Bug#818128: nethack: please make the build reproducible

[James Cowgill]

Hi,

On Mon, 2016-03-14 at 00:37 +0100, Reiner Herrmann wrote:
> While working on the "reproducible builds" effort [1], we have noticed
> that nethack could not be built reproducibly.

Thanks for the patch. It'll be applied at some point.

> There are three issues that prevent building it reproducibly:
> - the nhdat file is not reproducible, because the order of archive
>   members varies depending on the locale. Setting LC_ALL to C expands
>   the files in DATDLB always in the same way.

OK that seems reasonable.

> - makedefs uses the current date/time for the build date that gets
>   embedded. By using $SOURCE_DATE_EPOCH, the date from the latest
>   changelog entry is used instead.

OK, but should this be  tm = gmtime((time_t *) &clocktim);  ?
ie: drop the call to asctime.

> +@@ -1281,7 +1284,7 @@
> +     Fprintf(ofp, "\n");
> + #ifdef AMIGA
> +     {
> +-        struct tm *tm = localtime((time_t *) &clocktim);
> ++        struct tm *tm = asctime(gmtime((time_t *) &clocktim));
> +         Fprintf(ofp, "#define AMIGA_VERSION_STRING ");
> +         Fprintf(ofp, "\"\\0$VER: NetHack %d.%d.%d (%d.%d.%d)\"\n",
> +                 VERSION_MAJOR, VERSION_MINOR, PATCHLEVEL, tm->tm_mday,


> - parallel building causes that src/objects.o and src/monst.o are build
>   either from within src/ or from util/ (as those files are used by
>   nethack and by makedefs), which results in a different build path
>   embedded into the binary. This is solved by building makedefs
>   explicitely before nethack.

I have a feeling this caused the strange "random" build failure on
amd64. Hopefully I can get upstream to fix it properly at some point.

James
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: This is a digitally signed message part
URL: <http://lists.alioth.debian.org/pipermail/pkg-games-devel/attachments/20160314/54fe52d7/attachment.sig>