Join GitHub today
GitHub is home to over 31 million developers working together to host and review code, manage projects, and build software together.
Sign uplibrustdoc: set cap-lints=allow when compiling documentation #29529
Conversation
rust-highfive
assigned
Manishearth
Nov 2, 2015
This comment has been minimized.
This comment has been minimized.
|
Not entirely familiar with the complete reasoning behind capping ar Allow, deferring to Alex. |
rust-highfive
assigned
alexcrichton
and unassigned
Manishearth
Nov 3, 2015
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
|
I meant the unstable features stuff. I'm not sure if we should be capping at allow or warn here, and I'm not sure if this should even be a part of rustdoc (not cargo). Happy with the code changes, though, r=me if nobody has an issue with allow over warn and it being in rustdoc. |
This comment has been minimized.
This comment has been minimized.
|
So, one thing is that when docs run without lints, we get into situations like #29523 . I would have loved to have seen a "hey, this feature isn't needed" warning, so that we wouldn't have had the bug. |
This comment has been minimized.
This comment has been minimized.
|
So I guess we want to cap in docs but not doctests? |
This comment has been minimized.
This comment has been minimized.
|
@steveklabnik This change only applies to running Just confirmed that |
This comment has been minimized.
This comment has been minimized.
|
@Ryman ah! ignore my objection then, sorry :) |
This comment has been minimized.
This comment has been minimized.
|
@steveklabnik oh we actually have support for that, it's just not turned on by default. I also think that this is fine, seems kinda odd for a documentation generator to be warning you about various parts of your code! Thanks @Ryman! |
Ryman commentedNov 2, 2015
This sets the
cap-lintssetting to 'allow' for all doc compilations. There's precedent for this as rustdoc already whitelists unstable code when compiling documentation, with the expectation being that a regular compile will complain about any problems. I think the same justification applies here.Problem case in the wild: laumann/compiletest-rs#28
r? @Manishearth