FlashFeed

Full text

It might sound like a betrayal, but PlayStation legend Shuhei Yoshida has a Steam perched next to their TV. The former president even took time to share a bullet-point review covering their thoughts after a few hours of play, and oh boy, does he have mixed feelings about the new PS5 rival, its $1,049 price tag, and PC gaming capabilities.

In an X post appropriately titled "Thoughts after a few hours of playing with Steam Machine," Yoshida rhymes off a pretty long list of testing takes. I'm not sure how I'd feel if I were Valve, sifting through some of these for some industry veteran insight, as he kicks things off with "3D performance is just…meh." Yikes.

If I'm being real, that first bullet sounds like the sort of critique you'd see tied to the Sega Saturn or a cheaper graphics adapter back in the day versus the PS1. That said, seeing as Yoshida's final role at PlayStation was heading up Sony's Indie developer initiative, I imagine he means that smaller, non-AAA outings naturally run better on Valve's mini PC.

Thoughts after a few hours of playing with Steam Machine.- 3D performance is just…meh.- The system recommends to default to 1080p - am I going back to PS4 days?- Some games take a looooooong time to boot, what is it doing?- System UI is easy to use.- Being able to boot up… pic.twitter.com/qQL93AALpZ July 2, 2026

However, Yoshida's hit list does feature both positives and negatives. Yes, exclaiming, "Am I going back to the PS4 days?" regarding the Steam Machine apparently recommending 1080p is firmly still a burn, and the ex-PlayStation head also says "Some games take a looooooong time to boot, what is it doing?" Yet, there's praise in there, as he calls being able to boot up with a controller button press a "killer feature" and describes the system UI as "easy to use."

From where I'm sitting, Yoshida's Steam Machine review captures the typical reactions I'm seeing from PC and console players alike. Yes, it can serve as a solid alternative to the PS5 or Xbox Series X, but it's certainly not a killer, and it has landed on planet earth with a very specific objective.

"It allows me to play Steam games on my living room TV, which is a reason enough to keep it", is a closing line on the list that perfectly explains what the box is all about, and I'm pretty sure Valve would agree. The mini PC can serve as a lower-end desktop rig should you so desire, but breaking through to couch players and offering up access to a vast ecosystem with newfound controller-friendly access is the reason the system now exists, even if that $1,049 price tag is hampering the mission.

(Image credit: Valve) Naturally, the big "but" in Yoshida's Steam Machine assessment is Valve's unfortunate MSRP. "But the price was very unfriendly. Hard to recommend to people unless for research," is the closing line on this mixed bag of punches and praise, and it's exactly how I feel as a hardware reviewer discussing the mini PC. The $1,049 barrier makes the box feel like it's almost in early access, and only those privileged enough to have over a grand for a console rival will be experiencing what could be the future of mainstream gaming.

That said, Valve is already working on improving its newly released box, as the Steam Machine has already received updates that improve its VRAM abilities . I can certainly see a future where prices start to fall, and performance hits consistently harder than a base PS5, but by the time AI datacentres stop gobbling vital components, we might be looking at a sequel.

If you've got Valve's handheld, it's maybe time to grab a Steam Deck dock and use the portable as a low-spec Machine.

Full text

A rescue spacecraft is speeding toward a NASA telescope to prevent it from falling into Earth's atmosphere.

The first-of-its-kind mission launched at 5:09 a.m. EDT (0908 GMT) Thursday (July 2) from the Marshall Islands, carrying a robot-arm spacecraft named Link into the sky aboard a modified Lockheed Martin L-1011 airliner. In midair, a Northrop Grumman Pegasus XL rocket launched Link into orbit, where it will eventually rendezvous with NASA's Neil Gehrels Swift Observatory, a gamma-ray telescope that has been slowly falling toward Earth and headed for an untimely demise.

Link was developed by private company Katalyst Space for $30 million . Katalyst's goal is to meet up with the falling Swift Observatory, which launched in 2004, and pull it to a higher orbit, using Link's robotic arms and thrusters. Swift is still scientifically useful but is rapidly losing altitude due to the natural drag of Earth's atmosphere. Without help, Swift will likely meet its demise later this year, scientists say.

"This is a high-risk, high-reward mission," Shawn Domagal-Goldman , director of NASA's Astrophysics Division, said in a statement . "We have much to gain by attempting this boost, which is more affordable than trying to replace Swift's capabilities, and allows NASA to advance the nation's satellite servicing industry, for the benefit of all."

Swift cost $250 million in 2004, which is roughly $450 million today when accounting for inflation — a relatively cheap observatory compared with the $10 billion James Webb Space Telescope . The Swift mission was originally designed to study gamma-ray bursts ‪—‬ cosmic explosions that occur when a massive, dying star collapses into a black hole. Swift has made many observations of these bursts in the past 20-plus years while also watching for other signs of stellar activity, like X-ray flares or supernova explosions, as well as transient objects, like comets and asteroids .

In sum, Swift helps us see how the universe is changing over short periods.

"Swift is NASA's multitool when it comes to studying the cosmos," S. Bradley Cenko , Swift's principal investigator and an astrophysicist at NASA's Goddard Space Flight Center, said in the statement. "It observes the sky using a wide range of light, and rapidly points at short-lived outbursts, alerting other facilities in space and on the ground to help coordinate follow-up observations."

Katalyst engineers attach Link to a baseplate inside the Space Environment Simulator at NASA Goddard on April 28, 2026. The team practiced firing the satellite’s ion thrusters and operated one of the robotic arms while they cycled through space-like temperatures. (Image credit: NASA/Sophia Roberts) High-risk rescue mission

The rescue mission comes with several challenges. Swift was not designed to be serviced when it launched, making the logistics of the mission something of a puzzle. Moreover, NASA only gave the contract to Katalyst in September, after high solar activity in recent years ballooned Earth's atmosphere and unexpectedly accelerated the spacecraft's fall due to increased drag. This meant the mission had to be ready for space in less than a year after the usual punishing requirements of design, build and testing.

Katalyst accepted the challenge because the company hopes to do this kind of work more often. "If we're going to build an enduring presence beyond Earth, we need the capability to manipulate our environment in space," Katalyst CEO Ghonhee Lee said in the statement. "That means deploying robotic spacecraft that can reposition, repair, refuel, and refit satellites after launch."

Unless Swift altered its operations plan, however, the spacecraft would have been irrecoverable in July. So, to give Link as much time as possible to rescue Swift, the operations team at Penn State's Eberly College of Science made some changes.

For example, Swift's science was minimized so that the spacecraft looks at targets only if the telescope is put in a "streamlined position" to minimize drag, NASA explained. Power consumption has also been slashed to let Swift's solar panels fly "in a more aerodynamic orientation," which also reduces the drag that is making the spacecraft fall into the atmosphere.

Related stories

Can we refuel 'dead' satellites in space? Bold new missions aim to try.

Falling metal space junk is changing Earth's upper atmosphere in ways we don't fully understand

Paperclip-sized spacecraft could visit a nearby black hole in the next century, study claims

These changes will allow Swift to stay above the minimum rescue altitude of 185 miles (298 kilometers) until autumn, agency models suggest. Katalyst will use that time to do the usual spacecraft "commissioning" on Link to make sure all systems are well. It will take Link about a month to rendezvous with Swift, according to NASA.

Following those steps, Link will approach Swift for evaluation, and Katalyst will use Link's robotic arms to latch onto the NASA observatory. Link will then use its propulsion system to gradually bring Swift's orbit to about 370 miles (595 km) ‪—‬ well above the pathway of the International Space Station, which orbits roughly 250 miles (400 km) over Earth.

NASA did not say how much longerSwift would be able to keep observing, should it reach its new altitude safely. However, figures from the European Space Agency suggest that spacecraft at an altitude of 310 miles (500 km) reenter the atmosphere within about 25 years. This would suggest that as long as Swift's instruments hold out, scientists would have many years of observations to look forward to.

Full text

Frederick Douglass’ great lecture “What to the Slave Is the Fourth of July?,” delivered on July 5th, 1852, to a predominantly white audience at Rochester’s Corinthian Hall, captured the contradictions built into the nation’s July Fourth celebrations, then and to this day. The nation’s founding document may have declared that “All men are created equal,” but at the time of Douglass’ speech over three million Black people were enslaved. Douglass’ decision to deliver the speech on July 5th conveyed his view that the Fourth was not yet worth celebrating, for the holiday revealed to the American slave, “more than all other days in the year,” as Douglass put it, “the gross injustice and cruelty to which he is the constant victim.” The Thirteenth Amendment ended slavery in 1865, but “What to the Slave” continues to speak to many readers about the gross injustice and cruelty of a nation that fails to offer equal opportunities and justice for all.

That speech, however, may not be the most compelling or even relevant of Douglass’ speeches for this year’s Fourth of July. A better candidate would be his 1867 “Sources of Danger to the Republic,” delivered after the Civil War on several occasions, including as part of a Black-sponsored lecture series in Philadelphia. Douglass took stock of the Founding Fathers, the Constitution, and the democratic promise of the Declaration of Independence, and his concerns prophetically anticipate our present moment.

“Sources of Danger” was prompted by Douglass’ anger at President Andrew Johnson, whom he believed had betrayed the promise of Reconstruction by attempting to block all legislation intended to bring about racial equality. Though not delivered on or near July Fourth, the speech shares much of the holiday’s sentiment. Douglass begins with an appeal to “patriotic citizens of the United States,” and, stopping short of “Fourth-of-July extravagances,” he offers his “humble gratitude to the fathers who framed the nation’s founding documents.” (I quote from the version Douglass gave to the Black audience in Philadelphia on January 3, 1867.)

As in his more famous Fifth of July speech, Douglass laments the nation’s failure to live up to “the great principles of the Declaration of Independence.” In addition, he argues in “Sources of Danger” that the Constitution makes the presidency into something like a monarchy and thus a threat to democracy. Douglass concedes that there are electoral limits on the presidency, making it different from a monarchy. Nevertheless, the president, invested as he is with nearly unlimited power, “can reign long enough to commit any number of mischievous acts, and so defeat the most beneficent measures of our Government.”

How did it happen that some of the signers and supporters of the Declaration of Independence could create a government in which a president like Johnson is “invested with kingly powers”? Douglass speculates that the “fathers of this republic,” who were born under “monarchical institutions,” retained an unconscious nostalgia for monarchy. That nostalgia led to a key mistake in the Constitution—the elevation of the presidency to a kingship—which “must now be corrected.” As he instructs his audience, “You must have either a purely republican Government or you must have a monarchical Government, one or the other.” The burden of what could be called Douglass’ No-Kings speech is to identify the problem and suggest how to fix it.

Initially, the problem seems to be Andrew Johnson. Douglass states that the framers of the Constitution failed to anticipate a president like Johnson—a man willing to take full advantage of the lack of restraints on the presidency to become a “one-man power.” But even though Douglass refers to Johnson now and again, over the course of the speech he steps back in order to make a larger case about Constitutional failure and more disinterestedly offers his concerns about the nation’s future presidents and prospects.

“King Andy,” illustration by Thomas Nast. From Harper’s Weekly , November 1, 1866.

Douglass first calls attention to how the Constitution puts “ immense patronage in the hands of the President.” The president, he observes, has access to “hundred millions of dollars per annum in times of peace, and uncounted thousands of millions in times of war.” There are only limited constraints on how he might use the money. He can use it to advance his own policies or even to enrich himself and assume power over others. Douglass is particularly concerned about how easily the president can distribute the nation’s money “among his political friends.” At the thought, Douglass can only throw up his hands and declare: “What a power! What a corrupting power!”

And there’s more: Douglass calls attention to the corruption enabled by the president’s veto power. He is outraged that a single person—the president—can overrule a simple majority of congressional legislators. The idea that the president always out-votes an under two-thirds congressional majority is “an absurdity,” Douglass says, as well as an affront to democracy. Douglass leaves no doubt about what he thinks should be done to address the anti-democratic presidential veto: “I want that old, despotic, and aristocratic power of our Government utterly banished from our Constitution. It has no business in a republican form of government.” Even in England, Douglass points out, no such power is invested in the monarchy.

Even though Douglass refers to Johnson now and again, over the course of the speech he steps back in order to make a larger case about Constitutional failure and more disinterestedly offers his concerns about the nation’s future presidents and prospects.

Douglass next turns to the president’s pardoning power as yet another source of danger. He does not want to eliminate the possibility of pardons, but he sees corruption as the inevitable result of the president’s absolute right to grant them. In Douglass’ time, he was particularly angry about Johnson’s decision to pardon nearly all of the leaders of the former Confederacy. As a political theorist, however, he looked into the future and was especially concerned about how unrestricted pardons added to presidential power. With the help of the Constitution, a president’s ability to pardon felons could be deployed “to win personal friends and co-operation and alliance, instead of loyal obedience to the laws of the land.”

Among Douglass’ suggestions for limiting presidential power, perhaps the most pertinent for our present historical moment is his call to abolish the president’s ability to undertake “secret diplomacy.” which puts the nation at risk. As Douglass warns, a bad president will choose a bad cabinet, and with “a bad President and a bad Cabinet” doing the secretive diplomatic work, “this nation might be conducted into the jaws of a terrible war, and be perfectly helpless.”

Andrew Johnson, the immediate focus of Douglass’ critique, was impeached that next year. Douglass thought he should have been convicted for crimes against Black people, but he was impeached for something very different: his violation of the Tenure of Office Act by firing Secretary of War Edwin M. Stanton. But was impeachment the best response to the problem? In “Sources of Danger” Douglass insists that broad constitutional change, rather than rituals of impeachment, would help to insure the future of the country.

Even in an angry and at times despairing speech like “Sources of Danger to the Republic,” Douglass retained hope for the future, calling on Americans to secure the democratic promise of the nation. “Strike down the one-man power everywhere,” he says at the speech’s conclusion; “make your Government lean to the people, and away from the individual or the one-man power.” If Americans are willing to do that, through constitutional reform or other means, “you make sure the permanence, prosperity, and glory of this great republic.” This is Douglass’ wish and request to us from 1867. As in “What to the Slave Is the Fourth of July?,” he reminds Americans of the still unfulfilled egalitarian promise of the Declaration of Independence, and there is no better time for such a reminder than July Fourth.

Full text

Z.ai , the Beijing-based artificial intelligence lab formerly known as Zhipu AI, on Wednesday officially launched ZCode , a free desktop application it describes as an "Agentic Development Environment" purpose-built for its flagship GLM-5.2 large language model. The move marks the company's most aggressive push yet into the fast-growing AI-powered coding tool market, where it now competes directly with Cursor , Claude Code , GitHub Copilot , and Google's Antigravity .

"Introducing ZCode, the official development environment for GLM-5.2," the company wrote on X, noting the tool is available on macOS, Windows, and Linux, supports bring-your-own-key (BYOK) configurations for third-party models, and offers a 1.5x usage-quota bonus for subscribers to its GLM Coding Plan.

Read one way, ZCode is simply another entrant in a crowded market. Read another, it is a single product that crystallizes three of the most consequential trends in enterprise software today: the race-to-the-bottom pricing of frontier AI models, the geopolitical balkanization of the AI stack, and the rapid maturation of agentic coding agents into what Gartner now estimates is a roughly $10 billion market .

An AI coding tool designed to think in projects, not prompts

Unlike traditional IDEs that bolt on AI through a chat sidebar or autocomplete extension, ZCode is best understood as an agent-first development environment. Its core design is built around long-horizon tasks: the user describes an outcome, the agent plans the work, edits files, runs checks, reviews progress, and continues across multiple iterations until the goal is met.

ZCode organizes the development experience around the ZCode Agent , deeply tuned for GLM-5.2 , with emphasis on deep integration: the model, tools, and execution workflow are tuned together so the Agent fits continuous, multi-step real-world development tasks. The environment supports continuous follow-up across devices: desktop, mobile Remote, and Feishu / WeChat Bot can all keep the same workspace task moving. Sensitive commands, file changes, and high-permission actions go through confirmation before execution.

That remote-control feature — the ability to steer a running coding agent from WeChat , Feishu , or Telegram on a phone — is a differentiator that speaks directly to the Chinese developer market, where those messaging platforms dominate professional communication. You can keep checking progress and adding instructions while long-running work continues, from any device with these messaging apps.

The tool is free to download. Revenue flows through Z.ai's GLM Coding Plan subscription tiers , which start at $16.20 per month for a "Lite" plan and scale to $144 per month for "Max" — prices that undercut Anthropic's Claude Code and Cursor's comparable tiers by significant margins.

Through July 31, ZCode is offering a promotional 1.5x effective quota bonus for Coding Plan subscribers, with off-peak token consumption charged at a 0.67x coefficient. The platform also supports multiple AI models and agents, including Claude Code, Codex, Gemini, and OpenCode — a pragmatic concession to the reality that no single model wins every task.

GLM-5.2, the open-source model trained entirely on Chinese chips, powers the whole experience

ZCode's value proposition is inseparable from GLM-5.2 , the model it was designed to showcase. Z.ai released GLM-5.2 on June 16, first to its Coding Plan subscribers and subsequently as open-source weights under the MIT license on Hugging Face — a sequencing decision that prioritized distribution over the traditional benchmark-led launch.

The model's specifications are formidable. GLM-5.2 is a 744-billion-parameter mixture-of-experts architecture with 40 billion active parameters, a genuine one-million-token context window — five times the 200K limit on its predecessor — and training on 28.5 trillion tokens. It ranked second globally on Code Arena as of mid-June, trailing only Anthropic's Claude Fable 5, making it one of the highest-performing publicly available models for coding tasks.

Critically, the model was built entirely without American chips. As Decrypt reported, GLM-5.2 " runs entirely on Huawei silicon ." Stability AI founder Emad Mostaque estimated total training costs at roughly $25 million, with 80 percent spent on post-training — a figure that, if accurate, would make GLM-5.2 extraordinarily cheap relative to Western frontier models.

On benchmarks, GLM-5.2 performs within striking distance of the best proprietary systems. It trails Anthropic's Claude Opus 4.8 by just one percentage point on FrontierSWE , a benchmark measuring multi-hour autonomous engineering projects, while edging out OpenAI's GPT-5.5 .

Its API pricing — $1.40 per million input tokens and $4.40 per million output — are a cost reduction of up to 82 percent compared to Anthropic's Claude Opus 4.8 at $5 and $25, respectively. Because ZCode is a first-party tool from the same company that makes the model, it requires no manual endpoint configuration — the model is wired in.

The Anthropic export ban gave Chinese AI its biggest opening yet

ZCode's arrival cannot be separated from the geopolitical drama that has roiled the AI industry over the past three weeks. On June 12, the U.S. government, citing national security authorities , issued an export control directive suspending all access to Anthropic's Fable 5 and Mythos 5 models by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees. Enterprise clients in finance, healthcare, SaaS, and critical infrastructure found their core intelligence services abruptly disabled, without exception, prior warning, or effective recourse.

While the Trump administration lifted those controls just yesterday — Anthropic confirmed on June 30 that the Department of Commerce had rescinded the directive — the episode sent shockwaves through the developer community and accelerated interest in open-source, self-hostable alternatives. The government's crackdown on Anthropic coincided with a swift rise in Chinese open-source models that are proving to be almost as capable and significantly cheaper than some of the most powerful U.S. models.

Z.ai's timing was surgical. On the same day the Trump administration ordered Anthropic's most advanced models blocked for foreign nationals, Zhipu announced the open-source release of GLM-5.2 with no usage restrictions. The South China Morning Post reported that GLM-5.2 would be available to all users of Zhipu's new GLM Coding Plan subscription, "priced at just a tenth of Anthropic's premium Claude Code and Claude Max tiers."

The market responded accordingly. Zhipu AI's market capitalization crossed HK$1 trillion ( US$128 billion ) on June 22, driven by a 42 percent intraday share surge. JPMorgan raised its 2026–2030 revenue forecast for Zhipu by between 7 and 16 percent following the launch, projecting an over 534 percent revenue surge for 2026 and expecting the AI firm to turn a profit by 2028.

Why vendor lock-in now carries a geopolitical risk that no SLA can cover

The Fable 5 episode did more than embarrass Anthropic. It introduced a new risk category into enterprise AI procurement: sovereign access risk. When a government can disable a commercially deployed AI model overnight, the traditional evaluation criteria of developer experience, benchmark scores, and pricing become secondary to a more fundamental question: Will this tool still work tomorrow?

The event exposed the inadequacy of standard enterprise contract language. An investigation by FifthRow found that almost all standard Data Processing Addenda, SaaS agreements, and procurement SLAs "relied on vague 'force majeure' or 'compliance with law' catch-alls, not on precise, actionable regulatory suspension or kill-switch clauses."

ZCode's BYOK architecture and GLM-5.2 's MIT-licensed open weights offer a partial answer. A development team can download the model, host it on its own infrastructure, and run ZCode against it without ever touching Z.ai's cloud — eliminating both American export-control risk and Chinese data-sovereignty concerns in a single move. The catch is that anyone using Z.ai's cloud API remains subject to Chinese law, a consideration that evaporates only with pure self-hosting.

Gartner analysts have warned that governance, pricing, support, workflows, commercial maturity, and market durability matter as much as developer experience and model capabilities when evaluating coding agent vendors for enterprise-wide adoption. By that measure, ZCode faces a steep climb. It is not open source itself; Linux support remains in beta; and security reviewers have flagged the need for careful evaluation of its credential handling, particularly for remote development over SSH and messaging-platform-triggered tasks — an agent that can be summoned from WeChat involves access paths that should be mapped before trusting it with anything sensitive.

Inside the $10 billion race where model labs are becoming full-stack IDE companies

ZCode enters one of the most crowded and fastest-moving markets in enterprise software. Enterprise AI coding agents are capturing a growing share of enterprise software engineering spend, with the market estimated at roughly $9.8 billion to $11.0 billion annualized as of April 2026, according to Gartner . A defining shift this year, the analyst firm noted, is "the movement of frontier model providers into direct competition with application-layer vendors" — precisely the pattern ZCode embodies.

Gartner codified this evolution in May when it renamed its annual Magic Quadrant from "AI Code Assistants" to "Enterprise AI Coding Agents," defining the category as "autonomous or semiautonomous software engineering solutions that perceive context, translate human intent into multistep plans, and execute and verify those steps across code, tests and related engineering artifacts." The 2026 Magic Quadrant names Anthropic, Cursor, GitHub, and OpenAI as Leaders. Z.ai was not among the 12 vendors evaluated — an absence that underscores both the company's nascent enterprise sales presence outside China and the Western-centric lens through which the analyst community still views the market.

The competitive landscape is daunting. Cursor is the $2 billion ARR IDE that feels like VS Code with a supercharger. Claude Code reached approximately $2.5 billion in annualized revenue by early 2026. Google relaunched Antigravity 2.0 at I/O in May, and Cognition retired the Windsurf brand, relaunching the IDE as Devin Desktop with the Agent Command Center as the default surface.

Against these entrenched players, ZCode's pitch rests on three pillars: deep first-party integration with GLM-5.2 that no third-party editor can replicate, aggressive pricing that starts at a fraction of Western competitors, and MIT-licensed open weights that allow enterprises to self-host — eliminating the regulatory kill-switch risk that the Fable ban made viscerally real.

Z.ai's real challenge is turning a $128 billion valuation into a global developer tools business

Z.ai controls the model ( GLM-5.2 ), the subscription layer ( the GLM Coding Plan ), and the IDE ( ZCode ) — a tightly coupled stack that optimizes for performance but concentrates switching costs. For the company, the business logic is clear. Its most reliable revenue stream has been on-premises deployments for Chinese government agencies, state-owned banks, and energy conglomerates. In full-year 2025, on-premises deployment revenue reached RMB 534 million, growing over 100 percent year-over-year and accounting for 73.7 percent of total revenue with a gross margin of 48.8 percent. ZCode and the GLM Coding Plan represent the company's bid to build a comparable revenue engine in cloud-based developer tools — globally, not just in China.

The early signals are encouraging for Z.ai , if anecdotal. Community reception on X was enthusiastic, with one early user calling the tool "super stable" and others clamoring for more Coding Plan capacity. "Bro, can't snag your family's Coding Plan? When are you gonna stock up on more cards?" one user wrote in Chinese , suggesting demand is already outstripping supply.

But the hard questions loom large. Can a Chinese AI company build trust with Western enterprise buyers amid escalating technology tensions? Can ZCode's ecosystem mature fast enough to compete with Cursor's polished UX, Claude Code's deep agent primitives, and GitHub Copilot's unmatched distribution? And can Z.ai sustain a company valued at $128 billion while still losing money?

What is no longer in question is the competitive dynamic itself. Three weeks ago, a U.S. government directive proved that access to the world's best coding model can vanish overnight. Today, a Chinese lab is shipping a free IDE, an open-source model trained on zero American chips, and a subscription plan that costs less per month than a single lunch in Manhattan. The AI coding agent market did not just become global this summer. It became a market where the fallback option might be better than the thing it's falling back from — and that changes the calculus for every engineering leader choosing a toolchain in the second half of 2026.