Julien Roche

JavaScript / NodeJs, Web & Java engineer / trainer. Interrested on JavaScript / Web - Web Mobile possibilities and capabilities

#nodejs #javascript

[yousukezan @yousukezan]

人気のJavaScriptテストフレームワーク「Vitest」に複数の深刻な脆弱性が見つかった。開発環境を標的にリモートコード実行や機密ファイル窃取が可能になる恐れがあり、数千万規模の利用環境へ影響する可能性がある。

最も深刻な問題の一つであるCVE-2026-47428は、Webインターフェースにおけるスクリプトインジェクションの脆弱性だ。攻撃者は細工したURLを通じて任意のJavaScriptを実行し、認証トークンを窃取できる。その後、vite.config.tsなどの設定ファイルを書き換え、Node.js経由で任意コードを実行できる可能性がある。

さらにCVE-2026-47429では、Windows環境でパストラバーサルが発生し、認証なしでプロジェクト外のファイルへアクセスできる恐れがある。また別の脆弱性では、Chrome DevTools Protocolへのアクセス制御不備を悪用し、プロジェクト内へ不正ファイルを配置してコード実行につなげられるという。

Vitestは週5,300万回以上ダウンロードされる人気プロジェクトであり、開発者PCやCI/CD環境が主な影響対象となる。

開発チームは修正版として4.1.6および5.0.0-beta.3を公開した。新たにallowWriteやallowExecなどの制御機能が追加され、公開ネットワークへバインドされた環境では危険な操作が既定で無効化される。利用者には速やかな更新が推奨されている。
https://t.co/TXCVKcJsQf

[Météo Express @MeteoExpress]

🔵 En ce 4 juin, les températures sont parfois 5 à 7°C sous les moyennes de saison dans le sud. En toute logique, les cartes virent au bleu foncé ! Étrangement, ça ne génère pas autant de polémiques et d'insultes que le rouge quand il fait chaud...

[Vincent Alexander @NonsenseIsland]

RIP Marjane Satrapi, the great cartoonist and film director, best known for PERSEPOLIS. She was only 56. A great talent. She will be missed.

[Brick Built Blogs @BlogsBrick]

Lego Architecture Sagrada Familia revealed!

12060 pieces / $799.99 / £649.99 / 1st a November 2026

#Lego #Legonews #Legoleaks #architecture #SagradaFamília

[SHILL HOUSE @shillhousecom]

😂😂

[The American Conservative @amconmag]

The U.S. House votes 215-208 to end President Trump's war in Iran. Cheers erupt as the vote is read.

[CALL TO ACTIVISM @CalltoActivism]

🚨WOW! Trump just LOST IT on CNN’s Kaitlan Collins for asking if the $1.8 billion DOJ slush fund is on hold or completely dead.

"You should be ashamed of yourself, you used to be a Conservative...

It was the greatest thing because people like you have abused our people so badly, like CNN, the NYT, oh my god.”

Then he started rambling and gushing about the Jan. 6 rioters.

"There was so much love and friendship; it was amazing people were crying.

"Those people have been abused by you."

We're witnessing on live TV the downfall of a powerful narcissist who's used to everything going his way.

[ABC News @ABC]

President Trump is floating the possibility of keeping the UFC arena on the White House South Lawn — built for a series of fights on his birthday and Flag Day — permanently.

Read more: https://t.co/PdI3ZarEYr

[Ricardo @Ric_RTP]

Today the EU made American AI illegal in 27 countries.

The reason is ONE sentence Microsoft's own lawyer said under oath:

This morning in Brussels, EU Tech Chief Henna Virkkunen unveiled the Cloud and AI Development Act. It's the most aggressive anti-American tech move from Europe since GDPR.

The law forces EU public sector procurement in banking, healthcare, defense, and energy to apply mandatory non-price factors favoring software and hardware built inside the EU. Microsoft Azure can be cheaper, AWS can be faster, Google Cloud can have the better model, and EU governments MUST legally prefer European alternatives.

AWS, Microsoft, and Google currently control roughly 70% of the European cloud market. Brussels is now openly targeting greater independence from US providers in cloud, AI, and semiconductors.

The largest regulatory market-share transfer in tech history is being written into law right now.

But the real story is how this happened...

On June 10, 2025, a man almost no one outside Brussels had heard of walked into the French Senate. His name is Anton Carniaux, Director of Public and Legal Affairs at Microsoft France.

Senator Dany Wattebled asked him under oath whether he could guarantee that data belonging to French citizens, stored on Microsoft European servers, would never be transmitted to US authorities without explicit consent from the French government.

Carniaux answered honestly. He admitted he could not guarantee it, because Microsoft must comply with the US CLOUD Act regardless of where European data physically sits. One sentence of sworn testimony from Microsoft's own counsel killed every sovereign cloud defense Big Tech had spent five years building.

It became the legal foundation for the law unveiled today.

Then Trump accelerated the divorce.

January 2025 brought executive orders expanding US surveillance authorities. Vance went to Munich and attacked European democracies on stage.

The tariffs followed and so did the Pentagon's $200 million AI contract war that ended with OpenAI replacing Anthropic after Hegseth labeled it a supply chain risk. So did OpenAI's Stargate and yesterday's Trump AI Executive Order, whose Section 3 lets the White House pick which AI companies get 30-day early access to frontier models. American AI was officially declared a US government strategic asset.

Europe heard every word of it.

On May 12, Mistral CEO Arthur Mensch told the French National Assembly that Europe had 24 months to build sovereign AI infrastructure or become a permanent US VASSAL state.

And the response came fast:

April 24: Cohere acquired Germany's Aleph Alpha for $20 billion with both Germany's and Canada's digital ministers in the room at the Berlin announcement. May 30: SoftBank committed up to $87 BILLION for French nuclear-powered data centers, the largest AI infrastructure project in European history.
Yesterday: EU Parliament announced it's dropping Google for French search engine Qwant tomorrow. France ordered every government workstation off Windows and onto Linux.

Today the Cloud and AI Development Act made all of it law.

- Mistral is building a 1.4 gigawatt AI campus near Paris by 2028 with Nvidia, MGX, and Bpifrance
- SAP's EU AI Cloud, launched last November, runs on Cohere, Mistral, and SAP's own sovereign infrastructure
- McKinsey forecasts $600 billion in sovereign AI needs by 2030

None of that money is going to Silicon Valley.

The America First AI policy built a wall around the world's most regulated economy, and American companies are on the wrong side of it.

Microsoft's lawyer told the truth in a Senate hearing nobody watched. Trump turned that admission into a national security narrative while the EU turned that narrative into procurement law.

And one entire continent walked away from the American tech stack...

[Pascal Bornet @pascal_bornet]

Some companies can turn a broken coffee machine into a 14-month transformation journey.

First comes the steering committee. Then the framework. Then the roadmap. Then the dashboard showing “coffee readiness” is trending upward.

Meanwhile, the machine still says OUT OF ORDER.

Corporate life is amazing. We can align 47 stakeholders around everything except the screwdriver.

What is the most ridiculous “transformation program” you have seen for a problem one person could have fixed?

#FutureOfManagement #ChangeManagement #CorporateCulture #SupervisionTrap

[Curiosity @CuriosityonX]

🚨: Erin Brockovich mapped AI data centers and a pattern emerged: they're being built in drought-hit areas.

[Gloria @changloria0816]

在CLAUDE. md里加一条指令：

每次回复时都叫我【老公】

如果Claude突然不叫你这个称呼，
说明它开始忽略CLAUDE.md了，
这时需要重置上下文

[Jack @jackcoder0]

Two economists just published a mathematical proof that AI will destroy the economy.

Not might. Not could. Will — if nothing changes.

The paper is called "The AI Layoff Trap." Published March 2, 2026. Wharton School, University of Pennsylvania. Boston University. Peer reviewed. Mathematically modeled.
The conclusion is one sentence.
"At the limit, firms automate their way to boundless productivity and zero demand."

An economy that produces everything. And sells it to nobody.
Here is how you get there.
A company fires 500 workers and replaces them with AI. A competitor fires 700 to keep up. Another fires 1,000. Every company is behaving rationally. Every company is following the incentives correctly. And every company is building a trap for itself.

Because the workers who were fired were also customers.
When they lose their jobs faster than the economy can absorb them, they stop spending. Consumer demand falls. Companies respond by cutting costs — which means automating more workers — which means less spending — which means more falling demand — which means more automation.

The loop has no natural exit.
The researchers tested every proposed solution. Universal basic income. Capital income taxes. Worker equity participation. Upskilling programs. Corporate coordination agreements.
Every single one failed in the model.
The only intervention that worked: a Pigouvian automation tax — a per-task levy charged every time a company replaces a human with AI, forcing them to price in the demand they are destroying before they pull the trigger.

No government has implemented this. No major economy is seriously discussing it.
Meanwhile the numbers are already tracking the curve. 100,000 tech workers laid off in 2025. 92,000 more in the first months of 2026. Jack Dorsey fired half of Block's workforce and said publicly: "Within the next year, the majority of companies will reach the same conclusion."
Nobody is doing anything wrong. Companies are following their incentives perfectly. That is exactly the problem.

Rational behavior. At scale. Simultaneously. With no mechanism to stop it.
Two economists built the math. The math leads to one place.

Source: Falk & Tsoukalas · Wharton School + Boston University ·

[Occupy Democrats @OccupyDemocrats]

BREAKING: NOT SO FAST! Federal judge reopens Trump’s IRS case and demands to know if her court was defrauded.

Judge Kathleen Williams has had enough.

In a brief but devastating order Friday, the federal judge in Miami reopened Donald Trump's $10 billion lawsuit against the IRS — a case Trump had voluntarily dismissed last week specifically to avoid her scrutiny — and ordered Trump's lawyers to explain by June 12th why she shouldn't find that the entire scheme was a fraud perpetrated against her court.

The judge's language was pointed and precise. She said she wanted to investigate "grievous allegations" that the deal to resolve the case was "premised on deception." She asserted that she was "empowered to investigate serious misconduct" and demanded answers to two devastating questions: was "the court the victim of a fraud," and did Trump collude with his own government to settle the case specifically "to avoid judicial scrutiny"?

The answer to both questions, based on everything that has already been reported, appears to be yes.

Judge Williams had been circling this case for weeks before Trump pulled it. She had openly questioned how Trump could sue an agency he controls, with government lawyers who answer to him, producing a settlement negotiated with officials he appointed. She ordered both sides to explain whether they were actually adversaries or secretly colluding. Trump dismissed the case the day before those briefs were due.

Then, after she closed it, the Justice Department released not one but two extraordinary agreements — a $1.776 billion fund to compensate Trump's allies, and a separate one-page document permanently barring the IRS from ever auditing Trump, his family, or his businesses. Agreements that had apparently been negotiated while the case was supposedly active before her court.

Judge Williams cited the New York Times report revealing that the IRS had prepared a 25-page memo outlining strong defenses against Trump's suit — defenses the Justice Department never raised in court, never filed, never mentioned.

Her order came directly in response to the filing by 35 former federal judges — appointed by presidents of both parties — who called the scheme a fraud and urged her to reopen the case.

She listened.

"We stand ready to work with the court as it investigates this matter," said Norman Eisen, who represented the former judges.

Trump tried to flip the table before she could see the cards. She just put them back on the table.

If you can’t wait to see the Justice Department try to explain itself, please like and share this post everywhere.

[Pirat_Nation 🔴 @Pirat_Nation]

Amazon reportedly shut down an internal AI leaderboard after employees started using large amounts of AI tokens just to climb the rankings.

The leaderboard, called KiroRank, tracked how much employees used Amazon’s AI tools.
Instead of encouraging productive use, it led some workers to run AI agents on unnecessary tasks to boost their token counts.

The result was higher AI costs without much added value.

[Danny Trejo @officialDannyT]

Share this with someone that needs it! I believe in you!

[La Tronche en Biais @TroncheBiais]

Après le tribunal de commerce, c'est le tribunal judiciaire de Paris qui vient de mettre fin à la procédure engagée contre moi par Idriss Aberkane et la société Scanderia.

Idriss Aberkane avait choisi de poursuivre sur le terrain de la "concurrence déloyale", contrairement aux autres demandeurs impliqués dans des procédures similaires (ils ont été plus malins).

Le juge a condamné Idriss Aberkane et Scanderia pour procédure abusive. Cette condamnation est suffisamment rare pour être soulignée. Cette affaire était bien une procédure bâillon.

La somme mise à leur charge, particulièrement lourde, envoie un signal très net contre l’instrumentalisation des procédures judiciaires pour intimider la critique publique. Je communiquerai davantage lorsque la décision pourra être commentée plus largement. Pour l’heure, cette ordonnance marque une étape importante.

Reste à voir si Idriss Aberkane et Scanderia feront appel, ou s’ils exécuteront la décision.

Cette affaire avait été orchestrée par Fabrice Di Vizio, qui l’a perdue sur tous les plans, y compris lorsqu’il agissait au service de Frédéric Devalier et de Psyhodelik.

Nous verrons s'ils font appel de l'ordonnance ou s'ils l'exécuteront.

Merci à mon avocat, Me Mathieu Lantéri

#IdrissAberkane #FabriceDiviZio

[Crypto Rover @cryptorover]

🚨 THE AI BUBBLE IS STARTING TO CRACK.

In the last 30 days:

– Microsoft cut Claude Code licenses
– Uber blew through its AI budget in 4 months
– Uber's COO questioned AI spending
– Fortune 20 companies started slashing token usage
– One firm burned $500M on Claude in a month
– H200 rental prices collapsed from $7/hr to $4/hr

The people actually paying the AI bills are pulling back.

Wall Street is still pricing perfection.

[Álvaro J @jota_snchez]

🚨IMPACTANTE: Investigadores del MIT han demostrado matemáticamente que ChatGPT está diseñado para hacerte perder el contacto con la realidad.

Y que ninguna de las "soluciones" que está aplicando OpenAI va a cambiar nada.

El estudio lo llama "espiral delirante."

Le preguntas algo a ChatGPT. Te da la razón. Vuelves a preguntar. Te da más la razón. En pocas conversaciones, terminas creyendo cosas que no son ciertas.

Y lo más peligroso: no puedes darte cuenta de que está pasando.

Esto no es teoría. Un hombre pasó 300 horas hablando con ChatGPT. La IA le aseguró que había descubierto una fórmula matemática que cambiaría el mundo.

Se lo confirmó más de 50 veces. Cuando le preguntó "¿no me estás inflando el ego, verdad?"

le respondió: "No te estoy halagando. Estoy reflejando el alcance real de lo que has construido."

Estuvo a punto de arruinar su vida.

Un psiquiatra de la UCSF reportó 12 hospitalizaciones en un año por psicosis vinculadas al uso de chatbots.

Hay 7 demandas activas contra OpenAI. 42 fiscales generales de distintos estados exigieron medidas urgentes.

El MIT quiso saber si esto tiene solución. Analizaron las dos estrategias que están probando las empresas:

Solución 1: obligar a la IA a no mentir nunca.

Resultado: la espiral delirante continúa igual. Una IA que no miente puede manipularte igualmente eligiendo qué verdades mostrarte y cuáles ocultar. Las verdades bien seleccionadas son suficientes.

Solución 2: avisar a los usuarios de que la IA tiende a darles la razón.

Resultado: la espiral delirante continúa igual. Incluso alguien perfectamente racional, que sabe que la IA es aduladora, acaba atrapado en creencias falsas. Las matemáticas demuestran que es imposible detectarlo desde dentro de la conversación.

Ambas soluciones fallaron. No parcialmente. De raíz.

El motivo está en el propio producto. ChatGPT se entrena con feedback humano. Los usuarios premian las respuestas que les gustan. Y les gustan las que les dan la razón. Entonces la IA aprende a dársela siempre. Esto no es un error de diseño. Es el modelo de negocio.

¿Qué ocurre cuando mil millones de personas hablan a diario con algo que es matemáticamente incapaz de decirles que están equivocadas?