summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
...
* Vlan support: GUI - add vlan device (glade) (#906272)Radek Vykydal2013-03-131-0/+4
|
* Show that password was set by kickstart in TUIMartin Sivak2013-03-121-1/+4
|
* Add settable quit message to TUIMartin Sivak2013-03-122-4/+30
|
* Check _current_action for not being None before using itVratislav Podzimek2013-03-121-1/+1
| | | | | It may happen that _current_action is None even when self._actions is not None and not empty.
* New version.Brian C. Lane2013-03-113-148/+199
|
* Bonding support: GUI - hub status information (#906263)Radek Vykydal2013-03-111-22/+39
|
* Bonding support: GUI - generate kickstart network command for bonds (#906263)Radek Vykydal2013-03-112-10/+77
|
* Bonding support: GUI - add device dialog (#906263)Radek Vykydal2013-03-111-4/+12
|
* Bonding support: GUI - add device dialog (glade) (#906263)Radek Vykydal2013-03-111-0/+114
|
* Bonding support: GUI - device list, configuration and adding bond (#906263)Radek Vykydal2013-03-111-6/+57
|
* Bonding support: GUI - Slaves line in Wired tab (glade) (#906263)Radek Vykydal2013-03-111-14/+44
|
* Crypt the root we get from the user (#918991).Jesse Keating2013-03-081-3/+3
| | | | | | | No sense having it hang around in the kickstart object in plain text. Also add my name to the authors as now I've got a bunch of lines in this file too.
* Adapt to the new libxklavier's behaviourVratislav Podzimek2013-03-081-2/+3
| | | | | libxklavier now returns lists of ordinal numbers of chars instead of null-byte terminated strings.
* We use python-meh's interfaces instead of Anaconda'sVratislav Podzimek2013-03-081-1/+1
| | | | | There is no showError method in python-meh's interfaces, they have messageWindow instead.
* Redraw screen in case of valid input and nothing new scheduledVratislav Podzimek2013-03-081-0/+4
| | | | | If the screen successfully processes the input and doesn't schedule another screen, it should be redrawn, so that the new state is shown.
* Fix two places where we are locking up the main thread (#886680).Chris Lumens2013-03-072-10/+21
| | | | | | | | | | | | (1) Letting _first_refresh be the first time we reference payload.groups and payload.environments means we are potentially downloading lots of big data in the main thread, which means the UI cannot update. (2) Referencing anything in software.py's _initialize grabs the yum lock, possibly for a long time. When we later check the source spoke's status, we reference ready which checks baseRepo, which also attempts to grab the same lock. This isn't a deadlock because it will eventually get released, but it does make the UI unresponsive for a while.
* Log when we acquire and release the _yum_lock (dlehman, clumens).Chris Lumens2013-03-071-2/+18
| | | | | | | | We've got a temporary locking issue somewhere that causes the UI to become unresponsive for a bit (see #886680, perhaps other bugs). Examining the code indicates this isn't necessarily due to running things in the main loop that shouldn't be. There's something waiting for the _yum_lock. This patch adds logging to help debug.
* Do not fail when the logging stream cannot be opened (in initial-setup..)Martin Sivak2013-03-071-9/+12
|
* Mark DateTime spoke as usable for FirstbootMartin Sivak2013-03-071-1/+2
|
* Fix a missing import and move the addon KS output templateMartin Sivak2013-03-071-2/+2
|
* Make the TUI mainloop more resistant to screen implementation errorsMartin Sivak2013-03-071-17/+45
|
* Teach TUI how to react on async eventsMartin Sivak2013-03-075-11/+128
| | | | | | | | | | | | | This involves couple of places: - TUI is blocking and waiting for user input - Progress HUB is in charge and is reading the progress queue - any other part wanting to react to a message like READY, .. This patch directly fixes the first two and adds a mechanism to register callbacks that will react to the third one. We (me and vpodzime) expect python-meh integration to use this to file exception reports from text mode installation.
* Could not load UI file advanced_user.gladeMark Hamzy2013-03-061-1/+1
| | | | | | Fixes IOError: Could not load UI file 'advanced_user.glade' for object Signed-off-by: Martin Sivak <msivak@redhat.com>
* Don't try to remove the timer when it's None (DatetimeSpoke)Vratislav Podzimek2013-03-051-1/+2
|
* Log failed imports in the collect functionsVratislav Podzimek2013-03-051-1/+5
| | | | | Otherwise there is no other way to find out what happened other than rebooting and using pdb.
* remove the remnants of sparc supportDennis Gilmore2013-03-044-90/+5
| | | | Signed-off-by: Dennis Gilmore <dennis@ausil.us>
* Remove installmethod.py (dead code)Will Woods2013-03-011-46/+0
| | | | | | | | The only thing here was one completely unused function that tried to eject the media at the end of the install. We've got better ways to handle eject now, so this code is useless. Remove it.
* Make default media eject behavior match old behaviorWill Woods2013-03-012-1/+5
| | | | | | | | | | | | | | | | Previous versions of anaconda ejected the media by default for interactive installs. After commit a724f80 (May 2007), kickstart installs needed to pass "reboot --eject" to make the drive eject. (See bug #238711, #239002) Commit 2f601d2 (Sep 2010) added the "noeject" boot argument, which overrides both the default and the kickstart command. (See bug #477887) This patch should restore this behavior: a) Eject is enabled by default, but b) disabled for automatedInstall unless "reboot --eject" is used, and c) disabled if "noeject" is passed as a boot argument.
* Silence "cp: cannot stat '/etc/cmdline'..." error messageWill Woods2013-03-011-2/+4
| | | | | These files may or may not exist; check to see if they exist before copying them!
* New version.Brian C. Lane2013-03-013-117/+309
|
* Behave nice when root password is set by kickstartHEADmasterMartin Sivak2013-03-011-4/+21
|
* Password spoke is mandatory if the created user is not an adminMartin Sivak2013-03-011-1/+2
|
* Use the user data provided by kickstartMartin Sivak2013-03-011-7/+46
|
* Add the User creation spoke including the Advanced dialogMartin Sivak2013-03-015-1/+1238
|
* Bonding support: kickstartRadek Vykydal2013-03-012-14/+45
| | | | | | | TODO: - sync dracut and NM naming convention for slaves? perhaps better: don't write any ifcfg files for kickstart case - test with new NM (dhcp on slaves)
* Condense some duplicated and overly wordy code in custom.py.Chris Lumens2013-02-281-39/+39
|
* Add a new allMembers property that returns a list of pages and members.Chris Lumens2013-02-281-0/+6
|
* All Pages have a title, so get rid of the getattr games.Chris Lumens2013-02-281-2/+2
|
* Allow more than one Page to be expanded at a time.Chris Lumens2013-02-282-35/+5
| | | | | | | | | This should help a little bit with the confusion people have where it looks like filesystems are vanishing, when instead they are just moving to under a different root. However, note that anything calling _do_refresh will cause everything to be closed except the current page. I do not yet see a way around that.
* Get rid of the currentPage method.Chris Lumens2013-02-282-13/+22
| | | | | | When multiple pages can be expanded at the same time, the currentPage method doesn't make a lot of sense. Instead, make it a property of the custom spoke and have it depend on the current selector.
* Promote page._members to page.members.Chris Lumens2013-02-282-18/+16
| | | | | | We're using it all over the place anyway, so there's no point keeping the underscore on it. Also I gave CreateNewPage a members attr as well, so there's no need to test for it anymore.
* Require passing the title to a Page's constructor.Chris Lumens2013-02-282-14/+12
|
* Pressing F12 should do the same thing as clicking "Done" (#840998).Chris Lumens2013-02-282-2/+36
| | | | You could do this in the old UI, so let's bring it forward to the new UI.
* A bunch more "install" -> "installation" changes.Chris Lumens2013-02-286-7/+7
|
* When the user clicks "Reclaim Space", go back to the hub (#911792).Chris Lumens2013-02-281-7/+12
| | | | | | | | Going back to the options dialogs is arguably not really a bug (the user may want to go to custom partitioning), but they have other ways of doing those things. Also, watching people test this out showed it to be very confusing and led to people going down paths they didn't need to. Also, it's the way it worked in F17.
* Modify the logic that makes the reclaim button sensitive (#911793).Chris Lumens2013-02-281-1/+8
| | | | | | | | | | | | This dialog now serves two different purposes: (1) You do not have enough free space, and need to make some before you can continue to install. (2) You do have enough free space but want to free up additional space. Because the button is labeled "Reclaim space", we only want to allow clicking on it if you have chosen some destructive action. Buttons should be labeled with what they do, after all. However, we also want to take into account any free space that may already exist on disk when considering whether you have enough to continue.
* Add a free space line under every disk in the reclaim dialog.Chris Lumens2013-02-282-1/+17
| | | | | | This is to help communicate how much total free space there is, in the case where you've got enough but choose to go into the reclaim dialog to do some other things anyway.
* Remove the initial sentence from the top of the reclaim dialog (#911793).Chris Lumens2013-02-281-2/+1
| | | | | If you go into the reclaim dialog already with enough space, it's a very confusing thing to be told you do not have enough free space.
* Fix a traceback in verifying optical media on the source spoke.Chris Lumens2013-02-281-2/+2
| | | | | This was caused by 1821378de1aecbbf5806edcad34ee928f0892296, which changed what was being stored in the DiskOverviews.
* Revert "Hook up the "Remove Packages" button on the dep solving error ↵Chris Lumens2013-02-283-28/+2
| | | | | | | | | | | | screen." (#905899). This reverts commit f4ec3d682ffd93dfc7105eaa09acdd7fd672a3e8. This patch also removes the button entirely. All I've ever seen it do is reduce the installable package set to something like 30 packages and result in really cryptic bugs about commands that should be there failing to run. It doesn't seem to help in just removing the one or two packages that have a problem, since it looks like it's frequently a core package with the problem.
generated by cgit (git 2.34.1) at 2025-09-26 04:33:41 +0000