[Haskell-cafe] use of -Werror

[Adam Bergmark]

-Werror has unintended consequenses in a release. Say you are using
transformers/mtl and you want to stay backwards compatible, you can't use
Except since it's just in the new versions, but with -Werror you couldn't
use Error since it would produce deprecation warnings. CPP isn't an option
since that would change the implementation. I don't like the idea of
disabling deprecation warninings since that setting has at least module
level granularity, if other deprecations pop up I want to see those! By
enabling -Werror you are turning a non breaking change (deprecation) into a
breaking one and not even an upper bound of <= A.B.C.D would be safe since
the number of components are arbitrary.

The granularity of disabled warnings is part of the problem, it'd be really
nice if you could disable deprecation warnings for specific identifiers,
same with a lot of the other warnings.

- Adam



On Tue, Jul 1, 2014 at 7:34 PM, Carter Schonwald <carter.schonwald at gmail.com
> wrote:

> yeah, -Wall makes enough noise  to have a good ROI
>
> What I do though, is enable Werror on my Travis Ci formulae.
>
>
> On Tue, Jul 1, 2014 at 1:30 PM, Brandon Allbery <allbery.b at gmail.com>
> wrote:
>
>>
>> On Tue, Jul 1, 2014 at 1:25 PM, Johan Larson <johan.g.larson at gmail.com>
>> wrote:
>>
>>> Is there a consensus within the Haskell community on the use of -Werror
>>> in packages uploaded to Cabal?
>>>
>>
>> GHC's interpretation of Haskell, and the warnings it provides, change
>> often enough that -Werror is generally a good way to lock a package to one
>> or a small number of GHC releases. As such, last I heard Hackage complained
>> if -Werror was found in a package's cabal file.
>>
>> --
>> brandon s allbery kf8nh                               sine nomine
>> associates
>> allbery.b at gmail.com
>> ballbery at sinenomine.net
>> unix, openafs, kerberos, infrastructure, xmonad
>> http://sinenomine.net
>>
>> _______________________________________________
>> Haskell-Cafe mailing list
>> Haskell-Cafe at haskell.org
>> http://www.haskell.org/mailman/listinfo/haskell-cafe
>>
>>
>
> _______________________________________________
> Haskell-Cafe mailing list
> Haskell-Cafe at haskell.org
> http://www.haskell.org/mailman/listinfo/haskell-cafe
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/haskell-cafe/attachments/20140701/345ae6bb/attachment.html>