--nextPart3025226.he0EbxOkNQ Content-Type: text/plain; charset="iso-8859-6" Content-Transfer-Encoding: quoted-printable Content-Disposition: inline On Monday 17 October 2005 12:54, Benjamin Meyer wrote: > On Sunday 16 October 2005 7:07 pm, David Jarvie wrote: > > Here is my proposed new code for the kdelibs/kdecore time zone classes. > > I've attached new code rather than diffs since the changes are very > > substantial (although some parts of the existing code is still > > retained). The aim of the changes is to provide base classes to use for > > deriving non-system time zone classes. In particular, this will form a > > basis for implementing time zone classes for iCalendar time zone data, > > which should allow some long standing bugs in kdepim to be fixed. The > > current time zone implementation is only really suited to dealing with > > system time zones, not time zone information from other sources. > > > > See the discussion on 29th and 30th September in this mailing list > > (http://lists.kde.org/?l=3Dkde-core-devel&m=3D112800361918888&w=3D2 etc= =2E) for > > more details. > > > > Note that I haven't yet tested the KTzfile* classes - I'm in the > > process of doing that now. The system time zone handling doesn't depend > > on them, so they aren't so important. > > > > Can I commit? > > Docs and tests! Yah :) Also note, that he uses @defgroup and @ingroup for grouping. I think this i= s=20 a very good approach and we should use it all over KDE (where it makes=20 sense). Groups coming into my mind for kdecore are fox example: - Core Classes KApplication etc - DCOP (DBUS) Architecture - Localization KTimezone, KLocale, etc. - Calendar System Contains *Calendar* classes - Configuration Management KConfigBase etc. maybe even subgroups for KGlobal*, KStandardDirs - Completion System KCompletion etc. - Launching Processes KProcess, KShellProcess etc. - NET Window Manager Specification NET*, KWin, etc. - Loading Libraries at runtime KLibLoader etc. - Network Support KSocket etc. - Icons and Themes KIcon* This list was created by just looking over what classes exist. I'm sure it'= s=20 possible to bring it in a good shape ;) At some points subgroups make=20 sense. cu, =2D-=20 Dominik --nextPart3025226.he0EbxOkNQ Content-Type: application/pgp-signature -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.9.15 (GNU/Linux) iD8DBQBDU2snKaoWlv9k7bwRAh9jAJ9yZRt8E3Lay/M1KrxpGjCrRv2vkQCfQg0u 91H2J0kuEmAor26y01nE/dc= =3Rcx -----END PGP SIGNATURE----- --nextPart3025226.he0EbxOkNQ--