From ec6144a1513bb11f14bfc46222476233ce8a5be8 Mon Sep 17 00:00:00 2001 From: Pascal Rigaux Date: Fri, 31 Dec 2004 09:34:02 +0000 Subject: *** empty log message *** --- perl-install/ChangeLog | 375 ++++++++++++++++++++++++++++++++++++++++++++++++- 1 file changed, 374 insertions(+), 1 deletion(-) diff --git a/perl-install/ChangeLog b/perl-install/ChangeLog index 1fea36048..3b1529cb3 100644 --- a/perl-install/ChangeLog +++ b/perl-install/ChangeLog @@ -1,3 +1,376 @@ +2004/12/31 Pixel + + * install_steps_gtk.pm: simplify (thanks to perl_checker) + remove code using getAndSaveFile to get the X server + (we don't have live directory anymore, and anyway we only use one X + server) + + * interactive/gtk.pm: better layout: + - ensure buttons are at bottom using pack_end instead of packing a + growing empty vbox for ask_warn-like dialogs + - drop old code + - always allow box to grow (this may be wrong) + +2004/12/30 Pixel + + * ugtk2.pm: more spacing around the separator and the buttons (as + requested by ergonomy team) + + * install_steps.pm: finish backporting HEAD changes for finish-install + (any::write_passwd_user() doesn't exist anymore) + +2004/12/29 Pixel + + * lvm.pm, devices.pm: move device mapper initialisation out of lvm.pm into + devices.pm (since it will be useful for other things) + + * install_steps_interactive.pm: when computing group size, select + CAT_SYSTEM packages + + * pkgs.pm: simplify and speed up computeGroupSize() + in computeGroupSize() + - fix old ugly bug (CAT_FOO && CAT_BAR || CAT_FOO was reduced to + CAT_FOO && CAT_BAR) + - log the time spent + - ignore already selected packages + - group by same rpmsrate flags (nice speedup :) + +2004/12/28 Pixel + + * install_gtk.pm: fix overlap + re-organize steps display as requested by ergonomic team + + * install_steps_gtk.pm: fix displaying package tree (was empty since + moving from XXX to CAT_XXX) + + * mygtk2.pm: + - add can_default + - remove known_opts (alike the rest of the code) + + * pkgs.pm, install_any.pm, install_steps_interactive.pm: + - create set_rpmsrate_category_flags(), + set_rpmsrate_default_category_flags(), default_packages(), + rpmsrate_always_flags() replacing setDefaultPackages() + - create select_default_packages() + - apply "always" rpmsrate flags ASAP and skip packages with flag + "FALSE" in computeGroupSize() + => 30% speed-up + +2004/12/24 Pixel + + * share/rpmsrate, install_any.pm, install_steps_interactive.pm: + differentiate compssUsers flags from non-user modifiable flags (eg: + hardware flags) + + * network/drakfirewall.pm: "Samba server" is better named "Windows Files + Sharing (SMB)" (bugzilla #10585) + +2004/12/23 Pixel + + * standalone/XFdrake: fix Xconfig::default::configure() use + "XFdrake --auto" is now fully auto using auto_install configuration + instead of "as much auto as can be" + - don't ask X to restart in auto mode + - set exit status to 1 when configuration failed (esp. auto mode) + +2004/12/23 Rafael Garcia-Suarez + + * tools/checkusedmodules: Add a new check script in tools. It compares the + perl modules used by the .pm + files in perl-install against the ones listed in share/list, to detect + potential missing modules (and potential run-time problems during the + stage 2) + +2004/12/23 Thierry Vignaud + + * drakxtools.spec: bump require on perl in order to prevent beakage on + update due to + binary incompatibility (#12719 and the like) + fix typo in 10.2-0.7mdk's changelog + 10.2-0.7mdk + +2004/12/22 Pixel + + * pkgs.pm: simplify + fix previous commit + create read_rpmsrate_raw() out of read_rpmsrate(), this new function can + be easily used outside install + + * install_steps_interactive.pm: no need to use max_size, no need to + compute it + + * install2.pm, fsedit.pm: add option no_bad_drives (to workaround bugzilla + #12766) + +2004/12/22 Thierry Vignaud + + * harddrake/sound.pm: "snd-audigyls" ALSA driver was renamed "snd-ca0106" + in ALSA CVS; + let's have our tools be aware of that once our kernel got synced + handle new snd-pcxhr driver from alsa CVS + + * share/rpmsrate: install proper tools for Digigram PCXHR sound card + +2004/12/21 Pixel + + * install2.pm: /proc/sys/kernel/modprobe wants something non empty, + otherwise it doesn't change anything + + * install_any.pm: don't pass prefix, use $::prefix + don't pass prefix to network::netconnect::main() (it doesn't use it) + don't pass prefix, use $::prefix + + * fs.pm: + - fs::mount() wants a real device or a faked one, but doesn't accept + things like fd0 anymore (give it /dev/fd0) + - a little more robust analyze_wild_device_name() during install + - handle bad things in subpart_from_wild_device_name() + + * install_steps_interactive.pm: don't pass prefix, use $::prefix + don't pass prefix to network::netconnect::main() (it doesn't use it) + + * mdk-stage1/doc/UPDATEMODULES: much simpler code to handle post-install + update modules + - it handles new modules + - it calls depmod + + * install_steps_gtk.pm, commands, standalone/drakbug_report, pkgs.pm, + crypto.pm, commands.pm: don't pass prefix, use $::prefix + + * standalone/drakconnect, network/netconnect.pm, printer/printerdrake.pm: + don't pass prefix to network::netconnect::main() (it doesn't use it) + + * any.pm: + - fs::mount() wants a real device or a faked one, but doesn't accept + things like fd0 anymore (give it /dev/fd0) + - a little more robust analyze_wild_device_name() during install + - handle bad things in subpart_from_wild_device_name() + don't pass prefix, use $::prefix + + * install_steps.pm: + - fs::mount() wants a real device or a faked one, but doesn't accept + things like fd0 anymore (give it /dev/fd0) + - a little more robust analyze_wild_device_name() during install + - handle bad things in subpart_from_wild_device_name() + don't pass prefix, use $::prefix + don't pass prefix, use $::prefix + much simpler code to handle post-install update modules + - it handles new modules + - it calls depmod + + * detect_devices.pm: fix floppies detection (through dmidecode) + +2004/12/20 Pixel + + * install2.pm: don't die when /proc/sys/kernel/modprobe doesn't exist (for + BOOT kernels) + + * mdk-stage1/tools.c, mdk-stage1/tools.h, mdk-stage1/automatic.c: minimal + bootsplash stage1 support + create unset_automatic() + + * mdk-stage1/modules.c, mdk-stage1/cdrom.c, mdk-stage1/stage1.c, + mdk-stage1/network.c: create unset_automatic() + + * mdk-stage1/Makefile, mdk-stage1/bootsplash.h, mdk-stage1/bootsplash.c, + mdk-stage1/frontend.h, mdk-stage1/stdio-frontend.c, + mdk-stage1/newt-frontend.c: minimal bootsplash stage1 support + +2004/12/20 Thierry Vignaud + + * network/modem.pm (get_user_home) do not horribly die if USER environment + variable list + a non existing user + +2004/12/19 Funda Wang + + * lang.pm: jp console should not be localized + jp console should not be localized + +2004/12/18 rstandtke + + * share/po/de.po: added some translations + +2004/12/17 Pixel + + * install_gtk.pm: don't parse the theme rc file to set the root window + background color during install, + instead use the gc of a fake window named "background" + switch to mygtk2 + + * ugtk2.pm: add some more functions in mygtk2 + + * Xconfig/test.pm: remove unused code + + * modules.pm: packdrake changed a little, ensuring that if we ask no + modules it doesn't cause havoc + + * mygtk2.pm: don't parse the theme rc file to set the root window + background color during install, + instead use the gc of a fake window named "background" + oops, set_background *is* needed + remove unneeded ->set_background + add some more functions in mygtk2 + + * bootloader.pm: ia64 uses the lilo method (even if it has some important + differences) + on ia64, use method lilo + +2004/12/17 Thierry Vignaud + + * network/netconnect.pm: fix NETMASK autofilling + +2004/12/16 Gwenole Beauchesne + + * c/smp.c: merge smp detection from fedora/ydl + +2004/12/16 Pixel + + * install2.pm: help testing + + * mygtk2.pm: + - add DrawingArea + - add text_ref & format for buttons + remove debug code + we must allow more than one callback per ref & widget + add option allow_unknown_options for easy compatibility + + * detect_devices.pm, pkgs.pm, install_any.pm: create function + detect_devices::BIGMEM() which calls c::dmiDetectMemory(), but only + if we are root (this helps testings install without dying) + + * install_steps_gtk.pm: migrate to mygtk2 + + * ugtk2.pm: fix typo (mygtk2 wants active_ref, not val) + add option allow_unknown_options for easy compatibility + + * Xconfig/resolution_and_depth.pm: much simpler code using mygtk2 + +2004/12/15 Gwenole Beauchesne + + * mdk-stage1/insmod-modutils/include/util.h, + mdk-stage1/insmod-modutils/util/logger.c: rename log to do_log in order + to avoid clashes with builtins + + * mdk-stage1/dietlibc/ia64/start.S, mdk-stage1/dietlibc/ia64/fork.S, + mdk-stage1/dietlibc/ia64/pipe.S: ia64 fixes: fork, pipe, syscall weak + symnames + + * mdk-stage1/insmod-modutils/include/elf_ppc64.h, + mdk-stage1/insmod-modutils/obj/obj_ppc64.c: merge ppc64 support from + modutils 2.4.26 + + * mdk-stage1/insmod-modutils/obj/obj_gpl_license.c: add missing file from + older merge from modutils 2.4.26 + + * mdk-stage1/dietlibc/ia64/syscalls.h: add ia64 setjmp/longjmp from glibc + ia64 fixes: fork, pipe, syscall weak symnames + + * tools/ddcprobe/Makefile: ppc & ppc64 fixes + ppc64 support + + * mdk-stage1/insmod-modutils/obj/obj_kallsyms.c: fix build to expose + assignments better + + * mdk-stage1/insmod-modutils/obj/obj_ia64.c: add support for + R_IA64_PCREL60B reloc + build fix + + * tools/ddcprobe/ddcxinfos.c: ppc & ppc64 fixes + sounds like the (empty) line is expected + + * mdk-stage1/insmod-modutils/util/sys_oim.c, + mdk-stage1/insmod-modutils/insmod.c, + mdk-stage1/insmod-modutils/util/Makefile, + mdk-stage1/insmod-modutils/util/modstat.c: drop support for kernel 2.0 + + * tools/ddcprobe/vbe.h: ppc & ppc64 fixes + fix on big endian platforms (ppc, ppc64) + + * mdk-stage1/dietlibc/ia64/__longjmp.S, mdk-stage1/dietlibc/ia64/setjmp.S: + add ia64 setjmp/longjmp from glibc + + * mdk-stage1/rescue-gui.c, mdk-stage1/init.c: lib64 fixes on ppc64 + + * mdk-stage1/dietlibc/include/elf.h: ia64 elf definitions + + * tools/ddcprobe/of.c: ppc & ppc64 fixes + On PowerPC platforms, byteswap product_code so that ids match MonitorsDB + contents + +2004/12/15 Pixel + + * mdk-stage1/network.c: fix http directory not starting with a "/" (it was + handled for stage1, but not for URLPREFIX given to stage2) + +2004/12/15 Stew Benedict + + * standalone/drakTermServ: perl_checker, suggestions from Pixel, rework + buttons in main GUI + +2004/12/15 Thierry Vignaud + + * share/po/fr.po: update (stef, neoclust) + + * drakxtools.spec: 10.2-0.6mdk + +2004/12/15 Warly + + * share/rpmsrate.corpo-server, share/rpmsrate.corpo-desktop: remove + mdkonline-backend + +2004/12/14 Gwenole Beauchesne + + * mdk-stage1/dietlibc/include/termios.h, mdk-stage1/dietlibc/diet.c, + mdk-stage1/dietlibc/syscalls.h, mdk-stage1/dietlibc/include/sys/stat.h, + mdk-stage1/dietlibc/include/sys/ucontext.h, + mdk-stage1/dietlibc/include/sys/shm.h, + mdk-stage1/dietlibc/include/fcntl.h, + mdk-stage1/dietlibc/include/sys/mman.h, + mdk-stage1/dietlibc/include/asm/sigcontext.h, + mdk-stage1/dietlibc/include/elf.h, + mdk-stage1/dietlibc/include/sys/ioctl.h, + mdk-stage1/dietlibc/include/sys/types.h, mdk-stage1/dietlibc/Makefile, + mdk-stage1/dietlibc/include/endian.h, + mdk-stage1/dietlibc/include/sys/ptrace.h, + mdk-stage1/dietlibc/include/signal.h: ppc64 port + + * mdk-stage1/dietlibc/include/setjmp.h: fix setjmp.h (JB_SIZE) on ppc64 + ppc64 port + +2004/12/14 Pixel + + * drakxtools.spec: fix release (keep 101mdk for 10.1 packages) + + * network/netconnect.pm: for drakx-finish-install, we want drakconnect to + restart network + for drakx-finish-install, we want drakconnect to restart network + + * mygtk2.pm: export gtkval_register() and gtkval_modify() + + * mdk-stage1/pcmcia_/merge_from_pcitable: cleanup + +2004/12/13 Keld Jørn Simonsen + + * share/po/da.po: updates + soft/drakpxelinux/po/da.po soft/rpmdrake/po/da.po + soft/urpmi/po/da.po gi/perl-install/share/po/da.po + +2004/12/13 Pixel + + * detect_devices.pm: fix + + * drakxtools.spec: introduce in CVS 27.2.101mdk which was using a patch on + drakbug + + * standalone/drakbug: commit vdanen patch which makes drakbug send + everything to bugzilla (instead of anthill) + +2004/12/13 Rafael Garcia-Suarez + + * share/list: Add a gazillion modules required by the new packdrake for + now + 2004/12/13 Pixel * detect_devices.pm: fix @@ -42599,7 +42972,7 @@ * Makefile.config: remove dir security - * proxy.pm: add $Id: ChangeLog,v 1.993 2004/12/13 14:24:31 prigaux Exp $ + * proxy.pm: add $Id: ChangeLog,v 1.994 2004/12/13 17:47:19 prigaux Exp $ * interactive/gtk.pm: use my_gtk {isWizard} and {isEmbedded} which are more accurate than -- cgit v1.2.1