summaryrefslogtreecommitdiffstats
path: root/pyanaconda
Commit message (Collapse)AuthorAgeFilesLines
...
* Fix a traceback when clicking on the summary in custom spoke.David Lehman2012-08-201-4/+7
|
* Move device size calculation and setting into DeviceFactory.David Lehman2012-08-201-29/+39
|
* Stop pretending btrfs subvols can have a size.David Lehman2012-08-202-29/+1
|
* Fix a typo in StorageDevice._setSize.David Lehman2012-08-201-1/+1
|
* Don't fail on write of nonexisting IfcfgFile(SimpleConfigFile) (#849012, ↵Radek Vykydal2012-08-201-2/+5
| | | | | | | #849095) Due to commit 16a7c6e7af3df4816323a007f785155d307869bf, SimpleConfigFile.write of nonexisting files is failling.
* If dracut left the DVD mounted, don't try to remount it (#849152).Chris Lumens2012-08-201-7/+15
| | | | | Trying to remount was causing an exception which meant the DVD got thrown out as a possible installation source, so we were trying to use the network.
* Add support for most device editing functions.David Lehman2012-08-171-2/+127
| | | | | | | | | | | | | Changing label, fstype, size all mostly work. Changing device type between lvm and partition should mostly work. RAID options are completely untested, as is changing to a device type of RAID. Changing to and from a device type of BTRFS should probably work, but there's a SIGABRT from btrfs at filesystem/volume create time that will prevent you from actually using it.
* Various fixes, cleanups, and added logging for the custom spoke.David Lehman2012-08-171-20/+60
| | | | | | | | | | Fix size spinner values for new devices. Make sure the remove button is sensitive except on the "Create.." page. At the moment, selecting a page automatically selects one of the devices in that root, which means it's impossible to delete entire roots with the remove button.
* Work around some signal handling issues in the custom spoke.David Lehman2012-08-172-6/+14
| | | | | | | | For now, only connect the MountpointSelectors to button-press-event. Also, don't run the handlers for them during _do_refresh because, for some reason, they get triggered a whole bunch during setup of the left-hand side.
* Make choosing an auto-selected page after refresh slightly less fallible.David Lehman2012-08-172-12/+24
| | | | | | | Build a list, in decreasing order of preference, of pages that make sense to initially have selected. If a page was selected at the time of the refresh, it should probably be still selected afterwards. If that page is no longer present, no problem. Move on to the next one in the list.
* Raise an exception if a new device ends up with size 0.David Lehman2012-08-171-0/+7
| | | | It means the train is off the track.
* Split out logic to determine container based on factory and/or device.David Lehman2012-08-171-11/+15
|
* Allow adding disks to a container's disk set.David Lehman2012-08-172-21/+43
|
* Allow passing a device into newDevice for adjustment.David Lehman2012-08-171-5/+41
| | | | | This allows adjustment of a defined device instead of always having to create one from scratch.
* Add PartitionFactory class so partitions don't need a separate code path.David Lehman2012-08-172-53/+73
|
* Add a convenience method for scheduling resize actions.David Lehman2012-08-171-0/+18
|
* Return early from doKickstartStorage if there are no disks selected.David Lehman2012-08-171-0/+2
|
* Don't rely on having some network devices availableVratislav Podzimek2012-08-171-1/+3
| | | | | | | If there are no network devices available, itr returned from the devices_store.get_iter_first() is None and unchecked use raises an exception. This can happen e.g. when running 'make runhub' on a system with NetworkManager.service not running.
* Enlightbox mainExceptionWindowVratislav Podzimek2012-08-171-1/+9
|
* Put mainExceptionWindow in a WindowGroupVratislav Podzimek2012-08-171-1/+7
| | | | | Without WindowGroup, python-meh's window is insensitive if it appears above a spoke.
* Add a flag so we don't get spurious 'change' events from the treeview while ↵Bill Nottingham2012-08-171-0/+5
| | | | we're setting up the UI.
* Wire in the new environment logic through the UI.Bill Nottingham2012-08-172-56/+50
| | | | | | - s/desktop/environment/ - use the list of options for the environment, plus any user visible groups, to populate the add-on list once an environment is selected
* Add a local method for exposing group visibility from the comps file.Bill Nottingham2012-08-171-0/+12
|
* Add methods to yumpayload for handling environments.Bill Nottingham2012-08-172-0/+107
|
* Add some nicer wording to the column heads in the software selection UI.Bill Nottingham2012-08-171-2/+2
|
* Rename 'description' to 'groupDescription'.Bill Nottingham2012-08-172-2/+2
| | | | Other things may have descriptions later.
* Remove duplicate boot disk setting code (#848841).Chris Lumens2012-08-161-6/+0
| | | | | See 87b825c5ac14dfa16985f3115bec66c7ade68be1 for the GUI commit that does the same.
* Force authconfig to be installed on the target system (#848803).Chris Lumens2012-08-161-1/+6
|
* Fix attribution on common UI code.Chris Lumens2012-08-151-1/+2
|
* don't set armMachine in class definitionBrian C. Lane2012-08-151-1/+3
| | | | | armMachine can be set from the cmdline, via flags.armPlatform, so it needs to wait until instantiated before calling iutil.getARMMachine
* libudev now has a version of .1Mark Hamzy2012-08-151-2/+4
| | | | | libudev bumped the version number up to /usr/lib64/libudev.so.1 Also, change the error message from None.
* Enable text mode once again!Jesse Keating2012-08-141-5/+10
|
* Update text prompt to include c for continueJesse Keating2012-08-141-1/+1
| | | | Also changed the tab to two spaces to save on screen space
* Don't continue if incomplete spokes existJesse Keating2012-08-141-0/+7
| | | | | Not sure this belongs here, but it's where the text hub inherits the logic from.
* Return a bool for timezone completed propertyJesse Keating2012-08-141-1/+1
|
* Add a text progress hub to do the installJesse Keating2012-08-142-1/+117
| | | | | | | | This is a pretty simple hub that works with the doInstall function as a background thread. In the foreground we sit in a tight loop waiting for input to the Q and then handle messages or progress bar updates. Progress bar updates are handled by printing a pip (.) without a new line to the screen.
* text based storage spoke.Jesse Keating2012-08-141-0/+355
| | | | | | | | A fair amount of this code comes from the gui storage spoke. When we refactor things this will hopefully see a reduction in code. Initially we can only pick disks to use and what way to clear the partitions. Autopartitioning (plain) is forced.
* Prevent yum messages from showing on ttyJesse Keating2012-08-141-0/+6
| | | | | This is a hammer. A more subtle fix could be done to tie into yum's logging infrastructure and do something sensible with it.
* Remove unused imports from the installclasses.Chris Lumens2012-08-142-9/+0
|
* NoSuchGroup is provided by packaging now. yuminstall is on the way out.Chris Lumens2012-08-141-1/+2
|
* Set transaction color in case of multilib install.Chris Lumens2012-08-141-0/+5
|
* Add selinux-specific RPM macro setup.Chris Lumens2012-08-141-3/+12
|
* Add the user-agent to urlgrabber from the old yuminstall.py.Chris Lumens2012-08-141-0/+4
|
* Fix inheritance problems with the gui *Spoke classes.Chris Lumens2012-08-141-5/+14
| | | | | | | | | | | | | | | | | First, we need explicit __init__ methods on everything so the right superclass methods get called. This fixes a traceback I was seeing where the NormalSpoke.__init__ method was never getting called. Second, GUI spokes are inheriting from common.Spoke twice which seems likely to cause problems in the future. This is because they inherit from both gui.Spoke (which in turn inherits from common.Spoke) and a gui.*Spoke subclass which also eventually inherits from common.Spoke. The fix is to break some of the inheritance chains. There's no need to let gui.Spoke out via __all__. Nothing else should directly be inheriting from it. Then, gui.Spoke doesn't strictly need to inherit from common.Spoke. Any subclasses will get everything it needs via the common.Spoke inheritance chain.
* Call the correct method to schedule the screenJesse Keating2012-08-131-1/+1
|
* Add a missing import of osJesse Keating2012-08-131-0/+2
|
* Don't display indirect spokes in the hubJesse Keating2012-08-131-0/+3
|
* Revert "Remove unncessary __init__ definition. (clumens)"Jesse Keating2012-08-131-0/+3
| | | | This reverts commit 29c755f9ab05442b1278db2e94fcacfa8bc84d13.
* Merge master into newtuiJesse Keating2012-08-1331-509/+1011
|\ | | | | | | | | | | | | | | | | | | Conflicts: pyanaconda/ui/gui/__init__.py pyanaconda/ui/gui/spokes/custom.py One last merge before pushing to master. The conflicts were due to more development happening on those files, the result is what I hope is a clean merge.
| * Remove unused writeKS methods.Chris Lumens2012-08-1313-282/+0
| |
generated by cgit (git 2.34.1) at 2025-10-02 22:08:18 +0000