← Back to team overview

tortoisebzr-developers team mailing list archive

QBzr hybrid commands landed.

 

Hi all,
  I've landed support for the 'hybrid commands' I've been rattling on about
recently.  I've implemented 'qgetnew' (1 dialog) and 'qgetupdates' (2
dialogs depending on if a checkout or branch) commands, and landed a few
other cleanups.   A few UI cleanups remain and although everything works
AFAIK, but there are a few things that could do with polish.

Of note:

* new commands 'qgetnew' and 'qgetupdates' to implement the 'hybrid'
commands I've discussed previously with their dialogs.  The names of the
commands are obviously still up for discussions.  'qgetnew' has a whole
bunch of html for tooltip text added by QtDesigner that I fear will be a
problem for l10n, but we can deal with that as we go forward.

* consolidated the 'directory picker' and code for filling the combo boxes
into utils.py, resulting in chunks of duplicated code being removed and more
iterator/generator niceness/confusing-ness, depending on your perspective :)

* Part of the combo consolidation is support for saving 'pull' locations
globally for the user (ie, in qbzr.conf) - but at this stage, its only
support - the values aren't actually read or written.  Instead, a dummy
value of 'http://pretend/this/was/a/saved/location' is used to indicate
where the saved locations *would* be.  I'll either implement this or remove
the dummy value soon, but in the meantime, see the comments near the end of
utils.py for more details, and let me know if you think I'm on the right
track.

* The 'qgetnew' command dynamically updates the 'to location' combo
on-the-fly as the user types in the 'from location' box.  This mirrors tsvn,
and prevents the issue that Alex mentioned in a previous post, where the
'branch' dialog's "to location" was often incorrectly blank - if no one
complains, the 'branch' dialog should get the same behaviour (and the 'to
location' should not be a combo box in the 'branch' dialog either.)

* subprocess dialog allows the "default size" for a form to be None, thereby
using the default values specified in QtDesigner.

I hope I haven't broken anything (I did a 'checkin' to the trunk this time -
revision 409 ;) - but please let me know if I have, or if there are any
other concerns.  My tbzr branch has full support for these commands, incase
anyone is following along from home...

Thanks,

Mark






Follow ups