text 1.0.0.0 producing cabal hell?

Check this out: http://packdeps.haskellers.com/reverse/text 306 outdated dependencies! I am guessing that most libraries could easily change the boundaries to
=0.11 && <1.1 http://hackage.haskell.org/package/digestive-functors-blaze because the API hasn't changed a lot. As far as I can see only a function was added. Or is everybody using sandboxing tools? Maybe the author of a library should have the right to change the boundary of all reverse dependencies. So that Brian has to adjust all 306 boundaries if he knows that this would work :-) Just the boundaries, not the code.
-Tillmann

On 2013-12-17 22:03, Tillmann Vogt wrote:
Check this out:
http://packdeps.haskellers.com/reverse/text
306 outdated dependencies! I am guessing that most libraries could easily change the boundaries to
=0.11 && <1.1 http://hackage.haskell.org/package/digestive-functors-blaze because the API hasn't changed a lot. As far as I can see only a function was added. Or is everybody using sandboxing tools? Maybe the author of a library should have the right to change the boundary of all reverse dependencies. So that Brian has to adjust all 306 boundaries if he knows that this would work :-) Just the boundaries, not the code.
Devil's advocate question: Should bumping a package major version *even if there are no incompatible changes* be allowed according to the PVP? It seems it causes a lot of inconvenience for people who are following the PVP in their dependency declarations. @bos: Of course, I understand that this was probably to signal stability, and as such fully support the bump (not that you need my support), I'm just playing devil's advocate regarding the PVP. Regards,

Is there any way we can get packdeps to show the current maintainer (name +
email) so we can easily ctrl+f ourselves in the list?
- Clark
On Tue, Dec 17, 2013 at 4:20 PM, Bardur Arantsson
On 2013-12-17 22:03, Tillmann Vogt wrote:
Check this out:
http://packdeps.haskellers.com/reverse/text
306 outdated dependencies! I am guessing that most libraries could easily change the boundaries to
=0.11 && <1.1 http://hackage.haskell.org/package/digestive-functors-blaze because the API hasn't changed a lot. As far as I can see only a function was added. Or is everybody using sandboxing tools? Maybe the author of a library should have the right to change the boundary of all reverse dependencies. So that Brian has to adjust all 306 boundaries if he knows that this would work :-) Just the boundaries, not the code.
Devil's advocate question: Should bumping a package major version *even if there are no incompatible changes* be allowed according to the PVP? It seems it causes a lot of inconvenience for people who are following the PVP in their dependency declarations.
@bos: Of course, I understand that this was probably to signal stability, and as such fully support the bump (not that you need my support), I'm just playing devil's advocate regarding the PVP.
Regards,
_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
-- Clark. Key ID : 0x78099922 Fingerprint: B292 493C 51AE F3AB D016 DD04 E5E3 C36F 5534 F907

If you go to a URL like
http://packdeps.haskellers.com/feed?needle=cgaebel
you should get a nice report of all packages matching that string, along
with their restrictive dependency bounds.
There's an RSS feed also (linked from that page), if you're into that sort
of thing.
John L
On Tue, Dec 17, 2013 at 1:24 PM, Clark Gaebel
Is there any way we can get packdeps to show the current maintainer (name + email) so we can easily ctrl+f ourselves in the list?
- Clark
On Tue, Dec 17, 2013 at 4:20 PM, Bardur Arantsson
wrote: On 2013-12-17 22:03, Tillmann Vogt wrote:
Check this out:
http://packdeps.haskellers.com/reverse/text
306 outdated dependencies! I am guessing that most libraries could easily change the boundaries to
=0.11 && <1.1 http://hackage.haskell.org/package/digestive-functors-blaze because the API hasn't changed a lot. As far as I can see only a function was added. Or is everybody using sandboxing tools? Maybe the author of a library should have the right to change the boundary of all reverse dependencies. So that Brian has to adjust all 306 boundaries if he knows that this would work :-) Just the boundaries, not the code.
Devil's advocate question: Should bumping a package major version *even if there are no incompatible changes* be allowed according to the PVP? It seems it causes a lot of inconvenience for people who are following the PVP in their dependency declarations.
@bos: Of course, I understand that this was probably to signal stability, and as such fully support the bump (not that you need my support), I'm just playing devil's advocate regarding the PVP.
Regards,
_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
-- Clark.
Key ID : 0x78099922 Fingerprint: B292 493C 51AE F3AB D016 DD04 E5E3 C36F 5534 F907
_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
participants (4)
-
Bardur Arantsson
-
Clark Gaebel
-
John Lato
-
Tillmann Vogt