kaszanka

> Why isn't there a mechanism where a domain owner can specify which CA it allows?

There is DANE (TLSA) but seemingly only supported by e-mail software and not web browsers.

Looks like each message takes up a flat 15% of the 5-hour limit on the $20 subscription.

Only tangentially related, but knowing that remote attestation and co. will only ever get... "better" fills me with impotent rage like almost nothing else. https://www.lafkon.net/tc/ (2004) was so very prescient.

Is auto mode only in the IDE? I'm not seeing it in the CLI on my end, version 1.2.14.

If the direct link isn't working for anyone else (looks like it's some AWS pre-signed one maybe tied to IP): https://github.com/user-attachments/files/31802150/EXTERNAL....

Trending links don't work on Nitter, so here's the tweet: https://nitter.net/jietang/status/2067580270078030088

> Because this version of GPT‑5.6 Sol is optimized for everyday chats, it will only be available in the Chat experience in ChatGPT. The version of GPT‑5.6 Sol that powers Work and Codex is not changing as part of this release.

Hm, does this mean that 5.6 Pro in ChatGPT web is somehow different/not as good now? I found it really good for code review (upload your repo and patch and off it goes).

"Anti-Evil Operations" is a pretty grandiose name for spam filtering. Also I liked the House of Leaves reference

People have already been imprisoned for this, one case I can think of off the top of my head is https://en.wikipedia.org/wiki/Goatse_Security#AT&T/iPad_emai....

I wonder how Turbo Vision (still the only good TUI framework) does on screen reader accessibility.

Surprisingly only the headings (2.05) and links (3.72) fail the Firefox accessibility check, the body text is 5.74. But subjectively it seems worse and I definitely agree with you that the contrast is too low.

Is https://github.com/google-gemini/gemini-cli not 'the flagship agent' itself? It looks that way to me, for example here's a part of the prompt https://github.com/google-gemini/gemini-cli/blob/e293424bb49...

Is it AGentIc ai infrastructure? Or AGentic aI infrastructure? Or AGentic ai Infrastructure?

I expected better from the people who brought us the ARM architecture, with A, R and M profiles.

Also in Polish, which would mean "dog".

> no winhttp.dll, wininet.dll, or ws2_32.dll. offline validation only. all crypto is local, so theoretically extractable.

You can't possibly know that by the mere lack of these DLLs from the import directory.

TIL! That's neat, I wonder how much RAM that client uses compared to the desktop one.

That's still not a fair comparison, because on a console you don't have the option to do any of that.

> after booting my VM KDE just flashed because my external GPU was gone but everything went back working without a need for relogin.

What GPU are you using and how did you configure this, if you don't mind me asking? On my end I just can't unload the driver for it if I let KDE start with the external GPU available.

Good post on troubleshooting the failure to boot, but from the title I was kind of hoping for something like decryption and analysis of the blobs' contents, rather than just metadata. Very "cool" that 3 megabytes of unauditable malware (the public blobs) are still not enough to even boot the platform...

Anyone can do this by changing the author and committer date. Take a look at https://github.com/dspinellis/unix-history-repo, linked in the comments here, for example.

This is basically what most of the challenge types in go-away (https://git.gammaspectra.live/git/go-away/wiki/Challenges) do.

Probably even non-vetted firmware-to-browser chains, by requiring boot attestation to open a TLS connection or something.

To be even more pedantic, it's also not present on Android.

Oh joy. So when this sort of stuff comes to mobile phones, at least when the McDonalds app refuses to start on your pocket general purpose computer (because it's not running software that Google considers 'trustworthy') you'll be able to confidently say that the RTL for the part of the chip that is ultimately responsible for betraying your interests is open source. Surely consolation enough for missing out on your burger discount.

I like to bring up McDonalds as an example because IIRC it requires the highest, 'strong integrity' verdict from SafetyNet/Play Integrity/nom-du-jour. Maybe they should rename it to something with Open in the name when OpenTitan comes to Chromebooks.

BAR resizing (which is all that Smart Access Memory means, as far as I can tell) is just done by the kernel writing some value to the right field in PCI configuration space, why would it require firmware support (and if so, why would CSM preclude that support)?

Here's the live recording link for those who missed it and are too impatient to wait for the cleaned-up one: https://streaming.media.ccc.de/38c3/relive/233cb1d4-4833-538...

This is well on its way thanks to Microsoft's aggressive push to put a TPM in every Windows 11 PC.

This time it's sadly just another case of a web developer who thinks he knows better than you what browser you should use, it works fine with

  // ==UserScript==
  // @name     oasis.decart.ai Firefox fix
  // @version  1
  // @run-at   document-start
  // @match    https://oasis.decart.ai/*
  // ==/UserScript==
  
  unsafeWindow.chrome = 1;

You can actually turn off Stories in the settings, then the option also disappears from the bottom bar.

How do they ensure this under Linux?