El 23/6/25 a las 23:25, Karl Berry escribió:
jcb> Could the right answer be to limit the effect of SOURCE_DATE_EPOCH to timestamps that will appear in built files?I'm afraid I'm not sure what you mean, Jacob. At any rate, before flailing around further, I'd like to understand why the Debian build had problems with the original mdate-sh that did nothing with SOURCE_DATE_EPOCH. Santiago? Please? This is following up on the change I made to mdate-sh following your (indirect) report at https://bugs.gnu.org/77805.
The Debian build of m4 version 1.4.19 dropped UPDATED from doc/m4.texi because it was using autoreconf, which regenerated a lot of things that were already generated by the authors when creating the tarball. This is recommended by some people in Debian but I'm skeptical that it's a good thing, and I've decided not to do that unless there is a real need (for example, when not doing so makes the package not to build in some architecture). Version 1.4.20 of m4 in Debian will not use autoreconf, and I think it will be reproducible because I've already tested it. I'm really sorry that my silence in the last threads may have caused any trouble by implementing things that were not really requested (at least by me). Thanks.
