Conversation

Notices

1. Embed this notice
   Paul Cantrell (inthehands@hachyderm.io)'s status on Monday, 16-Oct-2023 11:33:20 JST Paul Cantrell

   • Ben Cohen

   This bang-on post from @airspeedswift is not getting the attention it deserves:
   https://mastodon.social/@airspeedswift/111234020859557888

   • Embed this notice
     Paul Cantrell (inthehands@hachyderm.io)'s status on Monday, 16-Oct-2023 12:09:21 JST Paul Cantrell

     in reply to

     • chris martens
     • Ben Cohen

     @chrisamaphone @airspeedswift
     Yes, in C/C++, which still accounts for a •substantial• portion of new code written, and of code in production, and of critical vulnerabilities.

     The C++ world in particular has been in turmoil in recent years over the battle between “just don’t write bugs” vs “mitigate with better tooling” vs “C/C++ are the cigarettes of p-langs.” This in particularly set the cat amongst the pigeons:

     https://media.defense.gov/2022/Nov/10/2003112742/-1/-1/0/CSI_SOFTWARE_MEMORY_SAFETY.PDF

   • Embed this notice
     chris martens (chrisamaphone@hci.social)'s status on Monday, 16-Oct-2023 12:09:22 JST chris martens

     in reply to

     • Ben Cohen

     @inthehands @airspeedswift … i feel like i’m missing some context. memory safety is generally taken as a given in most contexts, except where fine grained control is considered necessary, so where are folks arguing about this? rust vs c or something?

   • Embed this notice
     Paul Cantrell (inthehands@hachyderm.io)'s status on Monday, 16-Oct-2023 12:11:21 JST Paul Cantrell

     in reply to

     • chris martens
     • Ben Cohen

     @chrisamaphone @airspeedswift
     I think it’s hard for those of us who haven’t done serious work in C or C++ for ages (~25 years in my case), or who never have at all, to appreciate just what a large portion of the greater software development world those few remaining memory-unsafe languages still occupy.

   • Embed this notice
     Paul Cantrell (inthehands@hachyderm.io)'s status on Monday, 16-Oct-2023 12:12:24 JST Paul Cantrell

     in reply to

     • autism :verified:

     @jeff
     Yeah, no, sorry. This is wildly, fantastically, pants-on-head incorrect. You have a whole new set of very •different• problems.

   • Embed this notice
     autism :verified: (jeff@misinformation.wikileaks2.org)'s status on Monday, 16-Oct-2023 12:12:25 JST autism :verified:

     in reply to

     • chris martens
     • Ben Cohen
     @inthehands @chrisamaphone @airspeedswift you have the same problems in "safe" languages like rust, they just lie about it.
   • Embed this notice
     Paul Cantrell (inthehands@hachyderm.io)'s status on Monday, 16-Oct-2023 12:14:33 JST Paul Cantrell

     in reply to

     • autism :verified:

     @jeff You’re wrong. There’s even a study.
     https://dl.acm.org/doi/10.1145/3466642

   • Embed this notice
     autism :verified: (jeff@misinformation.wikileaks2.org)'s status on Monday, 16-Oct-2023 12:14:34 JST autism :verified:

     in reply to
     @inthehands go look at the CVEs for the rust stdlib in the last 2 years and tell me i'm wrong.
   • Embed this notice
     Paul Cantrell (inthehands@hachyderm.io)'s status on Monday, 16-Oct-2023 12:42:36 JST Paul Cantrell

     in reply to

     • chris martens

     @chrisamaphone No apology needed, and I hadn’t read any such implication in what you said! Happy I could help shed light.

   • Embed this notice
     chris martens (chrisamaphone@hci.social)'s status on Monday, 16-Oct-2023 12:42:37 JST chris martens

     in reply to

     • Ben Cohen

     @inthehands @airspeedswift sure, i completely acknowledge that it isn’t a world i’m privy to, hence asking for the context i was missing. i didn’t mean to imply “cases in which finer-grained control is necessary” was a small proportion of cases (though i likely would have underestimated them). that this is an ongoing conversation in the c++ community was i think the key info i was missing. thanks for explaining!

   • Embed this notice
     Paul Cantrell (inthehands@hachyderm.io)'s status on Monday, 16-Oct-2023 12:44:25 JST Paul Cantrell

     in reply to

     • Zorro Notorious MEB 😡

     @AlgoCompSynth
     The irony is something like generating audio on a Pi is totally something a memory-safe lang could do. Opening the door to memory errors for no reason!

   • Embed this notice
     Zorro Notorious MEB 😡 (algocompsynth@ravenation.club)'s status on Monday, 16-Oct-2023 12:44:29 JST Zorro Notorious MEB 😡

     in reply to

     @inthehands Case in point for the latter - the Raspberry Pi Pico examples for generating audio via PWM or an I2S DAC.

   • Embed this notice
     Zorro Notorious MEB 😡 (algocompsynth@ravenation.club)'s status on Monday, 16-Oct-2023 12:44:30 JST Zorro Notorious MEB 😡

     in reply to

     @inthehands Oh, we appreciate it, but it falls into the class of accepting the things we cannot change.

     I've never had to learn C or C++ and now I have no desire to either. And I am coming to the conclusion that a project whose only documentation is C source code is a waste of my time.