๐–‡๐–Š๐–”๐–—๐–“ || @beorn@rtfm.fr (@beorn__) profile picture

Been pushing buttons for too long to remember

Joined April 2011
Photo by ๐–‡๐–Š๐–”๐–—๐–“ || @beorn@rtfm.fr (@beorn__)Photo by ๐–‡๐–Š๐–”๐–—๐–“ || @beorn@rtfm.fr (@beorn__)Photo by ๐–‡๐–Š๐–”๐–—๐–“ || @beorn@rtfm.fr (@beorn__)Photo by ๐–‡๐–Š๐–”๐–—๐–“ || @beorn@rtfm.fr (@beorn__)Photo by ๐–‡๐–Š๐–”๐–—๐–“ || @beorn@rtfm.fr (@beorn__)Photo by ๐–‡๐–Š๐–”๐–—๐–“ || @beorn@rtfm.fr (@beorn__)Photo by ๐–‡๐–Š๐–”๐–—๐–“ || @beorn@rtfm.fr (@beorn__)Photo by ๐–‡๐–Š๐–”๐–—๐–“ || @beorn@rtfm.fr (@beorn__)Photo by ๐–‡๐–Š๐–”๐–—๐–“ || @beorn@rtfm.fr (@beorn__)Photo by ๐–‡๐–Š๐–”๐–—๐–“ || @beorn@rtfm.fr (@beorn__)
๐–‡๐–Š๐–”๐–—๐–“ || @[email protected] retweeted
Microsoft's preliminary report on last week's M365 outage is out. Here's what happened: - Outlook and Exchange Online were down from 11:20 AM ET Thursday to 11:10 AM ET Friday, and mail was delayed - Cause: a July update to the M365 network gateway servers had a latent bug that sat in production for over 3 months - Load tests missed it because the lab hardware had fewer cores than production, and the bug gets worse on newer machines - Trigger: Teams failed over onto the same gateway Exchange uses, which was already 14% short from an unrelated issue - At peak US traffic, over 80% of the gateway servers went unresponsive in under 2 hours - Teams, OneDrive and SharePoint could reroute. Exchange couldn't, so email stayed down all day. - Restarts didn't hold, and the old change couldn't be rolled back with the usual flags - Fix: an overnight update restored a known good version of the code, then servers were patched and restarted in batches - Changes so far: more capacity, better monitoring, guardrails against cascades, and a longer delivery window so queued mail didn't bounce - Final report due Nov 20 Preliminary Lesson: If your load test doesn't look like production, it isn't a load test
54
103
10
833
238,758
๐–‡๐–Š๐–”๐–—๐–“ || @[email protected] retweeted
Kate ็กฌไปถ้€†ๅ‘ไธŽ็”ตๅŽ‹ๆฏ›ๅˆบ่งฃ้”ๅฎŒๆ•ดๅค็›˜๏ผš blog.kate.land/posts/using-vโ€ฆ Hacker News ่ฎจ่ฎบไธฒ๏ผˆ353 ๅˆ†๏ผ‰๏ผš news.ycombinator.com/item?idโ€ฆ
6
31
2
255
25,372
๐–‡๐–Š๐–”๐–—๐–“ || @[email protected] retweeted
Both NeXT and 3DFX tried to โ€œdo the right thingโ€ with color spaces, and it wound up generally making them look worse than competing options. I didnโ€™t really understand it properly until years later. It is natural to think that a pixel with a value of 1.0 is emitting twice the photons as a pixel with a value of 0.5, so you can do the obvious linear math to calculate partial pixel coverage, blending a foreground over a background, or interpolating between colors in a texture. If displays are โ€œlinearโ€, that is correct. NeXT and 3DFX were much closer to linear than other contemporary displays. NeXT had crossover from the deep graphics understanding at Pixar, and 3DFX had the SGI heritage. They did this intentionally. If you have 10+ bits per channel like the professional rendering systems, linear displays are The Right Thing. Unfortunately, NeXTStation Colors were 4 bits per channel and 3DFX was 5 or 6 bits per channel, and thereโ€ฆ it was not the right thing. Human eyes arenโ€™t linear intensity sensors. We mostly register relative changes in intensity, so we can tell the difference between 0.01 and 0.02 as a huge doubling, while 0.99 to 1.00 is almost imperceptible. This distribution of intensities is often referred to as the โ€œgammaโ€ of the display system, but the actual curves like sRGB are a bit more complicated. If you only have a small number of discrete pixel values, you definitely donโ€™t want to spread them evenly from 0 to 1! If you do, you will have good precision in the brights, but very stark โ€œintensity bandsโ€ in the dark colors. If everything was authored specifically for 3DFX, you would still have the fundamental problem with the dark bands, but more accurate blending and filtering. The problem was that artists made all of their textures in paint programs on conventional non-linear displays, so when they were drawn by 3DFX cards, they got the characteristic โ€œwashed out colorโ€ look. You can mathematically correct textures for differences in display gamma, but I didnโ€™t really understand the issue back in the glQuake days, and we were so enthralled by the absolute performance of 3DFX that the color fidelity didnโ€™t get as much attention as it should have. Even today, tons of games donโ€™t properly handle gamma across their textures and render targets. High resolutions and 24 bit color masks most of it, but you still see it in some slight shimmerings in high contrast textures and edges.
Does anybody know why 3DfX 1 cards had such a washed out color output? In my research, circa '97 I found many people complaining about the colors. Is it gamma? 16-bit color? What can it be?
152
493
45
6,569
196,783
๐–‡๐–Š๐–”๐–—๐–“ || @[email protected] retweeted
Weโ€™ve monitored the status of the changes applied to improve serviceability and confirmed they're not taking as expected. Weโ€™re identifying alternate workstreams that can be applied to increase capacity in the face of sustained traffic management imbalances. More info under EX1550543 and MO1550821 in the Microsoft 365 admin center, as well as msft.it/6013xBHsf if the admin center is not available to you.
2,170
1,214
1,847
7,993
10,460,465
๐–‡๐–Š๐–”๐–—๐–“ || @[email protected] retweeted
I CANNOT believe the report: it proves that eMudhra is a clown show, and was never fit to be a CA: The point of a HSM (hardware security module) is that the private key NEVER leaves it. Guess what eMudhra did: it turned off FIPS mode (WTF?!) and copied out the private key ๐Ÿ˜ฑ
118
290
27
4,960
296,712
๐–‡๐–Š๐–”๐–—๐–“ || @[email protected] retweeted
AMD keeps its microarchitecture names simple: Zen 2, Zen 3, โ€ฆ Zen 6. Zen 6 is new and most of us still lack access. Chips and Cheese measured what changed from Zen 5. Zen 6 sustains 5.96 64-bit adds per cycle versus Zen 5โ€™s 5.25 (same six ALUs). (Yes! Your processor can do that many additions per cycle!) It also recovers faster from branch mispredictions: roughly 3โ€“3.5 ns extra versus 3.5โ€“4 ns on Zen 5. That's an important gain because lots of branchy code is limited by branch misprediction penalties. Overall, Zen 6 processors today can match the best of the previous generation (Zen 5) running at high clocks with more cache. So we might get an realistic performance boost of about 20% with comparable chips. Source: chipsandcheese.com/p/amds-zeโ€ฆ
9
13
2
231
13,402
๐–‡๐–Š๐–”๐–—๐–“ || @[email protected] retweeted
Un vrai rรฉgal ร  regarder ! #cablemanagement
41
162
13
1,188
55,332
๐–‡๐–Š๐–”๐–—๐–“ || @[email protected] retweeted
Today we're launching cve.now ๐Ÿšจ Yesterday was a rough day for security. php2shell, LiteLLM, Railway... How are we supposed to stay on top of all of this? cve.now aggregates 4 vulnerability sources into one free feed and uses AI to explain each one in plain English. No login. cve.now
128
273
81
1,858
588,501
๐–‡๐–Š๐–”๐–—๐–“ || @[email protected] retweeted
Surprisingly good video on a 1 cpu server challenge. Mad respect to the crazy person writing io_uring in assembly! youtu.be/-BFt49ula08?si=EJAtโ€ฆ
15
49
6
746
31,725
๐–‡๐–Š๐–”๐–—๐–“ || @[email protected] retweeted
๐Ÿšจ THE INTERNET BROKE TODAY ๐Ÿšจ Oct 21, 2026 will go down in history 1. eMudhra : Private key was literally "4". 8 keys forged ANY website for 17 months. 60M+ certs at risk. 2. SBI CONFIRMED. 3,897 customers' mobile, DOB, address + loan data found on dark web. 3. Rajasthan Police ๐Ÿ‡ฎ๐Ÿ‡ณALLEGED hack. Group claims 70,000 identity records + 59,460 numbers for sale. Unverified. 4. php2shell [9.2] Unauth RCE in PHP. 70% of the web. WordPress, Drupal, Laravel. CISA: Patch or GO OFFLINE. 5. GitLab [10.0/10] PERFECT SCORE. Unauth RCE, no login needed. 6. Node.js [CVE-2026-111497] Brotli decoder RCE + TLS bypass. 10 CVEs patched today. Also blocked eMudhra. 7. Railway BREACHED TODAY. Env variables decrypted. Your API keys & DB passwords exposed. Rotate all. 8. OpenCode BREACHED. OpenAI, Anthropic, Google API keys exposed for Console users. Distillation attack. Bills run up in $$$$. 9. LiteLLM PyPI v1.107.0 was MALWARE today morning. Steals credentials + backdoor. 10. Plex2Shell ACTIVELY EXPLOITED. Hackers installing cryptominers LIVE. CISA KEV, deadline Oct 23. AI found the bugs. AI is exploiting them. Humans can't patch fast enough. Patch. Rotate. Revoke. Update. RT to save someone's server โ™ป๏ธ #CyberSecurity #DataBreach #India #php2shell
131
1,576
191
7,339
735,198
๐–‡๐–Š๐–”๐–—๐–“ || @[email protected] retweeted
Emudhra, Opencode, Pingcap, Orange, PHP, Node, LiteLLM... what's next? Critical GitLab Flaw (CVSS 10.0) Could Allow Unauthenticated RCE Across 12 Years of Releases thehackernews.com/2026/10/crโ€ฆ
3
2
1
21
2,813
๐–‡๐–Š๐–”๐–—๐–“ || @[email protected] retweeted
I donโ€™t understand? How
a trusted intermediate CA just "leaked" (the private key was 4)
150
558
206
9,309
1,827,487
๐–‡๐–Š๐–”๐–—๐–“ || @[email protected] retweeted
โ€ผ๏ธ LA FAILLE PHP ร‰TAIT CONNUE DEPUIS 2024... La PHP Foundation vient de publier le rรฉcit officiel et il fait froid dans le dos Les premiers signes ont รฉtรฉ repรฉrรฉs en 2024, pendant un audit financรฉ par l'ร‰tat allemand. Puis le correctif est restรฉ bloquรฉ des mois. La fondation parle elle-mรชme d'un "รฉchec structurel" Et la sortie a tournรฉ au chaos - Le 20 octobre, les rรฉfรฉrences de la faille fuitent en ligne - Une sociรฉtรฉ de sรฉcuritรฉ publie par erreur une vidรฉo de son attaque - Ce matin, Chainguard rompt l'embargo et diffuse le correctif plus de 6 heures avant tout le monde ร€ SAVOIR... selon la @ThePHPF la sรฉcuritรฉ d'un langage qui fait tourner banques, boutiques et services publics repose sur "un petit groupe" de dรฉveloppeurs !! Ces mรชmes services publics que des ร‰tats ont dรป dรฉbrancher ce matin Bref... Des budgets entiers pour surveiller les citoyens... Une poignรฉe de dรฉveloppeurs pour protรฉger leurs serveurs ๐Ÿคฆโ€โ™‚๏ธ
We have posted more information about today's security releases, including the sequence of events that led to the discovery, how the PHP security team is improving their processes, and what we need from you to continue this critical work. #php #security thephp.foundation/blog/2026/โ€ฆ
25
372
34
1,333
117,669
๐–‡๐–Š๐–”๐–—๐–“ || @[email protected] retweeted
Not just PHP: Node also announced security releases nodejs.org/en/blog/vulnerabiโ€ฆ
14
219
32
560
152,037
๐–‡๐–Š๐–”๐–—๐–“ || @[email protected] retweeted
For the ZX Spectrum lovers, there was enough documentation around to also simulate the whole Spectrum PCB with very precise ULA emulation (cross checked with the actual gates implementing it).
19
24
7
275
14,925
๐–‡๐–Š๐–”๐–—๐–“ || @[email protected] retweeted
simd package is going to make Go great again! For creating faster Go programs than in Rust :) Thanks to @jespinog for writing such a clear article about simd support in Go!
My article about SIMD and Go is out! In this one I explain how one CPU instruction can work on 8 numbers at once, and how you can now use that from plain Go with the experimental simd/archsimd and simd packages. My intention with this post is to explain in simple terms how SIMD works at CPU level, to give you the mental model on how to use it, and then give you enough information about how you use it from Go to start working with it. You can read it here: internals-for-interns.com/poโ€ฆ #Go #SIMD
5
5
143
6,486
๐–‡๐–Š๐–”๐–—๐–“ || @[email protected] retweeted
I feel bad man ๐Ÿ˜ญ Apologies to the thousands of VPS and compute engineers worldwide spending their weekend emergency patching and rebooting hypervisors because of CVE-2026-99999. Security bugs really do love getting disclosed right before the weekend.
Introducing PwnEscape ๐Ÿฆ‰: a full KVM guest-to-host escape via a Linux kernel networking flaw affecting industry standard hypervisors. CVE-2026-99999 went from untrusted VM code to host-kernel memory reads, controlled writes, and root command execution. More below:
53
297
28
4,054
302,295
๐–‡๐–Š๐–”๐–—๐–“ || @[email protected] retweeted
Soon on GitHub. #C64 #retrocomputing
27
75
3
992
32,102