Forgot your password?
typodupeerror

Comment Re: Cool (Score 1) 90

I think if you took all the software devs on the planet and gave them unlimited budget and time you might just possibly be able to get near that goal, although you might have to outsource bits of the task to other planets. We've been using AI to audit our code, which is well-written, commented, and documented (the boss is a stickler for that), and it's taking us as much as several weeks per item to remediate it if it's a deep-seated design issue that's been in there for 20-30 years. That's for good-quality code. Now imagine what it's going to take for Cisco-quality code. And as others have pointed out, a lot of the code in there will be to deal with hardware bugs and quirks, or bugs and quirks with other devices and systems, documented only in the memory of ol' Bill who retired 18 years ago and is now in a care home and dribbles a lot.

The idea of doing all of this in a few weeks is beyond delusional.

Comment Re:Cool (Score 3, Interesting) 90

It sounds like someone at Cisco took whatever drugs Andy Yen is currently tripping on. This rewrite will have one of two outcomes, it'll either have to be bug-for-bug compatible with all of Cisco's existing products in order to no lose the decades of knowledge and existing configs that people have built up working with their stuff, or if it's a full rewrite with new capabilities and features and bugs, so essentially a new product, then why on earth would anyone go back to Cisco and not get something better and cheaper from anyone but Cisco?

Comment Re:"stop lovingly handcrafting all our code" (Score 2) 164

It was "AI in Open-Source Cryptography: Maintainers Compare Notes", from the programme here. They usually put the videos online within a few weeks, currently only the 2025 ones are up. For the exact wording (mine might not be 100% word-for-word correct), wait for the video to be posted. It was a really interesting panel covering, well, what the title says.

Comment Re:Proprietary software needs to be abandoned. (Score 2) 68

Linux distros also relentlessly expire signing keys so that even if the packages are still there the system will refuse to install them because some signing key has hit its arbitrary EOL. The result is an infinite number of posts to forums telling users how to completely disable all verification of packages in order to allow an older system to be updated.

The fix is pretty trivial, don't arbitrarily decide that a key that was perfectly fine on Tuesday can no longer be used on Wednesday, but that would sap and impurify the precious signing fluids so isn't permitted.

Comment Re:How Hard Can The Decision Be? (Score 1) 19

It's relatively easy to fire a certificate issuer; but it can get surprisingly fiddly surprisingly fast to ensure that all your stuff actually distrusts one

And does that matter? This fiasco has been running since April 2025 (not April of this year, 1 1/2 years ago) and no attackers have bothered with it because the PKI isn't defending against anything they're doing. Distrust, don't distrust, it may give you a warm feeling but the attackers won't care.

Incidentally, are all your PHP setups patched?

Comment Re:Vibe coded huh? (Score 2) 71

It's also completely the wrong thing to use. RS-coding will deal with small amounts (bits/bytes depending on the layout) of data corruption, not general filesystem corruption. And the overhead is enormous, unless you're streaming data through hardware that can correct and reassemble on the fly it's not practical, and for filesystems even useful.

This proposal is a bit like the hypothetical "why didn't we use the space shuttle to rescue the Apollo 13 astronauts" question. Yeah, you just go ahead and do that then.

Slashdot Top Deals

It's a naive, domestic operating system without any breeding, but I think you'll be amused by its presumption.

Working...