GNOME Bugzilla – Bug 437584
Compilation warning in evolution's plugins folder
Last modified: 2013-09-13 00:54:06 UTC
Please describe the problem: to follow, a patch to fix some compilations warnings (not all) in evolution/plugins Steps to reproduce: Actual results: Expected results: Does this happen every time? Other information:
Created attachment 87986 [details] [review] plugins-compilation-warning.patch Here is the patch (with ChangeLog). Note that it doesn't fix all warnings but it fixes a good number of the most verbose warning messages which helps in finding really annoying warnings.
exchange-operations/exchange-folder.c code-change needs rework. Use "folder-no-gc-error" and "no-user-error" error messages that are defined in the errors.xml Rest of the changes looks good. Commit to both TRUNK and stable branch.
Created attachment 87998 [details] [review] plugins-compilation-warning.patch ok, should be fixed wrt comment #2
Seems like you hand-edited the patch. It was not applying. In future. check if your patches are applying properly :) Modified the patch and committed in trunk and branch
Created attachment 88060 [details] [review] calendar-weather-warnings.patch this patch fixes warnings from calendar-weather that were left behind. I now have setup a stable branch to work on so it'll be easier/faster for me to provide patch against stable. Thanks for your time. Also I've found that some fixes that are in svn head are not in stable resulting in a big difference of volume of compilation warnings. Should I produce patches for stable branche specifically as well or just backporting those from head will do ?
reopening since there is one patch left.
(In reply to comment #5) > Created an attachment (id=88060) [edit] > calendar-weather-warnings.patch > > this patch fixes warnings from calendar-weather that were left behind. Patch committed. > > I now have setup a stable branch to work on so it'll be easier/faster for me to > provide patch against stable. Thanks for your time. > > Also I've found that some fixes that are in svn head are not in stable > resulting in a big difference of volume of compilation warnings. Should I > produce patches for stable branche specifically as well or just backporting > those from head will do ? Attach the patches in bugzilla, so that the maintainers will review/commit them. Do you see any such discrepancy in plugins ?
(In reply to comment #7) > Patch committed. thanks > Attach the patches in bugzilla, so that the maintainers will review/commit > them. Do you see any such discrepancy in plugins ? yes, actually this is where I found out. Example are bbdb and exchange plugin. Don't know how much it has changed and if it should be backported to stable anyway. Should we close this bug and discuss further on the mailing list ?
I just now compiled the plugins. For bbdb, in stable branch I see only one extra compilation warning than the trunk, which is due to a missing-cast. For exchange-operations I do not see any discrepancy at all. If you have any patches for plugins, just attach them in this bug so that we can review and take it in. For warnings in other areas, you can file individual bugs and use 437579 as the tracker.
Created attachment 88332 [details] [review] plugins-itip-formatter-warning.patch didn't looked much at diff stable/trunk yet but here is another patch for a warning in itip-formatter. Applys to trunk as well.
Looks fine to commit.
Committed to Subversion trunk (revision 33629). Leaving this bug open because there's still lots more warnings in the publish-location plugin.
Created attachment 89574 [details] [review] experimental-plugins.patch More cleanups in plugins. Please pay attention to the ipod-sync plugin. I've removed some stuff in the header that I thought was useless and was causing a warning because of glib-i18n / libintl overlap. Is there a better alternative to replace warn by g_warning in this case or was it just not justified to use a macro in the first place ?
Gilles, I merged the ipod plugin code from a bounty. I might have made sense as a separate project, but not as a plugin. I think it is fine, since the gi18n.h is included. Please commit to head.
Committed to head.