summaryrefslogtreecommitdiffstats
path: root/perl-install
Commit message (Collapse)AuthorAgeFilesLines
* use fsedit::is_same_hd in drakupdate_fstab and revert previous change in ↵Olivier Blin2004-06-292-3/+3
| | | | fs::subpart_from_wild_device_name (do not fill both devfs_device and device fields)
* in is_same_hd(), both hds can have {device} unset, it doesn't mean they are ↵Pascal Rigaux2004-06-291-1/+1
| | | | the same
* do not add default device in device hash, it will be filled with result from ↵Olivier Blin2004-06-291-1/+1
| | | | fs::subpart_from_wild_device_name
* always update $part{device} in fs::subpart_from_wild_device_name because ↵Olivier Blin2004-06-291-2/+2
| | | | /dev is stripped (fix #6982, #10175)
* - add bootloader-config (used by bootloader-utils and bootsplash scripts)Pascal Rigaux2004-06-291-4/+23
| | | | | | | | | | | | | | | | | | | | | | - drakboot (pixel): o major backend rewrite b/c of code sharing with new installkernel o when adding a new kernel, have a nicer new name for conflicting entry o when modifying kernel parameters in all entries, skip the "failsafe" entry (#10143) o when modifying a symlink, ensure we also use the long name for the old symlink in the existing entries - drakconnect (Olivier Blin): o never disable "DHCP host name" entry box, it shouldn't be linked with "Assign host name from DHCP address" checkbox (#2759, #9981) o unblacklist sis900 since its link beat detection works with latest kernels - draksound: remove unneeded "above" lines in modules::write_conf (Olivier Blin) (#8288) - ugtk2 layer: catch missing wizard pixmap, otherwise we end up with unshown windows and error messages can't pop up (pixel) - don't require mkbootdisk
* add missing share/net_monitor.desktopPascal Rigaux2004-06-291-1/+1
|
* add bootloader-config (tested for adding and removing in simple cases)Pascal Rigaux2004-06-292-1/+121
|
* - mkinitrd() calls make-boot-splash after building the initrdPascal Rigaux2004-06-291-47/+121
| | | | | | | | | | | | | (since /sbin/mkinitrd doesn't do it anymore) - change mkinitrd() parameters - change add_kernel() parameters and add option b_no_initrd (it is available in installkernel though i suspect it is unused) - after reading config file, get {perImageAppend} and {default_vga} (it was only done for lilo and {perImageAppend}) - keep the current kernel for linux-nonfb too (it was only done for failsafe) - add create_link_source(), action(), when_config_changed_{lilo,grub,yaboot} - create write_yaboot()
* add "Getopt::Long" (used by bootloader-config)Pascal Rigaux2004-06-291-0/+1
|
* make perl_checker a little more happyPascal Rigaux2004-06-281-1/+1
|
* after install, when modifying kernel parameters in all entries, skip the ↵Pascal Rigaux2004-06-281-1/+3
| | | | | | | "failsafe" entry (bugzilla #10143)
* catch missing wizard pixmap, otherwise we end up with WizardWindow being nonPascal Rigaux2004-06-281-1/+2
| | | | | shown and error messages can't pop up
* disallow switching to expert mode using Alt-ePascal Rigaux2004-06-281-3/+0
|
* fix titi's regexp for matching centrinoPascal Rigaux2004-06-281-1/+1
|
* don't set ICEAUTHORITY using $ENV{HOME} for kde (bugzilla #10144)Pascal Rigaux2004-06-281-2/+8
|
* drop useless linkage since X11 stuff was split out of c moduleThierry Vignaud2004-06-271-1/+0
|
* *** empty log message ***Pascal Rigaux2004-06-251-1/+83
|
* use internal_error() instead of die'ing with no argumentsPascal Rigaux2004-06-251-1/+1
|
* fix blino stage1 change (i think he likes breaking things ;p)Pascal Rigaux2004-06-252-2/+2
|
* remove deprecated rhimage symlinkPascal Rigaux2004-06-251-1/+0
|
* bootloader::add_kernel() do a good job at replacing symlinks with the realPascal Rigaux2004-06-251-17/+0
| | | | | kernel/initrd file name, so don't need to do it here
* when adding a new kernel, have a nicer new name for conflicting entryPascal Rigaux2004-06-251-3/+8
|
* when modifying a symlink, ensure we also use the long name for the old ↵Pascal Rigaux2004-06-251-15/+38
| | | | symlink in the existing entries
* Make perl_checker silentRafael Garcia-Suarez2004-06-251-6/+6
|
* fix 2 typosPascal Rigaux2004-06-251-1/+1
|
* never disable "DHCP host name" entry box, it shouldn't be linked with ↵Olivier Blin2004-06-251-1/+1
| | | | ""Assign host name from DHCP address" checkbox (#2759, #9981)
* unblacklist sis900 since its link beat detection works with latest kernelsOlivier Blin2004-06-251-1/+1
|
* *** empty log message ***Pascal Rigaux2004-06-251-1/+514
|
* drop obsolete functionPascal Rigaux2004-06-251-8/+0
|
* drop obsolete lnx4win_file()Pascal Rigaux2004-06-251-5/+0
|
* simplify using new bootloader.pmPascal Rigaux2004-06-251-3/+3
|
* adapt to new bootloader.pmPascal Rigaux2004-06-254-12/+12
|
* - adapt to new bootloader.pmPascal Rigaux2004-06-251-6/+3
| | | | | - simplify
* major cleanup and rewritePascal Rigaux2004-06-251-359/+419
| | | | | | | | | | | | | | | | | | | | - some functions have been renamed - some functions have changed prototype - no more bootloader::vga_modes, we use Xconfig::resolution_and_depth::bios_vga_modes() - no more detectloader, it is now detect_main_method(), but notice that bootloader::read() + bootloader::write() abstract all this! - update_for_renumbered_partitions() is surely broken :) - sort_hds_according_to_bios() and mixed_kind_of_disks() uses new function hd2bios_kind() - new data structure "kernel_str" - lilo doesn't need so much help as it used to be, so don't put as much "disk=/dev/sda bios=0x80" as before some goals: - don't rely on device names (this allows not to care too much about devfs vs udev vs ...) - kernels can be named something else than /boot/vmlinuz*, please use the various functions to construct the initrd file name, the symlink name...
* - enhance subpart_from_wild_device_name() to handle "sda" the same as ↵Pascal Rigaux2004-06-251-21/+35
| | | | | | | | | "/dev/sda" (when /dev/sda exists) - new function device2part() which uses subpart_from_wild_device_name and the given fstab to convert the device name to the corresponding structure (this allows not relying too much on the device name)
* - scim-uim requires scim, no need to have bothPascal Rigaux2004-06-251-2/+2
| | | | | - scim-chinese requires scim, no need to have both
* rollback service priorityThierry Vignaud2004-06-241-1/+1
|
* perl_checker complianceOlivier Blin2004-06-241-1/+1
|
* remove unneeded "above" lines in modules::write_confOlivier Blin2004-06-241-0/+1
|
* 10-48mdkThierry Vignaud2004-06-241-1/+19
|
* delete "above" lines when removing a sound alias with ↵Olivier Blin2004-06-241-1/+5
| | | | modules::remove_alias_regexp
* switch chinese to scim by defaultThierry Vignaud2004-06-241-2/+1
|
* fix gtk IM module for scim IMThierry Vignaud2004-06-241-1/+1
|
* - switch japanese to scim+uim and scimThierry Vignaud2004-06-241-2/+2
| | | | | - switch chinese to scim
* fix stupid typoOlivier Blin2004-06-241-1/+1
|
* in modules::when_load, try to find the best sound slot index instead of ↵Olivier Blin2004-06-241-1/+5
| | | | always overwritting sound-slot-0 (#7890)
* (acceptLicense) just reboot when one cancel the globetrotter first time wizardThierry Vignaud2004-06-241-0/+4
|
* fix xcin IMThierry Vignaud2004-06-241-3/+1
|
* - add gv (not installed by default) to have it in the package tree (bugzilla ↵Pascal Rigaux2004-06-241-2/+2
| | | | | | | #10127) - add pciutils (not installed by default) in MONITORING (but i don't think it will get to the package tree)
* module hid is now named usbhid (thanks to svetljo on cooker)Pascal Rigaux2004-06-241-1/+1
|