Age | Commit message (Collapse) | Author | |
---|---|---|---|
2010-12-09 | correctly return wait_for exitcode, and use it to die_error if needed | Dieter Plaetinck | |
2010-12-09 | make run_controlled actually return the right exitcode | Dieter Plaetinck | |
this fixed exit code feedback for package installation and mkinitcpio | |||
2010-12-09 | properly pass on $BACKGROUND_EXIT so that $CONTROLLED_EXIT is set | Dieter Plaetinck | |
2010-12-09 | lift no-hyphen restriction on run_{controlled,background} identifiers | Dieter Plaetinck | |
Since 44bc9f54734d9280c71959e56189044141bc1640 this is no longer an issue | |||
2010-12-09 | fix correct exitcode variables in run_{controlled,background} commands | Dieter Plaetinck | |
Since 44bc9f54734d9280c71959e56189044141bc1640, the variablenames for the exit code was still incorrect in the debug logmessage and a doc header Also added a debug line for run_controlled | |||
2010-04-16 | change infofy calls to inform, because libui-sh changed function name | Dieter Plaetinck | |
2010-04-05 | make sure to create the parent directories for the logfiles in lib-misc ↵ | Dieter Plaetinck | |
run_background and run_controlled. this caused breakage now since libui logs to another directory | |||
2010-04-04 | simplify check_is_in function | Dieter Plaetinck | |
2009-11-11 | variable assigment syntax error | Pavel Hofman | |
2009-08-01 | make run background/controlled exitcodes more robust & bugfree | Dieter Plaetinck | |
2009-06-29 | Modified timesetting procedure again | Gerhard Brauer | |
With our current code we get problems when the user want to set datetime manually in our dialog. Revisiting last changes i see: it's a bad thing to force any timesetting based on a *selected* timezone in an still UTC environment. So the easiest way (and i hope the last change on this!) is: a) user selected a timezone b) we set this timezonefile to /etc/localtime c) Now we ask: is your hwclock on UTC or localtime? c) UTC BIOSclock user now have both correct system and hw time. d) For localtime users we must re-read the hwclock into our new timezone, cause copy'ing above timezone file has recalculated the systemtime to a false time. I've tested these commit on my PC, both with hwclock on UTC and localtime, also with manual setting and via ntp(date) in the aif dialog. Signed-off-by: Gerhard Brauer <gerbra@archlinux.de> | |||
2009-06-28 | fixes for font/keymap substitutions | Dieter Plaetinck | |
2009-06-25 | Fixed timesetting for hwclock localtime users | Gerhard Brauer | |
When we set system time from the BIOS clock via hwclock we must honor the timezone the user previous selected during interactive_time(). After boot we are in an UTC environment. If we now - for localtime users - parse hwclcock value to systemtime, then we get later a false systemtime when copying the TZ file to /etc/localtime. Changing /etc/localtime always recalculate the time according to the timezone. Signed-off-by: Gerhard Brauer <gerbra@archlinux.de> | |||
2009-03-30 | fix hwclock calls for systohc and hctosys, fix /etc/localtime, show nice ↵ | Dieter Plaetinck | |
dialog with 'ok', 'manual date setting' and ntp using | |||
2009-03-30 | do hwclock (only) at the right time, optimize hwclock code, make UTC default ↵ | Dieter Plaetinck | |
(fixes FS#13752 - Make hwclock in UTC a "good idea" | |||
2009-03-17 | slight refactoring of how to run/monitor a program. i quite broke it though | Dieter Plaetinck | |
2009-03-13 | fix for kill dialogs forked process from the follow_progress call in wait_for | Dieter Plaetinck | |
2009-03-12 | support for storing the pid of background wrapper processes + support for ↵ | Dieter Plaetinck | |
non-blocking following of text files (eg tail --pid) | |||
2009-03-12 | various fixes re:keymap/font. + auto updating rc.conf in all procedures | Dieter Plaetinck | |
2009-03-10 | fix segfault caused by loop | Dieter Plaetinck | |
2009-02-28 | fix for bash read: not a valid identifier, and hence, got rid of the pacman ↵ | Dieter Plaetinck | |
install pkg workaround (finally!!) | |||
2009-02-28 | debug fix in check_is_in | Dieter Plaetinck | |
2009-02-28 | check_is_in fixes.. and this time i mean it.. i think. hopefully. + test ↵ | Dieter Plaetinck | |
script for check_is_in | |||
2009-02-28 | support for debug categories | Dieter Plaetinck | |
2009-02-25 | adding refactored/reordered date/time setting logic into aif. based on ↵ | Dieter Plaetinck | |
tpowa's tz script. this also means an ncurses implementation of timezone selection | |||
2009-02-22 | since we use /tmp now, no more need to hide files. some categorizing ↵ | Dieter Plaetinck | |
(tagging) of files is useful though | |||
2009-02-14 | tried to make aif FHS/hier compliant + updated howto/readme | Dieter Plaetinck | |
2008-12-31 | runfile should be hidden | Dieter Plaetinck | |
2008-12-20 | fix for incorrect shebangs | Dieter Plaetinck | |
2008-12-12 | cleanup stale files on start and stop + long overdue introduction of ↵ | Dieter Plaetinck | |
$RUNTIME_DIR | |||
2008-12-12 | check_is_in fix | Dieter Plaetinck | |
2008-12-12 | implementation if iterative, dependency-resolving underlying volume ↵ | Dieter Plaetinck | |
filesystem creater | |||
2008-12-12 | todo updates | Dieter Plaetinck | |
2008-11-16 | updated HOWTO for install using packages, got rid of patch-install-cd script ↵ | Dieter Plaetinck | |
+ renamed fifa into aif everywhere | |||
2008-11-13 | fixes to run_background | Dieter Plaetinck | |
2008-11-13 | dont overwrite previous logfile, rather append | Dieter Plaetinck | |
2008-11-13 | small stuff | Dieter Plaetinck | |
2008-11-13 | debugging info in run_background + better logfile location for pacman | Dieter Plaetinck | |
2008-11-12 | typo fix and todo updates (important one!) | Dieter Plaetinck | |
2008-11-12 | fix regression after wait_for running process refactoring | Dieter Plaetinck | |
2008-11-12 | cleaner handling of exit code and running state in run_background + ↵ | Dieter Plaetinck | |
installpkg exit code fix | |||
2008-11-12 | meh | Dieter Plaetinck | |
2008-11-03 | 2 bugfixes | Dieter Plaetinck | |
2008-11-02 | file hierarchy overhaul. now a structure that makes more sense: user and ↵ | Dieter Plaetinck | |
core separated. we now differentiate between procedures and libraries - no more "profiles" - + some bug fixed + probably quite a few introduced + runtime directory + separated my own stuff more |