2013-05-01 Martyn Russell Release 0.16.1 2013-05-01 Antoine Jacoutot Implement get_memory_total() on OpenBSD. https://bugzilla.gnome.org/show_bug.cgi?id=697719 tracker-dbus: add support for OpenBSD Add OS dependant glue for tracker-dbus on OpenBSD using kvm(3). 2013-05-01 Dominique Leuenberger firefox/thunderbird: prefer tracker 0.16 over tracker 0.14 https://bugzilla.gnome.org/show_bug.cgi?id=697809 2013-05-01 Antoine Jacoutot SCHED_IDLE is only available on Linux. https://bugzilla.gnome.org/show_bug.cgi?id=697728 Another need for XOPEN_SOURCE_EXTENDED missed in previous commit. Was missed in 3443706bbe0c1c2e465336c59f87cec90929ff1e OpenBSD needs _XOPEN_SOURCE_EXTENDED for strptime/strcasecmp These were introduced in XPG4. On OpenBSD, just defining _XOPEN_SOURCE gets you the older XPG3 wich doesn't include them, so we need to define _XOPEN_SOURCE_EXTENDED to get XPG4v2. This fixes several implicit declarations. https://bugzilla.gnome.org/show_bug.cgi?id=697688 2013-05-01 Martyn Russell libtracker-data: Fix unit tests for localtime/timezone functions The insert order != query order and the query has no ORDER BY. So it's not guaranteed anyway, but the diff expects an order. So this patch adds ORDER BY to ensure the diff doesn't fail due to ordering but only lack of content expected. 2013-04-30 Martyn Russell libtracker-sparql-backend: Fix connection singleton mutex issue Since moving from GStaticMutex to GMutex (due to deprecations in GLib), there has been an issue noticed with the way the mutex is created. Before, the GStaticMutex was created in the first instance from any API call and used there on after. This meant the chance of it not being initialised when we used it was almost 0. Now with GMutex, because we've initialised it differently (in the class constructor) with Mutex() this means that g_mutex_init(&_tmp0_) is the generated source from that. Sadly, the first instance of g_mutex_lock() will create the mutex for us if it is NULL or {0} anyway and this happens _BEFORE_ the g_mutex_init() call. So we were locking, initing and the unlocking. This is the cause of the problem. Now we don't initialise the mutex at all and let that initialise upon first lock. Note, atomic operations in GLib should protect race conditions affecting creation of the singleton here. https://bugzilla.gnome.org/show_bug.cgi?id=697316 2013-04-29 Andika Triwidada Updated Indonesian translation 2013-04-28 Žygimantas Beručka Updated Lithuanian translation 2013-04-25 Daniel Mustieles Updated Spanish translation 2013-04-18 Alexandre Franke Update French translation 2013-04-11 Мирослав Николић Updated Serbian translation 2013-04-10 Rafael Ferreira Updated Brazilian Portuguese translation 2013-04-08 Carlos Garnacho Fix AS ?foo handling in FTS queries FTS queries implicitly add an "AS var" clause to the translated SQL select query so values can be matched with the outer query that accesses FTS tables, which resulted in doubly added AS clauses if it was specified explicitly in SPARQL too. So, make sure the clause is just added once. 2013-04-01 Aurimas Černius Updated Lithuanian translation 2013-03-30 Matej Urbančič Updated Slovenian translation Updated Slovenian translation 2013-03-27 Marek Černocký Updated Czech translation 2013-03-25 Piotr Drąg Updated Polish translation 2013-03-25 Martyn Russell tracker-tag: Fix --list (-t) to use logical OR, not AND Also improve the documentation. tracker-tag: Fixed GLib-WARNING **: goption.c:2193: ignoring no-arg... Was using wrong FLAG in GOption