GitLab

#358

Ville Skyttä authored May 17, 2012 and David King committed Sep 18, 2021

#230

scootergrisen authored Feb 06, 2020 and David King committed Sep 17, 2021

!5

Ting-Wei Lan authored Jun 06, 2015 and David King committed Sep 17, 2021

#287
https://bugzilla.gnome.org/show_bug.cgi?id=750488

Bastien Nocera authored Sep 09, 2021 and David King committed Sep 16, 2021

The BD-R flags were simply never applied to the plugin, so the brasero
core didn't know how to burn that type of media.

See https://bugzilla.redhat.com/show_bug.cgi?id=1704341
and https://bugzilla.redhat.com/show_bug.cgi?id=1456971

Closes: #324

Bastien Nocera authored Sep 09, 2021 and David King committed Sep 16, 2021

The macro was never used and was missing an opening parenthesis.

Bastien Nocera authored Sep 09, 2021 and David King committed Sep 16, 2021

Print some debug when flags don't match.

Bastien Nocera authored Sep 09, 2021 and David King committed Sep 16, 2021

libbrasero-media/scsi-read-track-information.c: In function ‘brasero_read_track_info’:
libbrasero-media/scsi-read-track-information.c:116:33: warning: argument to ‘sizeof’ in ‘memset’ call is the same pointer type ‘BraseroScsiTrackInfo *’ {aka ‘struct _BraseroScsiTrackInfo *’} as the destination; expected ‘BraseroScsiTrackInfo’ {aka ‘struct _BraseroScsiTrackInfo’} or an explicit length [-Wsizeof-pointer-memaccess]
  116 |         memset (&hdr, 0, sizeof (info));
      |                                 ^

Robert McQueen authored Feb 19, 2020 and David King committed Sep 16, 2021

Make the gstreamer plugins tested for by the transcode plugin match the
plugins it uses.

Fixes: #325

Carlos Garnacho authored Dec 24, 2020 and David King committed Sep 16, 2021

Add the compile-time checks, and add the minimal code catering for
the API changes. In 3.0 connections are no longer singletons.

Carlos Garnacho authored Dec 24, 2020 and David King committed Sep 16, 2021

These resources are interpreted differently in Tracker 3.0 compared
to older versions, the newer version makes those 2 different resources
(one representing the "file", other the "content) that relate to each
other. In older tracker versions, these 2 aspects were conflated in
a single resource.

The query is implicitly crossing that barrier, so make it explicit.
With this change, the query may run with either version.

Carlos Garnacho authored Dec 24, 2020 and David King committed Sep 16, 2021

This fixes two things:
- The use of fts:match depends on priv->keywords being set, but
  fts:rank() has undefined behavior if fts:match is not present.
  Its value is returned in the cursor, but fortunately unused,
  so it can go away.

- Fixes "garbage" being added at the end of the query, concretely
  ") } ORDER BY ... " is appended at the supposed end of the query
  string. Older SPARQL parser implementations would allow garbage
  at the end of its parsing state, I suspect "LIMIT 0 OFFSET 10000"
  was just a mean to push it to that state that it'd allow the
  remaining string, so that is gone too.

Carlos Garnacho authored Dec 24, 2020 and David King committed Sep 16, 2021

The ?urn variable is not defined anywhere else in the query, so
ordering is left undefined. Access the url/filename properly here.

Carlos Garnacho authored Dec 24, 2020 and David King committed Sep 16, 2021

The is_available vmethod tries to create a connection if it's not
there. But the BraseroSearchTracker initialization already tries to
obtain the SPARQL connection. Let this method return wether
initialiation was successful.

Which means that gtk-doc can now list the object types and generate a
tree_index.sgml file.