-
-
Notifications
You must be signed in to change notification settings - Fork 480
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Update cddlib's autotooling #21952
Comments
Attachment: cddlib-0.94.tar.gz Attachment: 0001-Removed-generated-files-and-other-junk-that-should-n.patch.gz |
Attachment: 0005-Ran.patch.gz |
Attachment: 0006-Ran.patch.gz Attachment: 0007-Add-to-all-paths-in-EXTRA_DIST-so-that-VPATH-builds-.patch.gz |
comment:1
Do you have a git repository from which you created this? |
comment:2
I wouldn't have thought that There are other things I am thinking off that can be done (VPATH) but we may not want to do everything at once. |
comment:3
Replying to @kiwifb:
I'm fairly sure myself that it isn't strictly necessary, but I put it because autoconf suggested that I do anyways. |
comment:4
Or maybe it was libtoolize that told me that. I can't remember for sure. From https://www.gnu.org/software/libtool/manual/html_node/Invoking-libtoolize.html "In the future other Autotools will automatically check the contents of AC_CONFIG_MACRO_DIRS, but at the moment it is more portable to add the macro directory to ACLOCAL_AMFLAGS in Makefile.am, which is where the tools currently look." That future is now, but I guess it still suggests it just in case. |
comment:7
I think this is actually ready to go. I just don't know what the process is for adding a new tarball to the sage package mirrors. For this particular case I don't see a better workaround than a new source package since the original is a bit broken. |
comment:8
There's no branch, what is to be reviewed? Also, could you just share the git from which these patches are generated? (comment 1) Also, "cddlib-0.94.tar.gz" does not seem to have an unambiguous version number. |
comment:10
What I don't understand is what the process is (though it seems like there is one?) of providing an alternative upstream tarball for one of the packages in Sage, such that is distributed to all the mirrors, etc. I've attached such a tarball, as well as the patches that reproduce it from the original upstream tarball. The problem here was that the source tarball itself was broken in various ways. |
comment:11
Most of the process is described here: http://doc.sagemath.org/html/en/developer/packaging.html#when-to-patch-when-to-repackage-when-to-autoconfiscate (and in the following sections). The release manager adds the "upstream tar ball" to the server when the ticket is merged. Usually people just provide an URL in the description, rather than attaching it to trac. Instead of attaching patches here, use a public git repository. Make an |
comment:12
Forgot about this until now, but thanks for the info. I'll follow the process you described then. |
comment:13
embray: are you planning to work on this again anytime soon? I just stumbled upon this because Debian has some patches to fix Sage's integration of cddlib for version 0.94h. If you think that the patches proposed here are still the way this should be done, should we propose them upstream? |
comment:14
Just a note, from my experience in sage-on-gentoo, the debian patches are insufficient - as in I have doctest failures even with the debian patches. Otherwise I am all for pushing updated autotooling upstream. |
comment:15
This really needs #21196. IIRC I ran into some stumbling blocks with that, but I forget--it was a long time ago. |
Dependencies: #21196 |
comment:16
So, that's a dependency if I understood you correctly. |
comment:17
I'm dropping that dependency again, as I am in favor of dropping autotools from our SPKGs. |
Changed dependencies from #21196 to none |
comment:18
I wrote a mail to upstream, offering my help with this. Let's see what they say. |
Changed upstream from Not yet reported upstream; Will do shortly. to Reported upstream. No feedback yet. |
comment:19
I think the series of patches I originally attached to this ticket should do it. I forget why I just added a bunch of patches here as opposed to some other way of doing it... |
comment:20
Some of the patches don't apply anymore so I rebase them. I am trying to convince the author of the beauty of github (or similar) and he wants to have a look. I proposed a new home for upstream here: https://github.com/cddlib/cddlib. And a modified version of your patches is here: cddlib/cddlib#1 Let's continue the discussion there. |
Changed upstream from Reported upstream. No feedback yet. to Reported upstream. Developers acknowledge bug. |
comment:23
Upstream has upgraded their autotooling. Now we just need to package the next release #25344. |
comment:24
(feel free to reopen this, if you think that I am missing something here.) |
comment:25
closing positively reviewed duplicates |
As discussed in this thread, cddlib's autotools files are quite outdated and in due for modernization. In fact, somehow, its
configure.in
is not even compatible with the version ofautomake
that theMakefile.in
s were generated with. This makes it difficult/impossible to make update cddlib's Makefiles (see #15872).I reworked the latest version of cddlib (0.94h, though this work would also apply to at least the previous version, 0.94g) to use automake 1.14.1 (currently the highest version in sage, though #21196 would update that to 1.15).
I also cleaned up the contents of the tarball (removed .DS_Store files and other crud), and made it so that
make distcheck
works.I've attached an updated upstream tarball, as well as a set of patches to the original upstream that should reproduce it.
Upstream: Reported upstream. Developers acknowledge bug.
Component: packages: standard
Keywords: cddlib autotools
Author: Erik Bray
Issue created by migration from https://trac.sagemath.org/ticket/21952
The text was updated successfully, but these errors were encountered: