| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| ... | |
| | |
| |
| |
| |
| |
| | |
We don't get any notification that the post-trans has started, so this is the
best we can do. At least now it doesn't look like the last package is taking
forever to install.
|
| | | |
|
| | | |
|
| | |
| |
| |
| |
| |
| |
| |
| |
| | |
If you're on a spoke when all other spokes become ready and complete, the
install will continue automatically when you hit back. If you're on the hub,
the install will continue automatically when all spokes are ready and
complete.
It's a little jarring, but that's just how it's gotta be.
|
| | |
| |
| |
| |
| |
| |
| | |
It can only be sensitive both if there's no incomplete spokes and no not
ready spokes. By not tracking not ready spokes, it's possible to hit
continue while things are still downloading. This can lead to very bad
situations, and it's only by accident that we've not hit this yet.
|
| | |
| |
| |
| |
| |
| |
| |
| |
| | |
This first requires splitting out the chunk that calls execute methods into its
own function, then adding an execute method to spokes, then making sure it gets
called from everywhere. I have also added a way to make sure the execute
methods do not get run when spokes are ready, since at least the storage check
function will set the spoke not ready and then ready again which will lead to
an infinite loop.
|
| | |
| |
| |
| |
| |
| |
| | |
If you're doing an install with just Core, or with some special set of groups
that somehow doesn't include one of our desktop environments (even though Base
is currently a DE), make sure to set the spoke to complete and give a useful
status message.
|
| | |
| |
| |
| |
| |
| |
| |
| | |
This removes a lot of the special purpose code that made up the special
AnacondaKSHandler subclass. There's also no need for the _dataObjs
attribute anymore since we can't just blindly iterate over all kickstart
objects any more. Some need to be run at one point, and some at another
point.
|
| | |
| |
| |
| |
| | |
This gets their execute methods to be ready to be called from somewhere else,
but that does not yet happen.
|
| | | |
|
| | |
| |
| |
| |
| | |
When package metadata fails to download suggest that the user check the
repo url and, if enabled, the proxy settings.
|
| | | |
|
| | |
| |
| |
| |
| | |
It's easier to search through the available layouts manually if
they are sorted.
|
| | | |
|
| | |
| |
| |
| |
| |
| |
| |
| |
| | |
We still do the final check against the software's disk space
requirement with the assumption that the user is doing autopart.
Even when we have the tug of war, we're still checking free space
before the user has had a chance to choose between autopart and
custom. That will need to be addressed at some point.
|
| | | |
|
| | | |
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
This adds a way to hide devices from the devicetree without removing
them. The devicetree is initially populated without filtering because
we need to collect information about disks in the system as well as
free space in disks and filesystems. Disk selection in the storage
spoke establishes the set of disks to be used during install. We don't
want to re-populate the tree every time a disk is selected or
deselected. Instead, we just hide/unhide the disk and all devices it
contains.
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
CLEARPART_TYPE_ALL means clear existing partitions and create a new
disklabel.
CLEARPART_TYPE_LINUX means clear linux-native partitions (or linux-
native whole-disk formatting) and create a new disklabel if the
removals result in an empty disk.
CLEARPART_TYPE_LIST means clear the specified devices (and devices
built on top of them) and create a new disklabel if the removals
result in an empty disk.
Zerombr becomes an alias for 'clearpart --initlabel'.
To initialize all disks:
clearpart --all
To initialize uninitalized disks but otherwise clear no partitions:
clearpart --none --initlabel
This adds a flag to the storage discovery config class that determines
whether or not clearPartitions will remove not-yet-created devices and
disklabels. The default is to not clear non-existent devices/formats.
After the first trip to the storage spoke, we should not remove any
layout the user set up in previous visits unless they choose autopart
again, in which case we start from scratch each time, or de-select
disks. Since the disk set can change each time through we have to be
able to initialize uninitialized disks even on the way to the custom ui.
|
| | | |
|
| | |
| |
| |
| |
| |
| | |
This is all because I want the search box to only ever be as wide as the
language list view itself. Sure does take a lot of glade to accomplish
that.
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
cmdline proxy strings use the standard form:
[protocol://][username[:password]@]host[:port]
proxy url strings are stored in the ksdata, either in method.proxy for
the global proxy or per repo in repo.proxy
ProxyString handles parsing the proxy url and returning it in various
forms useful for urlgrabber, yum and the UI.
--proxyAuth has been removed. It has no meaning with noloader.
Also fix a typo with sslverify in the repo configuration file.
|
| | |
| |
| |
| |
| |
| |
| |
| | |
If a kickstart pre/post fails we read the logfile and add it to the
anaconda log. This doesn't work well for device nodes like /dev/console
so skip reading the log it isn't a regular file.
Resolves: rhbz#835563
|
| | |
| |
| |
| |
| |
| | |
This basically reverts a commit of mine from January. Keeping the entry
and the list in sync is really not a big deal, and we need to provide a more
full-featured search capability than the default GTK model one anyway.
|
| | | |
|
| | | |
|
| | | |
|
| | |
| |
| |
| |
| | |
Not having a root in self.storage.roots for the newly created set means
deletion was broken in this case.
|
| | | |
|
| | |
| |
| |
| |
| | |
It also works better, in that using new_devices means when the last of the
autopart-created set is deleted, we know to re-display the CreateNewPage.
|
| | | |
|
| | |
| |
| |
| |
| | |
This also has the effect of making sure the RHS has something relevant
displayed in it after the filesystem is deleted.
|
| | | |
|
| | | |
|
| | |
| |
| |
| |
| | |
Even if the boot disk has the correct disklabel type, we were asked
to reinitialize the disk(s), so do it.
|
| | |
| |
| |
| |
| |
| | |
- Fix a typo
- Enable remove button for devices
- Fix a tb on removal of unused devices
|
| | | |
|
| | | |
|
| | | |
|
| | | |
|
| | | |
|
| | | |
|
| | |
| |
| |
| |
| | |
If autopart is selected, we re-run clearpart and autopart on apply.
If not, we do neither.
|
| | |
| |
| |
| |
| |
| |
| |
| |
| | |
Unset autopart flag after running ksdata.autopart.execute since we
are still techically doing custom partitioning.
Don't show devices queued for removal in the custom spoke.
Don't include Roots whose root device has been queued for removal.
|
| | | |
|
| | | |
|
| | | |
|
| | | |
|
| | |
| |
| |
| |
| |
| | |
Otherwise, we end up hung waiting for XklWrapper.get_interface to return
which it appears to never do. This only happens on kickstart installs in
my testing.
|
| | | |
|
