summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Generalize value_select, cleanup and more commentsdupMartin Sivak2013-03-251-47/+23
|
* Add manual loading from media, directory and image fileMartin Sivak2013-03-251-10/+129
|
* Add rmmod blacklistMartin Sivak2013-03-251-0/+6
|
* Fix the rmmod logicMartin Sivak2013-03-251-3/+3
|
* Add driver updates code to anaconda install phaseMartin Sivak2013-03-254-7/+91
|
* Remove debug shells and copy data to anaconda environmentMartin Sivak2013-03-253-2/+17
|
* Mark variables as local and fix the copying part of DUD extractionMartin Sivak2013-03-252-33/+58
|
* Add comments and use iterative removal of modulesMartin Sivak2013-03-251-7/+30
|
* Install the driver updates module to initrdMartin Sivak2013-03-252-2/+11
|
* Add dracut module for driver updatesMartin Sivak2013-03-251-0/+237
|
* Tweak the driver update utilitiesMartin Sivak2013-03-255-12/+47
|
* Port of driver updates utilitiesMartin Sivak2013-03-256-0/+845
|
* New version.anaconda-19.13-1Brian C. Lane2013-03-223-216/+253
|
* Set Tip text on the create user spoke.David Cantrell2013-03-221-1/+1
| | | | | | | | | | | | | Remove the 'lorem ipsum...' text and set it to: Keep your username less than 32 characters and do not use spaces. I do not think it's necessary for the Tip text to explain all of the username rules in place. After all, if it fails, we can display more detailed error text for the user. The important things are length and whether or not you can use spaces. I'm not sure how common it is for people to try to use symbols in their name, but I've never heard of it and anyways, the text above is better than "lorem ipsum..."
* Use space instead of underscore when user uses the timezone name (#924352)Martin Sivak2013-03-221-3/+3
|
* Use only self.data in TUI timezone spoke's statusMartin Sivak2013-03-221-3/+1
|
* Use the named tuple in root password dialog (#924138)Martin Sivak2013-03-222-3/+3
|
* Add message instructing users they can type to search for language.Samantha N. Bueno2013-03-212-2/+3
| | | | | Persistently show the placeholder text when nothing is typed in the language search bar.
* Don't unbusy the cursor until the first action is ready to display.Chris Lumens2013-03-211-4/+4
| | | | My goal here is to make it seem like we're taking less time to set up.
* Move custom storage setup into its own thread.Chris Lumens2013-03-212-5/+14
| | | | | | | | | Instantiating all those device formats can result in lengthy processes like loading kernel modules, some of which are very large. This results in lag between when the user leaves the welcome spoke and the hub being displayed. In my thoroughly unscientific testing on one computer, this knocks about two seconds off the amount of time between the welcome screen and the hub.
* When you turn off NTP, clear the warning along the bottom of the screen.Chris Lumens2013-03-211-0/+4
| | | | | The show_alls are for re-displaying the warning when setting NTP enabled again later.
* If the disk has no serial number, don't give the DiskOverview a popup.Chris Lumens2013-03-211-1/+4
| | | | | Without a serial, it just says "None", and there's no other useful data we could give for a disk.
* Reorder the columns on the shopping cart so name is next to description.Chris Lumens2013-03-211-12/+12
| | | | This brings it in line with the reclaim dialog's column order.
* Add device node names to the resize dialog as a new column.Chris Lumens2013-03-212-4/+21
| | | | | This is how we do it on the shopping cart, too, so it's good to try to do the same thing in multiple places.
* Use an emblem for indicating spokes have not been completed.Chris Lumens2013-03-211-31/+81
| | | | | | | On the one hand, this fixes the problem of the little warning triangles floating of in space. On the other hand, this gives me no control over the color of icons and they are now grey, which looks insensitive. I think this is net progress, though.
* If you remove all the disks in the shopping cart, disable the buttons.Chris Lumens2013-03-211-0/+5
| | | | They don't do anything anyway.
* Apply a style to the network spoke's toolbar.Chris Lumens2013-03-211-0/+3
|
* Add a little more space between the updates checkbox and the add repo stuff.Chris Lumens2013-03-211-0/+1
|
* Remove the partition scheme expanders.Chris Lumens2013-03-211-86/+52
| | | | They only hide one thing, and there's no need for it to be hidden.
* Don't error out if a ks %include is missing when looking for sshpw (#923627).Chris Lumens2013-03-211-1/+1
|
* Do not guess username immediately when user clears it (#924184)Martin Sivak2013-03-211-1/+0
|
* Do not require password when no user is requested (#924150)Martin Sivak2013-03-211-1/+3
|
* Refresh the checkboxes on AdvancedUser dialog properly (#924257)Martin Sivak2013-03-211-1/+9
|
* Allow setting the default GID of the new user.Martin Sivak2013-03-213-5/+8
|
* Add call to new-kernel-pkg --rpmposttrans (#922988)Brian C. Lane2013-03-212-1/+4
| | | | Also exclude /run/ from the rsync in live installs.
* Make our gtk_* decorators safer and more intelligentVratislav Podzimek2013-03-217-31/+43
| | | | | | | | | | | | | | | | The gtk_thread_wait and gtk_thread_nowait decorators cannot be used from the main thread as it would cause a deadlock by queueing actions to the Gtk main loop and then blocking the loop by waiting for result to appear in the temporary queue. However with Thread Manager's ability to tell if we are in the main thread or not we can make these decorators more universal so that the decorated functions can be called from anywhere. Since they are now not only for threads, I've renamed them to gtk_action_wait and gtk_action_nowait. The question is if we can start decorating all methods and functions somehow manipulating with Gtk or just methods and functions that are expected to be called also from threads. That depends on the overhead these decorators bring.
* Add method for checking if in main thread to the ThreadManagerVratislav Podzimek2013-03-212-8/+24
| | | | | Also create the instance of ThreadManger in initThreading and call it from the main thread as soon as possible.
* Port the mandatory logic for User and Password spokes from GUI to TUIMartin Sivak2013-03-212-2/+3
|
* Use only self.data to determine completeness in User spokeMartin Sivak2013-03-211-1/+1
|
* Make firstboot kickstart command aware of initial-setupMartin Sivak2013-03-211-1/+3
|
* Add command and data updates to AnacondaKSHandler's __init__Martin Sivak2013-03-211-2/+2
|
* New version.Brian C. Lane2013-03-193-227/+454
|
* _model -> model in filter.py.Chris Lumens2013-03-191-11/+11
| | | | | Turns out it is used outside of the pages, so it shouldn't have an underscore name.
* Add some documentation to FilterPage.Chris Lumens2013-03-191-0/+52
|
* Add the advanced storage UI and hook it up.Chris Lumens2013-03-195-20/+1721
| | | | | | | This large patch adds the new filter spoke off the storage spoke, consisting of all the UI needed for searching and multipath, plus the code to make those two work and the framework necessary for making the other tabs work. I've added some preliminary zSeries stuff, but have not yet tested it.
* Don't wrap the DO creation in gtk_thread_wait.Chris Lumens2013-03-191-21/+15
| | | | | | This can either be called from initialize, or from refresh. Both of these happen in the main thread, so we do not want an additional wait. That just leads to a lock.
* Add a button to the specialized window to bring up the add dialog.Chris Lumens2013-03-192-5/+35
|
* Filter out multipath devices from the getDisks results.Chris Lumens2013-03-192-5/+9
| | | | We don't want these included as local standard disks.
* Reduce duplicated code between the GUI and TUI.Chris Lumens2013-03-195-101/+76
| | | | getDisks and size_str now exist in precisely one place.
* Set the horizontal and vertical scales to what we want.Chris Lumens2013-03-191-2/+2
|