src - OpenBSD base system

Age	Commit message (Collapse)	Author
2009-12-13	force gzip to write package even if things don't compress.	Marc Espie

2009-12-12	update will ask about conflicts, so we need that in updateinfo...	Marc Espie

2009-12-12	autovivified	Marc Espie

2009-12-12	fix the fam->libgamin issue. Look for update candidates in self before	Marc Espie
	asking the repository. This solves this kind of conflicts. also, register packages we keep, this makes for simpler scenarios (todo: split updateset into further categories)
2009-12-12	have pkg_add recheck there are no internal conflicts in an updateset.	Marc Espie
	provides a different failure for the fam/gamin update (better one actually, since we go further...)
2009-12-12	remove everything from to_update, including hints	Marc Espie

2009-12-12	tie conflict lists to packing lists and use them	Marc Espie

2009-12-12	do the same dance with older packages: build before list, and use it when	Marc Espie
	deleting.
2009-12-11	record internal updateset dependencies for new packages, and respect them	Marc Espie
	while installing (need doing something similar for old packages)
2009-12-07	don't use $set->print, consistent messages.	Marc Espie

2009-12-07	simplify tracker a bit now we know what we're doing.	Marc Espie
	kill code that never got used
2009-12-06	-U before -u;	Jason McIntyre

2009-12-05	a bit more documentation, including revisit of OpenBSD::Intro to reflect	Marc Espie
	the current state of things.
2009-12-05	use eval {} everywhere around quirks.	Marc Espie
	also, remove quirks from the list of packages to update, since it deserves special treatment...
2009-12-05	since we don't redisplay missing libraries message, tell why we can't install	Marc Espie
	in general terms...
2009-12-05	quirks exists	Marc Espie

2009-12-05	avoid progressmeter warning	Marc Espie

2009-12-05	better quirks interface: pass the search object that was built, so we can	Marc Espie
	adjust it (eventually, pass the whole list ?)
2009-12-05	change the internal representation of Search::Stem to allow the addition	Marc Espie
	of supplementary stems (useful for quirks)
2009-12-05	oops	Marc Espie

2009-12-05	cosmetic	Marc Espie

2009-12-05	fix quirks object construction. Simplify its handling: just install/update	Marc Espie
	it first, and try to build the Quirks object. Make ourselves resilient to errors in there. For removing sets with only older packages, we have to call complete_old even if there are no conflicts at all !
2009-12-05	make quirks silent. Track # of updates found, instead of newer size, since	Marc Espie
	we may end up updating sets with newer == 0
2009-12-05	fix off-by-one, noticed by frantisek holop	Marc Espie

2009-12-04	create new updatesets when needed	Marc Espie

2009-12-04	fix the case where you look to update one package: the dependencies of the	Marc Espie
	new package must first be updated too.
2009-12-04	set-up for quirks, just don't do it yet.	Marc Espie

2009-12-04	simpler unified lookup for libs inside set/installed packages.	Marc Espie

2009-12-04	internal API: is_known, for pkgnames that are known by the tracker,	Marc Espie
	so that we decide whether to create new updatesets. (not efficient, but isolate from rest of code before Tracker refactor)
2009-12-03	fix typos and blanks	Marc Espie

2009-12-03	zap useless debug code	Marc Espie

2009-12-03	register plists from newer elements, in case they're useful.	Marc Espie

2009-12-02	fix some inter-dependencies case:	Marc Espie
	- when we merge updatesets, invalidate the solver (there isn't any, usually). - look inside set for check_depends. - do register_dependencies after committing all plists to cache.
2009-12-01	document more or less what's going on now.	Marc Espie
	With updatesets, technical details are ways more complicated.
2009-12-01	okay, this one needs tagging as well	Marc Espie

2009-12-01	for every case we can't install stuff, we have to let the tracker now, and	Marc Espie
	close the set.
2009-11-30	tweak the rules a bit: the search filter for "more recent than" ought to	Marc Espie
	notice it sees older stuff. Even if they get filtered later on, this would turn into a "no need to update". (in actual use, I probably will need to open those packages and look at the pkgpaths to make sure they are not valid candidates, and to compare full signatures... oh well)
2009-11-30	better errors: have $set->cleanup mark set as finished.	Marc Espie
	in case there is an issue, register set as not updateable with tracker.
2009-11-30	better: have can_install return 0/1 as definitive answer, and return "undef"	Marc Espie
	for try again later. Don't bother cleaning up set from can_install, since complete does it too.
2009-11-29	some major changes have happened	Marc Espie

2009-11-29	compute updates on the fly, instead of "at start"	Marc Espie

2009-11-29	set things up for updates on the fly. In particular, we may have to look	Marc Espie
	into "to_update" stuff...
2009-11-29	in the worst case, we find a new handle with the right pkgpath,	Marc Espie
	independently of the name...
2009-11-29	show old->new as everywhere else in sets.	Marc Espie
	micro-optimisation: if old pkgname is the same as new pkgname, we just need to build the RequiredBy file, no need to adjust Requiring...
2009-11-29	clean up conflict handling and dependency adjustement: build list	Marc Espie
	of conflict globally, add it as old handles, and then use it everywhere.
2009-11-29	show missing libraries as "\|detail" for the can't install message.	Marc Espie

2009-11-29	slightly simpler error handling	Marc Espie

2009-11-29	have SharedLibs use state to print error messages.	Marc Espie

2009-11-29	handle more general conflicts:	Marc Espie
	- match old spec against any of the new packages in the set. - manage conflicts globally if several newer packages are there. - remove the limit of 5 conflicting packages, it's no longer really relevant.
2009-11-28	oops	Marc Espie