[darcs-users] darcs patch: Refactor version machinery in Setup.lhs. (and 3 more)

Eric Kow kowey at darcs.net
Sun Jan 25 17:50:35 UTC 2009


Hi,

On Sat, Jan 24, 2009 at 23:08:55 +0100, Petr Rockai wrote:
> NB. These patches break autoconf horribly (that was not the intention, but
> there's no other way around). I will also send in a patch to fix the hspwd
> problem somehow and I will ask all these patches to be applied on mainline. We
> will need to switch all of the testing infrastructure over to cabal along
> that. I will rely on volunteers to fix up autoconf, if there are any

What would we need to do to get the autoconf stuff working again?

> (or otherwise, send a patch removing the autoconf build altogether).

You know exactly what reaction I'm going to have to this one :-D

I'm happy to switch the defaults, but I think it would be prudent to
keep the autoconf stuff up to date for at one more release, if not two.
Carefully, carefully.  It may slow us down, but I think it pays off.

> Sat Jan 24 22:10:15 CET 2009  Petr Rockai <me at mornfall.net>
>   * Refactor version machinery in Setup.lhs.
> 
>   Sanctify the notion that 97, 98 and 99 are special in a darcs version
>   number. Assign fancy names to them, for prettier darcs --version.

I'm guessing this one is perfectly safe for me to apply immediately
 
> Sat Jan 24 22:51:49 CET 2009  Petr Rockai <me at mornfall.net>
>   * Replace ThisVersion.hs generation within Setup with some simple CPP.
> 
> Sat Jan 24 22:52:00 CET 2009  Petr Rockai <me at mornfall.net>
>   * Produce -DPACKAGE_VERSION="..." programatically in Setup.lhs.
> 
> Sat Jan 24 23:00:19 CET 2009  Petr Rockai <me at mornfall.net>
>   * Drop autogeneration of Autoconf.hs, use CPP instead.

And I'm going to wait a while on the rest of these.  Trent, perhaps you
could look into the autoconf issue?

Thanks!

Drop autogeneration of Autoconf.hs, use CPP instead.
----------------------------------------------------
> -      subst "DIFF"          = "diff"
> -      subst "BIGENDIAN"     = show bigendian

I didn't spot any mechanism to define BIGENDIAN as needed... could you
point it out to me?

-- 
Eric Kow <http://www.nltg.brighton.ac.uk/home/Eric.Kow>
PGP Key ID: 08AC04F9
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 194 bytes
Desc: not available
Url : http://lists.osuosl.org/pipermail/darcs-users/attachments/20090125/c6d87c57/attachment.pgp 


More information about the darcs-users mailing list