It's a sad day for #rustlang: actix-web is dead. I am not feeling very good about our community today. https://words.steveklabnik.com/a-sad-day-for-rust …
-
-
Replying to @steveklabnik
I've been extremely disappointed with how everyone has acted in this repeating thunderstorm, thanks for writing this out
1 reply 0 retweets 29 likes -
Replying to @ManishEarth @steveklabnik
I've said this before, but we don't have enough documentation on unsafe best practices (nomicon helps, isn't everything). When it's your own opinion vs a Reddit mob and there's no "higher authority" to refer to, what the fuck do you think is going to happen
3 replies 0 retweets 34 likes -
Replying to @ManishEarth @steveklabnik
I mean in this case we do have a pretty clear statement in the nomicon to refer to in this case.pic.twitter.com/ulrYhH4wUH
1 reply 0 retweets 3 likes -
Granted, this could probably use a paragraph explaining that this is true even if the unsafe function is private, and that having this boundary is to help audit and maintain code, not just providing safe public APIs
1 reply 0 retweets 4 likes -
Replying to @sgrif @steveklabnik
I think there's a TON of nuance involved when it comes to private functions, around where the boundary should be. We typically pick "the module" but it's not quite that in particular the standard you have set makes it impossible to effectively write FFI code
4 replies 0 retweets 2 likes -
And yes, there's a generally agreed upon standard that the community has mostly settled on, that's close to what you say and also works for FFI, but it's super nuanced and not documented anywhere. Very understandable if someone makes the wrong conclusion from one line of docs
1 reply 0 retweets 2 likes -
Replying to @ManishEarth @steveklabnik
Sure. And nobody was pushing on "use unsafe correctly", which is part of the problem. IMO if the function folks were complaining about was itself marked `unsafe` it'd be fine
1 reply 0 retweets 1 like -
Replying to @sgrif @steveklabnik
yes, but that doesn't contradict what I'm saying. Yes, the code isn't fine and would be fine with that one change, however this is totally not obvious from current documentation. One line in the docs with nuance behind it doesn't cut it.
1 reply 0 retweets 1 like -
Replying to @ManishEarth @steveklabnik
Yes, I agree it needs to be expanded more than it already is
1 reply 0 retweets 2 likes
Honestly this makes me want to write a blog post about this, and how maintainers should consider it. Probably not a great time to post such a thing though >_>
-
-
Replying to @sgrif @steveklabnik
Yeah there are multiple blog posts worth of content to be written about this and I've wanted to for ages but never had the willpower or time
0 replies 0 retweets 4 likesThanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.