intended to fix a problem reported by tb@: pkg-readmes docs used to have
the FULLPKGNAME encoded, which makes for gratuitous differences from version
to version.
- use PKGSTEM inside the README for less variation
- ditch the FULLPKGNAME as the install location, PKGSTEM is enough.
PKGSTEM ?= is overrideable, because stuff like php branches need to install
several readmes.
Next commit will fix all affected ports: all readmes change location, so lots
of plist tweaks and bumps
- move the logic for choosing junk into add_tasks
- create specific finalize hooks at end of extract and configure
so that nojunk only starts at end of extract, thus gaining a few
(precious) minutes for big nojunk ports
- introduce noconfigurejunk logic, so that a port may configure without
junking opening. This should fix sporadic bugs in automake-based ports
without locking junking for the full build (roughly 200 ports are
affected)
Job::Port and Job::Fetch, do not kill_on_timeout stuff that's not really stuck
(e.g., waiting for locks and checksumming).
You can now put fairly agressive timeouts on fetch, since checksum is out
of the way.
stop displaying the 100% message and frozen, you could think it's frozen
"near" 100%. Instead display "cksum..." and get rid of frozen indications
entirely
- they don't makesum
- they fetch the final name in case of renames.
Likewise in dpb:
- grab MASTER_SITE_BACKUP as well
- do a subclass that uses the correct filename
That way, we get stuff from MASTER_SITE_BACKUP correctly
option usage.
- final tweak to directory names, make sure they end with a /
a double / is not a problem for pkg_create, and it makes things
more readable for puny humans.
- add an assertion for unsubst for debugging (commented out)
- chomp the unsubst string so that the assertion works for full matches
- document where to add the new stuff
it might make sense to have this a bit wider, but currently the
marker is added in quite a few places where it doesn't help.
Expand the match to include truetype/opentype collections and
pfa files. ok espie
- use context in prefix/suffix substitution
specifically, if we have keywords, explicitly allow for them in prefix
Allow for substitution "anywhere" in actions
And allow for manpage extension for suffixes
- never propose a substitution for TRUEPREFIX, it shouldn't even be there
(still need PREFIX vs LOCALBASE desambiguation)
- create the sorted variables list in finalize, so we can add more meat
- allow for MACHINE_ARCH-openbsd (perl)
- be more specific for new substs, we allow substs if the value just appeared
- do not count variables we don't backsubst as duplicate value, this should
solve DISTNAME vs FULLPKGNAME
