching all the way up to Chappaqua, and out into Connecticut and New Jersey. That is a massive area. One could just as easily say it was within thirty-five miles of any number of New York-based landmarks and be just as accurate.

Second, the ability to “facilitat[e] anonymous, encrypted communication between potential threat actors and criminal enterprises” is common to basically any internet-connected device. The scale of this one is notable, but you do not need a hundred-thousand SIM cards to make criminal plans. And the apparent possibility of “shut[ting] down the cellular network in New York” is similarly common to any large-scale installation. This is undeniably peculiar, huge, and it seems to be nefarious, but a lot of this seems to be a red herring.

Andy Greenberg, Lily Hay Newman, and Matt Burgess, Wired:

Despite speculation in some reporting about SIM farm operation that suggests it was created by a foreign state such as Russia or China and used for espionage, it’s far more likely that the operation’s central focus was scams and other profit-motivated forms of cybercrime, says Ben Coon, who leads intelligence at the cybersecurity firm Unit 221b and has carried out multiple investigations into SIM farms. “The disruption of cell services is possible, flooding the network to the degree that it couldn’t take any more traffic,” Coon says. “My gut is telling me there was some type of fraud involved here.”

These reporters point to a CNN article by John Miller and Celina Tebor elaborating on the threat to “senior U.S. government officials”: they were swatting calls targeting various lawmakers. Not nothing and certainly dangerous, but this is not looking anything like how many reporters have described it, nor what the U.S. Secret Service is suggesting through its word choices.

⌥ Permalink

Pixel Envy
‘How A.I. Helped Locate a Viral Video’s True Origin’ Was by Ignoring A.I. 23 September 2025 at 23:23

‘How A.I. Helped Locate a Viral Video’s True Origin’ Was by Ignoring A.I.

Pixel Envy

By: Nick Heer

23 September 2025 at 23:23

This story of how Full Fact geolocated a viral video claiming to be shot in London is intriguing because it disproves its own headline’s claim that “A.I. helped”.

Charlotte Green, Full Fact:

But in this case, directly reverse image searching through Google took me to a TikTok video with a location marker for ‘Pondok Pesantren Al Fatah Temboro’, in Indonesia.

This is enough information to give the Full Fact team a great start: translated, it is a school in Temboro.

Green:

We found a slightly different compilation of similar videos on Facebook, seemingly from the same area, also with women in Islamic dress, but with more geographical features visible, such as a sign and clearer views of buildings.

Using stills from this video as references, we asked the AI chatbot ChatGPT if it could provide coordinates to the location, using the possible location of the Al Fatah school in Indonesia.

Up to the point where ChatGPT was invoked, there is no indication any A.I. tools were used. After that — and I do not intend to be mean — it is unclear to me why anyone would ask ChatGPT for coordinates to a known, named location when you can just search Google Maps. It is the third one down in my searches; the first two would quickly be eliminated when comparing to either video.

Green:

But this did not match the location of the original video we were trying to fact check—or anywhere in the near vicinity. While we were very confident the video had been filmed in Temboro, we needed to investigate further to prove this.

After this, no A.I. tools were used. ChatGPT was only able to do as much as a basic Google Maps search. After that, Full Fact had to do some old-fashioned comparative geolocation, and were ultimately successful.

I found this via Charles Arthur, who writes:

And thus we see the positive uses of geolocation by chatbots.

On the contrary, this proved little about the advantages of A.I. geolocation. These tools can certainly be beneficial; Green links to an experiment in Bellingcat in comparison to Google’s reverse image search tools.

I think Full Fact did great work in geolocating this video and deflating its hateful context in that tweet. But a closer reading of the actual steps taken shows any credit to ChatGPT or A.I. is overblown.

⌥ Permalink

Pixel Envy
Amazon to End Inventory Commingling 22 September 2025 at 23:38

Amazon to End Inventory Commingling

Pixel Envy

By: Nick Heer

22 September 2025 at 23:38

Allison Smith, Modern Retail (via Michael Tsai):

Amazon revealed at its annual Accelerate seller conference in Seattle that it is shutting down its long-running “commingling” program — a move that drew louder applause from sellers than any other update of the morning.

The decision marks the end of a controversial practice in which Amazon pooled identical items from different sellers under one barcode. The system, intended to speed deliveries and save warehouse space, had also allowed counterfeit or expired goods to be mixed in with authentic ones, according to The Wall Street Journal. For years, brands complained that commingling made it difficult to trace problems back to specific sellers and left their reputations vulnerable when customers received knockoffs. In 2013, Johnson & Johnson temporarily pulled many of its consumer products from Amazon, arguing the retailer wasn’t doing enough to curb third-party sales of damaged or expired goods.

I had no idea Amazon did this until I complained on Mastodon about how terrible its shopping experience is, and Ben replied referencing this practice, nor did I know it has been doing so for at least twelve years. I am certain I have received counterfeit products more than once from Amazon, and I think this is how it happened.

⌥ Permalink

Pixel Envy
Meta’s Steak Sauce Demo Should Have Been Dumber 20 September 2025 at 18:50

Meta’s Steak Sauce Demo Should Have Been Dumber

Pixel Envy

By: Nick Heer

20 September 2025 at 18:50

John Walker, Kotaku

Rather than because of wifi, the reason this happened is because these so-called AIs are just regurgitating information that has been parsed from scanning the internet. It will have been trained on recipes written by professional chefs, home cooks and cookery sites, then combined this information to create something that sounds a lot like a recipe for a Korean sauce. But it, not being an intelligence, doesn’t know what Korean sauce is, nor what recipes are, because it doesn’t know anything. So it can only make noises that sound like the way real humans have described things. Hence it having no way of knowing that ingredients haven’t already been mixed — just the ability to mimic recipe-like noises. The recipes it will have been trained on will say “after you’ve combined the ingredients…” so it does too.

I would love to know how this demo was supposed to go. In an ideal world, is it supposed to walk you through the preparation ingredient-by-ingredient? If Jack Mancuso had picked up the soy sauce, would it have guided the recipe-suggested amount? That would be impressive, if it had worked. The New York Times’ tech reporters got to try the glasses for about thirty minutes and, while they shared no details, said it was “as spotty as Mr. Zuckerberg’s demonstration”.

I think Walker is too hard on the faux off-the-cuff remarks, though they are mock-worthy in the context of the failed demo. But I think the diagnosis of this is entirely correct: what we think of as “A.I.” is kind of overkill for this situation. I can see some utility. For example, I could not find a written recipe that exactly matched the ingredients on Mancuso’s bench, but perhaps Meta’s A.I. software can identify the ingredients, and assume the lemons are substituting for rice vinegar. Sure. After that, what would actually be useful is a straightforward recitation of a specific recipe: measure out a quarter-cup of soy sauce and pour it into a bowl; next, stir in one tablespoon of honey — that kind of thing. This is pretty basic text-to-speech stuff, though it would be cool if it can respond to questions like how much ginger?, and did I already add the honey?, too.

Also, I would want to know which recipe it was following. A.I. has a terrible problem with not crediting its sources of information in general, and it is no different here.

Also — and this probably goes without saying — even if these glasses worked as well as Meta suggests they should, there is no way I would buy a pair. You are to tell me that I should strap a legacy of twenty years of privacy violations and user hostility to my face? Oh, please.

⌥ Permalink

Pixel Envy
U.S. Federal Trade Commission Sues Live Nation and Ticketmaster 19 September 2025 at 23:47

U.S. Federal Trade Commission Sues Live Nation and Ticketmaster

Pixel Envy

By: Nick Heer

19 September 2025 at 23:47

In 2018, the Toronto Star and CBC News jointly published an investigation into Ticketmaster’s sales practices:

Data journalists monitored Ticketmaster’s website for seven months leading up to this weekend’s show at Scotiabank Arena, closely tracking seats and prices to find out exactly how the box-office system works.

Here are the key findings:

Ticketmaster doesn’t list every seat when a sale begins.

Hikes prices mid-sale.

Collects fees twice on tickets scalped on its site.

Dave Seglins, Rachel Houlihan, Laura Clementson, CBC News:

Posing as scalpers and equipped with hidden cameras, the journalists were pitched on Ticketmaster’s professional reseller program.

[…]

TradeDesk allows scalpers to upload large quantities of tickets purchased from Ticketmaster’s site and quickly list them again for resale. With the click of a button, scalpers can hike or drop prices on reams of tickets on Ticketmaster’s site based on their assessment of fan demand.

Ticketmaster, of course, disputed these journalists’ findings. But the very existence of TradeDesk — owned by Ticketmaster — seems to be in direct opposition to Ticketmaster’s obligations to purchasers. One part of the company is ostensibly in the business of making sure legitimate buyers acquire no more than their fair share of tickets to a popular show, while another part facilitates easy reselling at massive scale. The TradeDesk platform is not something accessible by just anyone; you cannot create an account on demand. Someone from Ticketmaster has to set up your TradeDesk account for you.

These stories have now become a key piece of evidence in a lawsuit filed by the U.S. Federal Trade Commission against Live Nation, the owner of Ticketmaster:

The FTC alleges that in public, Ticketmaster maintains that its business model is at odds with brokers that routinely exceed ticket limits. But in private, Ticketmaster acknowledged that its business model and bottom line benefit from brokers preventing ordinary Americans from purchasing tickets to the shows they want to see at the prices artists set.

The complaint’s description (PDF) of the relationship between Ticketmaster and TradeDesk, beginning at paragraph 84 and continuing through paragraph 101, is damning. If true, Ticketmaster must be aware of the scalper economy it is effectively facilitating through TradeDesk.

⌥ Permalink

Pixel Envy
Sponsor: Magic Lasso Adblock: Incredibly Private and Secure Safari Web Browsing 19 September 2025 at 18:00

Sponsor: Magic Lasso Adblock: Incredibly Private and Secure Safari Web Browsing

Pixel Envy

By: Nick Heer

19 September 2025 at 18:00

My thanks to Magic Lasso Adblock for sponsoring Pixel Envy this week.

With over 5,000 five star reviews, Magic Lasso Adblock is simply the best ad blocker for your iPhone, iPad, and Mac.

Magic Lasso Adblock: No ads, no trackers, no annoyances, no worries

Designed from the ground up to protect your privacy, Magic Lasso blocks all intrusive ads, trackers, and annoyances. It stops you from being followed by ads around the web and with App Ad Blocking it stops your app usage being harvested by ad networks.

So, join over 350,000 users and download Magic Lasso Adblock today.

⌥ Permalink

Pixel Envy
Meta’s Whiffed Its Live Demos at Connect 18 September 2025 at 20:12

Meta’s Whiffed Its Live Demos at Connect

Pixel Envy

By: Nick Heer

18 September 2025 at 20:12

Rani Molla, Sherwood News:

While the prerecorded videos of the products in use were slick and highly produced, some of the live demos simply failed.

“Glasses are the ideal form factor for personal superintelligence because they let you stay present in the moment while getting access to all of these AI capabilities to make you smarter, help you communicate better, improve your memory, improve your senses,” CEO Mark Zuckerberg reiterated at the start of the event, but the ensuing bloopers certainly didn’t make it feel that way.

I like that Meta took a chance with live demos but, in addition to the bloopers, Connect felt like another showcase of an inspiration-bereft business. The opening was a more grounded — figuratively and literally — version of the Google Glass skydive from 2012. Then, beginning at about 52 minutes, Zuckerberg introduced the wrist-based control system, saying “every new computing platform has a new way to interact with it”, summarizing a piece of the Macworld 2007 iPhone introduction. It is not that I am offended by Meta cribbing others’ marketing. What I find amusing, more than anything, is Zuckerberg’s clear desire to be thought of as an inventor and futurist, despite having seemingly few original ideas.

⌥ Permalink

Pixel Envy
Reviewing the iPhone 17 Models as Cameras 18 September 2025 at 05:07

Reviewing the iPhone 17 Models as Cameras

Pixel Envy

By: Nick Heer

18 September 2025 at 05:07

If you want reviews of the iPhone 17 — mostly the Pro — from the perspective of photography, two of the best come from Chris Niccolls and Jordan Drake of PetaPixel and Tyler Stalman. Coincidentally, both from right here in Calgary. I am not in the market for an upgrade, but I think these are two of the most comprehensive and interesting reviews I have seen specifically about the photo and video features. Alas, both are video-based reviews, so if that is not your bag, sorry.

Niccolls and Drake walk you through the typical PetaPixel review, just as you want it. The Portrait Mode upgrades they show are obvious to me. Stalman’s test of Action Mode plus the 8× zoom feature is wild. He also took a bunch of spectacular photos at the Olds Rodeo last week. Each of these reviews focuses on something different, with notably divergent opinions on some video features.

⌥ Permalink

Pixel Envy
WSJ Says TikTok Divestiture of U.S. Operations Nears Completion, I Say It Will Make Everyone Mad 17 September 2025 at 03:49

WSJ Says TikTok Divestiture of U.S. Operations Nears Completion, I Say It Will Make Everyone Mad

Pixel Envy

By: Nick Heer

17 September 2025 at 03:49

Raffaele Huang, Lingling Wei, and Alex Leary, Wall Street Journal:

The arrangement, discussed by U.S. and Chinese negotiators in Madrid this week, would create a new U.S. entity to operate the app, with U.S. investors holding a roughly 80% stake and Chinese shareholders owning the rest, the people said.

It must be at least an 80% stake. That is the letter of the law this administration has been failing to enforce.

This new company would also have an American-dominated board with one member designated by the U.S. government.

A golden share, perhaps?

Existing users in the U.S. would be asked to shift to a new app, which TikTok has built and is testing, people familiar with the matter said. […]

“Asked”?

[…] TikTok engineers will re-create a set of content-recommendation algorithms for the app, using technology licensed from TikTok’s parent ByteDance, the people said. U.S. software giant Oracle, a longtime TikTok partner, would handle user data at its facilities in Texas, they said. […]

And I am sure this will satisfy everyone who has found TikTok’s success alarming. Oracle already has access to TikTok’s source code and — at best — will allow TikTok employees to rewrite it to get a “Made in the USA” stamp. It is possible the recommendations system will be unchanged.

Of course, Chinese investors will still have a stake in the U.S. company and, unless the U.S. company is entirely siloed from TikTok everywhere else, users will still be recommended videos the U.S. government framed as a national security threat. But now the U.S. app will seem suspicious to anyone who has been skeptical of the country’s increasing state involvement in the tech industry.

Some TikTok users are going to be furious about this. Some people who viewed its Chinese ownership as inherently problematic are not going to be satisfied by this. It is going to make everyone a little bit upset. It is unclear if it will solve any of the pressing concerns, either. From a distance and in summary, what it looks like is the U.S. government panicked over the only massively successful social media app not based in the U.S., then wrested control of the app and gave it to people friendlier to this government. That is too simplified but, also, not inaccurate.

⌥ Permalink

Pixel Envy
Liquid Glass Pours Out to Apple Devices Today 15 September 2025 at 18:31

Liquid Glass Pours Out to Apple Devices Today

Pixel Envy

By: Nick Heer

15 September 2025 at 18:31

Craig Grannell, Wired:

Apple revealed Liquid Glass as part of its WWDC announcement this June, with all the pomp usually reserved for shiny new gear. The press release promised a “delightful and elegant new software design” that “reflects and refracts its surroundings while dynamically transforming to bring greater focus to content.” Today it launches globally onto compatible Apple devices.

If you haven’t encountered it yet, brace yourself. Inspired by visionOS — the software powering the Apple Vision Pro mixed reality headset — Liquid Glass infuses every Apple platform with a layered glass aesthetic. This is paired with gloopy animations and a fixation on hiding interface components when possible—and showing content through them when it isn’t.

Grannell interviewed several developers for this piece, which is ultimately quite critical of Liquid Glass.

I, too, have thoughts, but life got in the way of completing anything by today’s release. Luckily, there is no shortage of people with opinions about this new material and the broader redesign across Apple’s family of operating systems. I trust you will find their commentary adequate, and I hope you will still be interested in mine whenever I can finish.

In the meantime, I think a chunk of Dan Moren’s iOS 26 review, for Six Colors, is quite good:

Apple has designed extensive rules to try and minimize some of the most distracting impacts of Liquid Glass. For example, if you’re viewing black-on-white content and suddenly scroll past a darker image, the UI widgets will only flip from light to dark mode based on the speed of your scrolling: scroll past it quickly and they won’t flip; it’s only if you slow down or stop with the widgets over the image that they’ll shift into dark mode.

While clever, this also feels remarkably over-engineered to work around the fundamental nature of these devices. It’s a little reminiscent of the old apocryphal story about how the American space industry spent years and millions of dollars designing a pen that could write in space while the Soviets used a pencil. Perhaps they should have used a design that doesn’t require adjusting its look on the fly.

Also, Federico Viticci has published his extraordinary annual review. In addition to the section on design, I am also looking forward to his thoughts in particular on iPadOS 26. Lots to read and lots to discover.

⌥ Permalink

Pixel Envy
Embargoed Reviews for New Apple Stuff Begin With the AirPods Pro 3 15 September 2025 at 18:17

Embargoed Reviews for New Apple Stuff Begin With the AirPods Pro 3

Pixel Envy

By: Nick Heer

15 September 2025 at 18:17

Nicole Nguyen, Wall Street Journal:

My husband, who grew up in Switzerland, helped me test: He spoke French, which turned into English audio in my ears. I responded in English, and he read the French translation on-screen.

There was a delay between his speech and my in-ear translation, which made the conversation stilted. This is par for the course for real-time translators, including the Google Meet and Google Pixel versions I’ve tried. But the AirPods delay was long and it didn’t always transcribe speech correctly, leading to nonsensical translations. (“Down” became “done,” “smoothie” became “movie,” etc.)

Live Translation is still in beta, so I’ll try it again down the line.

Kate Kozuch, Tom’s Guide:

The AirPods Pro 3 are the first AirPods to include a dedicated heart rate sensor.

You can start about 50 different workouts from the iOS 26 fitness app on your iPhone, and your AirPods Pro 3 become the heart rate source, no Apple Watch required. They even sync with Workout Buddy for Apple Intelligence-based workout guidance and Apple Music to launch a workout playlist automatically.

I do not use an Apple Watch, so this feature is compelling for tracking my cycling trips more comprehensively. A similar sensor is in the Beats Powerbeats Pro 2; I wonder if the workout tracking features will work with those, too.

Apple’s AirPods remain, for me, the most difficult product not to buy. I enjoyed my AirPods 2 while they lasted, and using a set of wired headphones afterwards does not feel quite right. But these new models still do not have replaceable batteries. It is hard to write this without sounding preachy, so just assume this is my problem, not yours. I continue to be perplexed by treating perfectly good speaker drivers, microphones, and chips as disposable simply because they are packaged with a known consumable part. The engineering for swappable batteries would be, I assume, diabolical, but I still cannot get to a point where I am okay with spending over three hundred Canadian dollars every few years because of this predictable limitation.

It is difficult to resist, though.

⌥ Permalink

Pixel Envy
Sponsor: Magic Lasso Adblock: Block Ads in iPhone, iPad, and Mac Apps 15 September 2025 at 13:30

Sponsor: Magic Lasso Adblock: Block Ads in iPhone, iPad, and Mac Apps

Pixel Envy

By: Nick Heer

15 September 2025 at 13:30

Do you want to block ads and trackers across all apps on your iPhone, iPad, or Mac — not just in Safari?

Then download Magic Lasso Adblock — the ad blocker designed for you.

Magic Lasso: No ads, No trackers, No annoyances, No worries

The new App Ad Blocking feature in Magic Lasso Adblock v5.0 builds upon our powerful Safari and YouTube ad blocking, extending protection to:

News apps
Social media
Games
Other browsers like Chrome and Firefox

All ad blocking is done directly on your device, using a fast, efficient Swift-based architecture that follows our strict zero data collection policy.

With over 5,000 five star reviews, it’s simply the best ad blocker for your iPhone, iPad, and Mac.

And unlike some other ad blockers, Magic Lasso Adblock respects your privacy, doesn’t accept payment from advertisers, and is 100% supported by its community of users.

So, join over 350,000 users and download Magic Lasso Adblock today.

⌥ Permalink

Chris's Wiki :: blog
Syndication feed fetchers, HTTP redirects, and conditional GET 29 September 2025 at 03:49

Time since last seen	Old rate multiplied by
1x interval	0.37
2x interval	0.13
3x interval	0.05
4x interval	0.02

Total requests	Theoretical interval (rate)
6	Four hours
12	Two hours
24	One hour
32	45 minutes
48	30 minutes
96	15 minutes
144	10 minutes
288	5 minutes
360	4 minutes
480	3 minutes
720	2 minutes
1440	One minute
2880	30 seconds
5760	15 seconds
8640	10 seconds
17280	5 seconds
43200	2 seconds
86400	One second

Normal view

Telephone Carriers

T-Carrier

T-Carrier Infrastructure

DS1 Protocol

The Internet

A faster T1

Miscellaneous

Sidebar: rate limit policies and feed reader behavior