Hi, This week I have continued working on the reproducible timestamp approach through the exported variable SOURCE_DATE_EPOCH.
I submitted a bug for debhelper with last week's patch (I had already commited a fix exporting SOURCE_DATE_EPOCH in our local git repo, and uploaded the package in our local repo): http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=791823 I submitted bug to Debian for libxslt to honour SOURCE_DATE_EPOCH: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=791815 I also commited the fix to our git repo and puloaded the package our repo to perform tests: https://anonscm.debian.org/cgit/reproducible/libxslt.git/commit/?h=pu/reproducible_builds&id=571bb438df0b11288598c2b05010099d1861a8ed After the upload, test were performed on the packages affected by issues: 1- timestamps_generated_by_docbook_to_man (here some packages where miss tagged) - there where 79 unreproducible packages and 39 became reproducible 2- timestamps_in_manpages_generated_by_docbook_xsl - there where 89 unreproducible packages and 33 became reproducible. NOTE: there are still packages generating docs with libslt not reproducible. This is because the patch for dh to export SOURCE_DATE_EPOCH only works for calls to doc generation under dh. Some packages have the calls to generate docs directly in debian/rules so SOURCE_DATE_EPOCH is not set there. A solution needs to be found for that. I updated some info in the wiki/ExperimentalToolchain regarding packages using the SOURCE_DATE_EPOCH proposal. I tagged some packages affected by locale_in_documentation_generated_by_javadoc. I started working on the issue lc_messages_randomness: https://reproducible.debian.net/issues/unstable/lc_messages_randomness_issue.html I'll continue working on that next week. Best regards, Dhole
Description: OpenPGP digital signature
_______________________________________________ Reproducible-builds mailing list Reproduciblefirstname.lastname@example.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds