- Don't create a PartitionDevice for devices that do not exist (#489122).
(clumens)
- A getter doesn't usually take a parameter (#489965). (clumens)
- Do not write "Running..." to stdout, as that could be tty1. (clumens)
- Call storage.exceptionDisks, not diskset.exceptionDisks. (#489615)
(dlehman)
- Fix typo. (jgranado)
- Fix typo. (dlehman)
- Add udev rules for handling for mdraid arrays. (dlehman)
- Honor the zerombr kickstart directive. (dlehman)
- currentSize is expected to be a float, so convert it to one (#489882).
(clumens)
- It's clearPartDisks, not clearPartDrives. (clumens)
- Get rid of the mappings and ksID as well. (clumens)
- Make sure the device has a diskType before attempting to check what it is.
(clumens)
- Update the volgroup command to work with the new storage code. (clumens)
- Update the raid command to work with the new storage code. (clumens)
- Update the part command to work with the new storage code. (clumens)
- Update the logvol command to work with the new storage code. (clumens)
- addPartRequest is no longer needed. (clumens)
- Don't set default partitioning in every kickstart case. (clumens)
- Clear partitions before scheduling requests. (clumens)
- Always go through doAutoPart. (clumens)
- Format modules import fix (mgracik)
- Fixed the format modules import (mgracik)
- Allow overriding the anaconda udev rules from an updates.img (hdegoede)
- If a pv somehow does not contain a vg_name, do not try to get other vg
info (hdegoede)
- Fix a few bugs in the lvm dialog. (#489022) (dlehman)
- Modify livecd.py to work with new storage backend. (dlehman)
- Be explicit about resetting Disks' partedDisk attribute. (#489678)
(dlehman)
- Deactivate devices after we've finished scanning them. (dlehman)
- Handle the case of removing an unallocated partition from the tree.
(dlehman)
- Try again to set up LVs when we've just added a new PV to the VG. (dlehman)
- Set partition flags in format create/destroy execute methods. (dlehman)
- Make sure we use the newly committed parted.Partition after create.
(dlehman)
- Make device teardown methods more resilient. (dlehman)
- Initialize storage in rescue mode so we can find roots (#488984). (clumens)
- We also need to pack up the extra args tuple, too. (clumens)
- doLoggingSetup keeps growing new arguments, so put them into a dict
(#489709). (clumens)
- Fix anaconda udev rules to not require pre-existing device nodes (hdegoede)
- Hook up 'Shrink current system' dialog to new storage code. (dcantrell)
- Fix _getCheckArgs() in class FS. (dcantrell)
- Fix action pruning to handle more complex scenarios. (dlehman)
- Schedule destruction of any existing formatting along with the device.
(dlehman)
- Add a size attribute to mdraid arrays. (dlehman)
- Speed up partitioning screen redraws by trimming workload where possible.
(dlehman)
- Create partitions with exactly the geometry we calculate. (dlehman)
- Fix name collision between formats.mdraid and devicelibs.mdraid. (dlehman)
- Destruction of the member device formatting will be handled elsewhere.
(dlehman)
- Fix a typo (jkeating)
- Fix pruning between two destroy actions on the same device (rvykydal)
- Use the pyblock functions when possible. (jgranado)
- We are searching a list, not a dict now (rvykydal)
- Move the recursive teardown of all devices out of processActions. (dlehman)
- Clean up handling of /proc, /sys, /dev/pts, /dev/shm entries. (dlehman)
- Fix several minor bugs preventing upgrade/rescue mount. (#488946) (dlehman)
- Only populate the device tree on demand. (dlehman)
- Prune actions by device based on path, not object-id. (dlehman)
- Rewrite action sort so it works correctly. (dlehman)
- Do a separate disk.commit for each partition add/remove. (dlehman)
- Fix bug keeping track of best free region/type/disk info. (dlehman)
- Return early if doAutoPart is False, but clearpart first if kickstart.
(dlehman)
- Recognize PS3 as a valid machine type (#489263). (clumens)
- Move the mdRaidBootArches logic into the platform module. (clumens)
- stdout and stderr may also need to be created. (clumens)
- Fix booty for dmraid (hdegoede)
- It's self.origrequest, not self.origreqest (#489036). (clumens)
- Added crypto.py unittest; Updated devicelibs tests baseclass.py and lvm.py
(mgracik)
- Start storage before parsing the kickstart file. (clumens)
- Make sure autopart without any clearpart command will fail. (clumens)
- Update storage flag on ks autopart (rvykydal)
- Use correct storage attribute for ks clearpart (rvykydal)
- Catch the new _ped.DiskLabelException for unrecognized disklabels.
(dlehman)
- Catch all failures from making parted objects in exceptionDisks. (dlehman)
- various dmraid fixes. (jgranado)
- Implement the format disk question as a callback. (jgranado)
- Add dmraid functionality to new storage code. (jgranado)
- Do not pass None values into nonmandatory arguments, you are screwing the
default values.. (msivak)
- Schedule device destroy actions for partitions last. (dlehman)
- Pass storage.disks, not storage, to createAllowed.... (#488860) (dlehman)
- Nodev filesystems always exist. And the device is arbitrary. (dlehman)
- Include proc, &c filesystems in fstab and FSSet.{mount/umount}Filesystems.
(dlehman)
- Remove FSSet.writeFSTab. That job is handled elsewhere. (dlehman)
- Add properties to FSSet to provide the nodev entries. (dlehman)
- Fix incomplete format in Storage.deviceImmutable. (dlehman)
- Make sure we use the same disk the free space is on. (#488807) (dlehman)
- Prevent clobbering of name 'mdraid' by qualifying it. (dlehman)
- Handle unformatted disks and cdroms in Storage.exceptionDisks. (dlehman)
- Add resizeArgs property for resizable filesystems. (dcantrell)
- Fill out class NTFS a bit more. (dcantrell)
- Add fsckProg property to class FS. (dcantrell)
- Ext2FS.migratable(self) -> Ext2FS.migratable (dcantrell)
- Fix StorageDevice.minSize() and PartitionDevice.maxSize() (dcantrell)
- Center resize window on the screen. (dcantrell)
- Do not raise DeviceError if not bootable device is found. (dcantrell)
- Do an even more thorough job of ignoring disks libparted doesn't like.
(clumens)
- Fix a couple problems on the "Change device" bootloader dialog. (clumens)
- Fix a typo when writing out the mdadm config file. (clumens)
- Remove all uses of isys.cdromList, which no longer exists. (clumens)
- Check to see if we're on S390 on the congrats screen (#488747). (clumens)
- Handle non-fatal errors more gracefully in addUdevDevice. (dlehman)
- partRequests no longer exists, so don't try to import it (#488743).
(clumens)
- When building the exceptionDisks list, skip devices libparted doesn't
like. (clumens)
- Iterate over devicetree.devices.values, not devicetree. (dlehman)
- Add a get() method to Flags, since it pretends to be a dictionary.
(clumens)
- Stop with the fsset usage. (dlehman)
- Format message string after translation not before (msivak)
- We need newer python-cryptsetup because of the default values for cipher
and keysize for luskFormat (msivak)
- If a drive is not initialized, offer reinitialization or ignoring the
drive to the user (msivak)
- More syntax errors / traceback fixes (hdegoede)
- Fix syntax errors (rvykydal)
- Implement Storage.sanityCheck, mostly from old partitions code. (dlehman)
- Remove old content from utils/ (dcantrell)
- Ensure request.drive is always a list (#485622) (dcantrell)
- Pick up pyblock if it exists in block/ on an updates.img. (dcantrell)
- Don't check for a swapfs on things that aren't partitions (#485977).
(clumens)
- Perform ext3->ext4 filesystem migration if ext4migrate is given (#484330).
(clumens)
- Translate i?86 into i386 as a base arch. (jkeating)
- Teach upd-instroot about i586 arch, for sake of glibc.i586/openssl.i586
(jkeating)
- Fix the obvious typo. (clumens)
- filer.login raises an exception with it can't login, not returns None
(#486454). (clumens)
- Take into account that a parted.Partition's _fileSystem can be None
(#485644). (clumens)
- Fix indentation on upd-instroot (kanarip)
- Fix the indentation in mk-images (kanarip)
- Remove unused iface_netmask2prefix() function. (dcantrell)
- A parted.Disk has no attribute named "dev". It's named "device"
(#486007). (clumens)
- Use brandpkgname for the efi art too (katzj)
- Let's use the product string for a brandpackage name. (kanarip)
- Fix indentation in mk-images.efi (kanarip)
- Fix indentation in buildinstall script (kanarip)
- It's part.active, not part.is_active(). (clumens)
- File the basic traceback as the first comment instead of a generic
message. (clumens)
- Encode our upgrade policy in productMatches/versionMatches and enforce it.
(clumens)
- If we'd show package selection on kickstart installs, also show tasksel.
(clumens)
- If ks=nfs:... is given, don't try to find the file via boot options
(#480210). (clumens)
- Fix cdrom install on machines with no network devices (wwoods)
- updated fuzzy strings (jsingh)
- Use modinfo to find out what firmware we need in initrd (wwoods)
- Use the preconf object for yum configuration now (jantill). (clumens)
- Updated Dutch translation adn only 1 -fuzzy- string left (zuma)
- Add a boot target for the xdriver=vesa parameter and document it. (clumens)
- repo.proxy is now a property, so check before setting it (#481342).
(clumens)
- Workaround compile error due to (# 478663) (hdegoede)
- Various packaging fixed from review (#225246) (hdegoede)
- Show the header in certain non-lowres cases (#478765, alsadi AT
ojuba.org). (clumens)
- Remove doMultiMount. (clumens)
- Use mount -t auto instead of passing a list of valid fstypes (#477328).
(clumens)
- Fix case sensitivity when searching for headers (kanarip)
- Fix a traceback in checking for network install (ricky AT
fedoraproject.org). (clumens)
- Reduce direct D-Bus calls in isys/iface.c. (dcantrell)
- Allow 'ks' to function as it once did (#471812) (dcantrell)
- Fix telnet install support (#471082) (dcantrell)
- Call 'udevadm settle' instead of 'udevsettle'. (dcantrell)
- When using anaconda with kickstart file with UI mode - do not show the VNC
question (#476548) (msivak)
- Check error from asprintf() correctly for dhcpclass handling. (dcantrell)
- Use libnm_glib in net.c:get_connection() (dcantrell)
- Add libnm_glib CFLAGS and LIBS to loader's Makefile. (dcantrell)
- BR NetworkManager-glib-devel. (dcantrell)
- Only write the short hostname to the localhost line (#474086) (dcantrell)
- Updated Tajik Translation - Victor Ibragimov (victor.ibragimov)
- Copy /etc/dhclient-DEV.conf file to target system (#476364) (dcantrell)
- Use macros for D-Bus paths (dcantrell)
- Let X tell us when it's launched rather than just sleeping. (ajax)
- When there's no baseurl, set a default of [] instead of [''] (#476208).
(clumens)
- cracklib now raises exceptions on bad passwords (rzhou, #476312). (clumens)
- Make sure ssh doesn't get duplicated in the open port list (#474937).
(clumens)
- mdraid1: default to putting grub on partition instead of mbr (#217176)
(hdegoede)
- Don't install the games group as part of office/productivity (#472324).
(clumens)
- Don't dump encryption passphrases. (dlehman)
- Write anacdump.txt upon receipt of SIGUSR2 (from clumens). (dlehman)
- Use stacks instead of tracebacks in traceback handlers. (dlehman)
- Unmount swap devices when migrating filesystems, then reactivate
(#473260). (clumens)
- Handle both /dev/sr0 and sr0, since that's what cdromList gives (#475083).
(clumens)
- In iface_ip2str(), make sure to advance to next item before continue.
(dcantrell)
- We already have _GNU_SOURCE defined in Makefile.inc (dcantrell)
- Remove XXX comment in net.c about GATEWAY (dcantrell)
- Use strverscmp() from glibc in place of rpmvercmp() (dcantrell)
- Remove readLine() function from loader/loadermisc.c (dcantrell)
- Do not write SEARCH line to ifcfg-DEVICE file (#474858) (dcantrell)
- Preserve existing network configuration files during install (#461550)
(dcantrell)
- Send unique vendor class identifier unless user specifies one. (dcantrell)
- Avoid tracebacks when filling in static network config fields (#474275)
(dcantrell)
- Prevent network install when no network devices are found (#470144)
(dcantrell)
- Remove markup from text before printing it in cmdline mode (#470253).
(clumens)
- Move strip_markup() into iutil. (clumens)
- Fix up plural forms header so that python doesn't blow up for us (katzj)
- Change text to reflect Jesse's comments (katzj)
- Add support for the Tajik language (#455963). (clumens)
- Add a button to the UI to ignore all missing packages. (clumens)
- First small eu.po transtation, just to be sure that the system is set up
OK. (mikel.paskual)
- mini-wm: Turn on automatic window redirection. (ajax)
- Better naming for LVM volume groups and logical volumes (#461682)
(dcantrell)
- Partition requests can be None when populating the tree. (#474284)
(dlehman)
- Say we are unable to configure the network interface (#467960) (dcantrell)
- Match textw/network_text.py strings to iw/network_gui.py (#470145)
(dcantrell)
- In addSnap(), check snapshots for data key before continuing (#433824)
(dcantrell)
- Load FCP modules early for CD/DVD install (#184648) (dcantrell)
- Update mk-s390-cdboot.c to work with large kernel images (#184648)
(dcantrell)
- Make sure fstype exists before we try to test it (#473498). (clumens)
- Updated a small correction in kn locale (svenkate)
- Use modules.* files for finding modules of a type rather than modinfo
(katzj)
- Make complete text mention updates (#244431) (katzj)
- Make text for autopartitioning types clearer (#441350) (katzj)
- Allow installing grub on the MBR if /boot is on mdraid (#217176) (hdegoede)
- Fix some spelling errors in German translation (fabian)
- Make the required media dialog less wordy (#469557). (clumens)
- returnNewestByName now raises an error instead of returning [] (#472462).
(clumens)
- Fix death on login of an OLPC on a live image (katzj)
- Fix ld-*.so globbing for glibc-2.9 . (pjones)
- Do not bring up network for non-remote kickstart locations (#471658)
(dcantrell)
- Resolve dm-X devices returned by pvdisplay. (#448129) (dlehman)
- More shell script syntax fixing (katzj)
- Only bring up the network dialog on package failures if required
(#471502). (clumens)
- Don't have the key icon take up so much space on the LUKS dialog
(#470338). (clumens)
- Avoid getting linux-base in the kernel list (katzj)
- Deselect groups when we reset things also (#469854) (katzj)
- make iscsi login code wait for udev to create the devices (#466661,
#470073) (hdegoede)
- Set the correct path when using the directory chooser. (clumens)
- We always need a wait window, not just when the repo has a name. (clumens)
- Set initial state of IP configuration fields in text mode (#469933)
(dcantrell)
- Prevent traceback when there are no network devices (#469339) (dcantrell)
- Indentation fix. (pjones)
- Let users edit net settings on network failure in stage 1 (#465887)
(dcantrell)
- Move startNewt later to avoid printing extra messages on the screen
(#469687). (clumens)
- Revert "Make sure dialog deletions take effect sooner (#455676)." (clumens)
- Don't set up the launcher for the installer on XO (katzj)
- Whitespace cleanups for timezone.py (dcantrell)
- Do not store mount options in loaderData->instRepo (#467760) (dcantrell)
- Make sure we look up the IP address for the correct device (#469439)
(dcantrell)
- Remove unused bool() function. (dcantrell)
- Check for required space for / on live installs (#468867) (katzj)
- Add a basic method for checking the minimal size needed for a backend
(katzj)
- Fix typo that somehow snuck in (katzj)
- If there's no language selected, don't traceback (#469578). (clumens)
- Improve filtering of non-available groups (#469438) (katzj)
- filer.py: set defaultProduct in __init__ (hdegoede)
- Fix indentation error in filer.py (again) (hdegoede)
- Rebuild keymaps to get rid of trq.map (#469433). (clumens)
- Provide sample punch card reader script for s390x (#462953) (dcantrell)
- Fix a typo that shouldn't have even gotten though. (clumens)
- Check that the platform and product are also correct (#469367). (clumens)
- Remove cio_ignore functionality for s390x (dcantrell)
- Remove bootdisk/s390 (dcantrell)
- If method=nfs: is given, check if it's really an NFSISO install (#468885).
(clumens)
- Get the right list elements for the iscsi text interface (#466902).
(clumens)
- Don't traceback when displaying error messages (#469372). (clumens)
- Make sure we differentiate locked luks devs from deleted ones. (dlehman)
- Fix a typo that breaks kickstart with encryption. (#469318) (dlehman)
- Call startNewt earlier than network bring up (#469171). (clumens)
- Write out the path to the repo, not anaconda-ks.cfg (#467753). (clumens)
- Allow specifying devices by path if they're files (#468504) (katzj)
- Fix the last pychecker warnings in master (hdegoede)
- Add --strict option to runpychecker.sh (hdegoede)
- Don't sleep(5) after xrandr (ajax)
- Force DPI to 96 even harder (#458738) (ajax)
- Don't try to switch VT to the one that X is on (ajax)
- Only copy /etc/resolv.conf if instPath != '/' (dcantrell)
- 'is not' -> '!=' (dcantrell)
- Write --dhcpclass instead of --class to the anaconda ks file. (jgranado)
- Fix 2 issues in pyparted found by pychecker (hdegoede)
- Add a bit of documentation to the top of runpychecker.sh (hdegoede)
- Add runpychecker.sh script and pychecker-false-positives file (hdegoede)
- Fix saving tracebacks via scp while in text mode. (clumens)
- Search for the hash in the whiteboard, not as the entire whiteboard.
(clumens)
- Fix various syntax errors caught by PyChecker (hdegoede)
- Wouldn't it be nice to have some real documentation in filer.py? (clumens)
- Make sure the productVersion given by .treeinfo exists in bugzilla
(#468657). (clumens)
- Let DNS lookups work from %post scripts (#468132) (dcantrell)
- Do not use /.tmp for temporary files (#468720) (dcantrell)
- Don't treat encrypted PVs as available if we don't have the key. (#465240)
(dlehman)
- Do all new device passphrase prompting from partitioningComplete. (dlehman)
- Fix the obviously stupid typo. (clumens)
- There's a new version of the firewall command for F10 (#467753). (clumens)
- Another fix for printing package summaries in text mode (#468283).
(clumens)
- Fix traceback in network.bringUp() (#468651) (dcantrell)
- lvresize requires a --force arg now (#468478) (katzj)
- Include return code on resize failure error message (#468479) (katzj)
- Catch UnicodeDecodeError so traceback messages display anyway. (dcantrell)
- Do not write NM_CONTROLLED=yes to ifcfg files (#468028) (dcantrell)
- Log D-Bus messages at ERROR or INFO level. (dcantrell)
- Write dhcpclass to the dhclient conf file for the device (#468436)
(dcantrell)
- Tell NetworkManager not to touch network interfaces when / is a netfs
(hans)
- Catch more X failures and fallback to text (#467158). (clumens)
- Fix a typo when using network --gateway (#468364). (clumens)
- Fix icon (#468273) (katzj)
- Remove extra debug info. (pjones)
- Fix the damn spinner in the progress bar. (pjones)
- Fix whitespace. (pjones)
- Fix "looking for installation images" when there's no disc at all. (pjones)
- Make sure dialog deletions take effect sooner (#455676). (clumens)
- Make cdrom drive door status messages be INFO not DEBUG. (pjones)
- Don't switch to tty6 on vnc installs. (clumens)
- Update font list (#462295). (clumens)
- Don't display the entire lengthy device description (#467825). (clumens)
- Fix ext4 detection on existing partitions (#467047) (rvykydal)
- Make sure we handle the /tmp/method file for FTP correctly (#467753).
(clumens)
- Do not write NM_CONTROLLED=yes to ifcfg files (#468028) (dcantrell)
- Revert "dhclient-script not needed for NetworkManager" (clumens)
- Skip Installation Repo when writing out repo kickstart lines. (clumens)
- Correct media check docs (#468061). (clumens)
- Many translation updates
- Make kickstart installs work again (#374271, #392021, #448096, #466340,
#466304) (dcantrell)
- Let users go Back when loading updates. (dcantrell)
- Write ifcfg files to /etc/sysconfig/network-scripts instead of /.tmp
(dcantrell)
- Handle unknown hosts in getDefaultHostname (#466775) (dcantrell)
- Try to look up the hostname by the IP address NM reports (#466775)
(dcantrell)
- NM no longer provides the hostname as a property (#466775). (clumens)
- ext4dev -> ext4 (esandeen). (clumens)
- Move persistent network udev rule to under /etc (#464844). (clumens)
- Update keymaps to include latest Romanian settings (#466117). (clumens)
- Take ip= parameter values by not resetting ipinfo_set. (dcantrell)
- Remove unnecessary STEP_IP code. (dcantrell)
- Fix how configureTCPIP() returns. (dcantrell)
- Write new sysconfig data to a tmpdir first, then move in place. (dcantrell)
- Write NM_CONTROLLED=yes rather than NM_CONTROLLED= (dcantrell)
- Get rid of some iface flags that were not doing anything anymore.
(dcantrell)
- Generate new config files in /.tmp in writeEnabledNetInfo() (dcantrell)
- Remove unused variables from configureTCPIP() (dcantrell)
- Do not call get_connection() twice for DHCP. (dcantrell)
- Ask for language and keyboard in rescue mode (#466525). (clumens)
- Fix bringing up the network in rescue mode (#466523). (clumens)
- If we don't have a translation for a lang name, just use the English
(#466515) (katzj)
- Disable some more IPv6 checks. (clumens)
- Fix a typo (second part of #466374) (katzj)
- Pull in static network settings from the boot: line (#465270) (dcantrell)
- Do not segfault when going back to select a new interface (#465887)
(dcantrell)
- Do not test for DNS settings in mountNfsImage() (dcantrell)
- Populate struct iface correctly in setupIfaceStruct() (dcantrell)
