Search Criteria
Package Details: recutils 1.9-2
Package Actions
Git Clone URL: | https://aur.archlinux.org/recutils.git (read-only, click to copy) |
---|---|
Package Base: | recutils |
Description: | Set of tools and libraries to access plain text databases called recfiles. |
Upstream URL: | https://www.gnu.org/software/recutils/ |
Licenses: | GPL |
Submitter: | fgrsnau |
Maintainer: | fgrsnau |
Last Packager: | fgrsnau |
Votes: | 18 |
Popularity: | 0.002368 |
First Submitted: | 2015-12-05 13:03 (UTC) |
Last Updated: | 2024-06-09 13:47 (UTC) |
Required by (1)
- gnunet-git (make)
Latest Comments
« First ‹ Previous 1 2 3 Next › Last »
matthias.lisin commented on 2020-10-20 22:13 (UTC)
@fgrsnau: How about 1.8.90 alpha meanwhile? It includes the fix according to a report on the ML (that I can't find right now)
http://alpha.gnu.org/gnu/recutils/
fgrsnau commented on 2020-10-20 22:07 (UTC) (edited on 2020-10-20 22:08 (UTC) by fgrsnau)
Hi exclumine and OJFord!
I did have a look and it seems that incorporating the changes is currently not easily done (other opinions welcome though!)
The python subdirectory is missing from the upstream source tarball (at least for the latest stable release). While we could fetch these files or include them I think it makes more sense to wait for them to appear in an official release, because this is not a *-git package.
Regarding the renaming/tmpfs issuse: The patch unfortunately does not apply nicely. If you look at the patch it is also modifying bootstrap.conf which is (i) not part of the source distribution, (ii) used by gnulib to copy a "source code library" into the source distribution and doing proper modifications of the whole build system (namely lib/Makefile.am which is then mangled by automake). I currently hesitate to patch these automatically generated autotools/automake things.
Depending on the release date of 1.9 it should be easier to wait for the change to appear there. (A workaround right now is to set the TMPDIR environment variable to a directory on the same filesystem.)
As a side note: The fix seems to unconditional introduce a copy+unlink step which is not an atomic operation. If a problem occurs during the copy (overwrite) the database will probably become corrupted.
OJFord commented on 2020-10-17 20:36 (UTC)
I'm not sure what's most appropriate here, but could you include the python lib? There is no separate
python-rec
currently, and since it's in the same repo maybe it makes sense to provide in this package too? (That is what packages such asgit
do with their 'contrib' dirs, for example.)exlumine commented on 2020-10-07 16:03 (UTC)
Hi fgrsnau, Could you include the upstream commits which fix a /tmp/rec* renaming error? Currently selected utils fail as they cannot handle files in other filesystems well. It seems that jemarch fixed that in https://git.savannah.gnu.org/cgit/recutils.git/commit/?id=86f662a8202408134a235572ec60141d3082f975. Thx and best exlumine
chikker commented on 2020-03-16 20:16 (UTC)
Hey there. I've updated PKGBUILD to verify source code using PGP signatures and sha256 hashes instead of relying on broken md5 algorithm.
fgrsnau commented on 2020-03-12 21:39 (UTC) (edited on 2020-03-12 21:39 (UTC) by fgrsnau)
I updated the package according to your request. As I do not use bash as an interactive shell, hence I did not notice any problems. Hope it works now, otherwise I can have a look again :)
matthias.lisin commented on 2020-02-23 15:17 (UTC)
Hey fgrsnau, please remove
--with-bash-headers
frombuild()
. It seems to set an empty path and prevents bash builtin compilation. The bash headers are detected automatically. Also add package tocheckdepends=(check)
for additional unittests.Diff:
fgrsnau commented on 2019-11-10 19:30 (UTC)
Sorry for the delay. I simply forgot that I wanted to bump the version.
gour commented on 2019-01-04 12:46 (UTC)
1.8 builds cleanly here. ;)
fgrsnau commented on 2019-01-04 08:51 (UTC) (edited on 2019-01-04 08:53 (UTC) by fgrsnau)
Thanks, I will have a look and also update this to v1.8 which was released yesterday.
edit: This seems to be an upstream bug, see <https://bbs.archlinux.org/viewtopic.php?id=240499>.
« First ‹ Previous 1 2 3 Next › Last »