27 votes

A Sad Day For Rust

11 comments

  1. [3]
    HoolaBoola
    Link
    As a newbie Rust developer, I hadn't really even heard about actix-web, let alone the heated debate concerning it. However, I'm still a bit sad that something like this ended up happening. I hope...

    As a newbie Rust developer, I hadn't really even heard about actix-web, let alone the heated debate concerning it. However, I'm still a bit sad that something like this ended up happening. I hope the developer returns to developing open source some day

    7 votes
    1. aphoenix
      Link Parent
      If I had to guess, I'd wager that he will not. The cycle of releasing an open source project all too frequently goes like this: get excited about something and write some code for it release the...

      I hope the developer returns to developing open source some day

      If I had to guess, I'd wager that he will not. The cycle of releasing an open source project all too frequently goes like this:

      • get excited about something and write some code for it
      • release the source
      • work to fix the issues that people bring up, merge pull requests, be engaged
      • ignore the trolls
      • some flashpoint that consists of a non-trivial amount of angry people about something you no longer have time to fix quickly
      • personal attacks
      • personal attacks ++
      • personal attacks all the way down
      • close the fucking project and forget about it

      Don't get me wrong, I love Open Source, but there are just too many assholes.

      15 votes
    2. reese
      Link Parent
      Sounds to me like the maintainer was dogpiled for no good reason. If there is/was truly unsafe code in actix-web, and the maintainer refused to pull in patches people wanted, then they could have...

      Sounds to me like the maintainer was dogpiled for no good reason.

      If there is/was truly unsafe code in actix-web, and the maintainer refused to pull in patches people wanted, then they could have forked the repo, called it actix-web-safe or something snarky like that, publicized it, updated it, and moved on. It sounds to me like the maintainer was receptive to some patches, but not to others, perhaps because he had concerns about quality, style, or whatever. Maybe he's just busy working or has too much shit going on.

      And maybe he doesn't/didn't fully understand the safety of his code, and was therefore hesitant to accept PRs when he shouldn't have been. Maybe he needs time to mull over the docs, which is fine. What's not fine is for hordes of people, most of which I can assure you do not understand concurrency, to jump on a bandwagon to collect the maintainer and burn him at the stake. Says a lot more about them than it does him. Talk about biting the hand that feeds.

      I wouldn't blame him if he wants go pre-anger-management Linus Torvalds on people, I would be irate.

      I should also mention that I don't know anything about Rust, so I can't speak more specifically or technically on this matter. Maybe I'm wrong. Maybe the maintainer is a giant asshole, but I strongly doubt it. He maintained this codebase on his own volition for three years so others could benefit from his work. Again, the common convention is just to fork stuff that isn't actively maintained, or is going in a direction people disagree with, so long as it's permissively licensed for doing so.

      11 votes
  2. HoolaBoola
    Link
    The repo is back up with a different maintainer!

    The repo is back up with a different maintainer!

    4 votes
  3. skybrian
    Link
    Setting expectations is really important, particularly when you make unusual choices. If the project's home page had said that they aim to build the fastest web framework, not necessarily the...

    Setting expectations is really important, particularly when you make unusual choices. If the project's home page had said that they aim to build the fastest web framework, not necessarily the safest one, then it would attract users who are okay with that and there wouldn't be much to complain about.

    3 votes
  4. [3]
    vegai
    Link
    This sounds like some sort of a failure at project management level. The focus was mostly on code and not so much on social interactions and processes. These things happen all the time, but when...

    This sounds like some sort of a failure at project management level. The focus was mostly on code and not so much on social interactions and processes. These things happen all the time, but when it happens to a high profile library such as this one, everyone notices. And of course open source libraries don't tend to hire project managers.

    Personally, I was less than impressed by Actix when I realized its high benchmark values were partly due to clevereness in the benchmark code, something which you wouldn't normally do in application code. Reminds me a bit of the Volkswagen thing, although obviously much much less serious and totally legal in every way.

    2 votes
    1. [2]
      spit-evil-olive-tips
      Link Parent
      This sounds interesting, could you elaborate? I don't follow Rust terribly closely so I hadn't heard about this before.

      Personally, I was less than impressed by Actix when I realized its high benchmark values were partly due to clevereness in the benchmark code, something which you wouldn't normally do in application code.

      This sounds interesting, could you elaborate? I don't follow Rust terribly closely so I hadn't heard about this before.

      1 vote
      1. vegai
        (edited )
        Link Parent
        https://github.com/TechEmpower/FrameworkBenchmarks/issues/3845 illustrates some of it. I'm not entirely sure if this still applies today. Practical example:...

        https://github.com/TechEmpower/FrameworkBenchmarks/issues/3845 illustrates some of it. I'm not entirely sure if this still applies today.

        Practical example:
        https://github.com/TechEmpower/FrameworkBenchmarks/blob/master/frameworks/Rust/actix/src/main_raw.rs#L26

        The headers, along with content lengths are inserted as static strings.

        Seems like Actix's maintainer saw being on top of those benchmarks as a target, which is I think is totally fine goal for him. I just don't like it when it gets to "end justifies the means" territory.

        3 votes
  5. [3]
    HoolaBoola
    Link
    I assume those tags were created automatically? I totally forgot to add them myself :D I'm actually a bit impressed by them, pretty accurate for what I assume was automatic.

    I assume those tags were created automatically? I totally forgot to add them myself :D

    I'm actually a bit impressed by them, pretty accurate for what I assume was automatic.

    2 votes
    1. [2]
      hungariantoast
      (edited )
      Link Parent
      I wish they were added automatically, that would be super neat! Unfortunately though, they were added by me, a very boring human... In addition to being able to edit tags on your own topics, you...

      I wish they were added automatically, that would be super neat!

      Unfortunately though, they were added by me, a very boring human...

      In addition to being able to edit tags on your own topics, you can also get permissions to edit the tags of other users' topics, but you have to ask Deimos for them.

      As for seeing who edited the tags on your topic:

      If you are on mobile, in the top right there should be a button labeled Sidebar. Tapping it will reveal the sidebar.

      On desktop, the sidebar is just visible on the right side of the screen.

      At the bottom of the sidebar you should see something like: ▶ Topic log (1)

      That means one edit has been made to your topic. Click it to reveal what users did what things.

      So far, users can be given permissions to edit tags, titles, and links, as well as move topics to different groups. Each time an action like this is performed, the topic log states which user did what action.

      Finally, if you have not checked it out yet, there is a link to the site's documentation at the bottom of every page. It contains info for just about every feature the site has.

      15 votes
      1. HoolaBoola
        Link Parent
        Ah, thanks for the tags, then! And for explaining this, apparently there's a bunch of stuff I did not know about Tildes :D

        Ah, thanks for the tags, then! And for explaining this, apparently there's a bunch of stuff I did not know about Tildes :D

        7 votes
  6. Comment deleted by author
    Link