fa2c4bcc52
* Add some munging on the Release files if we're not building
...
for "testing". Used in the etch release, and I'll back it out shortly.
2007-04-07 11:34:51 +00:00
ab7e1d0171
In list2cds, don't add udebs in the final COMPLETE run - they're no
...
use there.
2007-04-06 18:23:14 +00:00
b695b182c4
* In start_new_disc, cope with $DEBVERSION containing spaces
...
* Only put the release notes and installation guide on disc#1
2007-04-04 16:40:18 +00:00
64bd969a0c
* Fix up boot-alpha to calculate the size of the ramdisk needed, rather
...
than hard-coding to 16384. Also slight refactoring.
2007-03-20 01:01:16 +00:00
a2d50b9f65
* Fix a silly bug in start_new_disc that would break on discs with
...
local packages. Thanks to Ronny Aasen for the patch. Closes : #414818
* Final *final* upload for etch; urgency high to get through the
freeze.
2007-03-13 23:38:53 +00:00
1fc78c6086
* Fix the extra_image() function in both boot-i386 and boot-amd64 to make
...
sure we don't end up overwriting extra images from one arch with those
from the other. Now the multi-arch netinst should work for the installgui
target for both arches
2007-03-13 02:21:13 +00:00
9006e6dd31
* Add the DEFAULT64 handling to boot-amd64 as well as boot-i386; needed
...
in both so that things are setup the same regardless of the order in
which the boot-$arch scripts are called
2007-03-13 01:06:07 +00:00
ec87877896
Fix path for creating splash image in boot-amd64. Closes : #413850 .
2007-03-12 01:33:39 +00:00
46369d0dc5
Disable pattern metacharacters in some regex with included variables.
...
Closes : #413506
2007-03-05 16:09:46 +00:00
7b1b3973f8
* Updated isolinux.bin to the version from syslinux 3.31-3.
...
* Re-enable use of 64 bit detection, fixed now.
2007-03-01 19:18:45 +00:00
2ca98a7f32
* Cope with rolling back more than one package in case we go well
...
over the disc size
* Use a multiplier of 1.2 for HFS space - overkill, but it fixes
the multi-arch DVD overflowing. Also log if we're doing HFS.
* Tweak where we look for s390 boot files to make the weeklies work again.
2007-02-27 20:05:32 +00:00
a2d9587f1a
Don't rely on PATH containing the tools directory.
2007-02-21 14:13:35 +00:00
f47d23be43
* Temporarily disable use of 64 bit detection due to syslinx's bug #410729 .
2007-02-13 01:14:35 +00:00
3d5259c5ef
* Use the 64 bit autodetection feature of syslinux 3.31-2, if available,
...
to default to the 64 bit target in multiarch isos.
2007-02-04 23:41:44 +00:00
06efc9c4d7
Add debug for when get_file_blocks fails
2007-01-22 23:06:30 +00:00
ba8e154a12
* In start_new_disc, make sure that we call the (badly-named) add-bin-doc
...
so that we get README.html and README.txt created even for source discs.
2007-01-02 01:56:44 +00:00
55f4db876a
+ Slightly change the output naming of the disc images to make it safer
...
to produce more then one type of multi-arch disc
2006-12-30 15:13:07 +00:00
c7ece8748b
* Added an update-tasks target to the Makefile and moved the code out of
...
file header comments to there. Fix some bugs in passing.
* List packages for the various specific desktop tasks ahead of the non-key
packages for the general desktop task. Along with changes in tasksel 2.60
this should lead to a generally more useful set of packages on the desktop
CDs; instead of filling up with OOo, it will fill up with things specific
to the desktop environment.
* Updated task lists.
2006-12-29 06:05:30 +00:00
9d376d233e
Re-indent add_packages
2006-12-29 05:57:21 +00:00
17bb18194f
* Re-implemented exclude/unexclude support in make_disc_trees.pl instead
...
of in list2cds. Slight changes in semantics - the new code supports
regular expressions for matching package names (to make it easier to
exclude groups of related packages), and EXCLUDE<m> should be used in
preference to EXCLUDE for packages that are just being shifted to
later in the set by using UNEXCLUDE<n>.
2006-12-29 05:56:43 +00:00
4d4bba1b7e
Don't list the md5sum.txt file inside itself, it will always fail
2006-12-29 02:50:35 +00:00
35743cde5f
As a last-minute step before building images, fill in the TOTALNUM
...
value in the README files in the root of each image.
2006-12-29 02:44:22 +00:00
cccbdd469f
Gunzip the installation manual in .txt and .pdf format, so it's useful
...
on other platforms too.
2006-12-29 01:48:26 +00:00
7af2565ecf
* Add linux-image-2.6-prep for powerpc.
2006-12-29 01:14:28 +00:00
eb1131b3df
Make make_disc_trees.pl and start_new_disc fail better on errors
2006-12-23 03:03:58 +00:00
528c0072f2
remove Suggests on delo - we grab and use the correct version direct
...
from the archive instead these days
2006-12-23 03:02:37 +00:00
342bcac67b
don't suggest binutils-multiarch, we haven't needed that for mipsel CDs since woody
2006-12-23 01:17:47 +00:00
37a2704f3b
Remove scansources ; no longer used
2006-12-23 00:55:48 +00:00
9ae3366b88
Lose the "|| true" when we call boot-$arch; make sure we fail if it fails...
2006-12-23 00:00:36 +00:00
82aa893404
Minor update to boot-sparc
2006-12-22 14:24:28 +00:00
5f185be61f
Removed old "set -x"
2006-12-21 16:19:34 +00:00
4a2a622525
Make the boot scripts more consistent in how they call add_mkisofs_opt
2006-12-21 16:05:32 +00:00
67d9909bbe
Make sure the mkisofs options file goes in the right place
2006-12-21 15:27:50 +00:00
77e3c9015c
Add extra logic to source-tracking code. If we have more than 1 binary
...
arch along with source, there's a chance that the different arches may
have refs to a particular binary package (and hence its source) at
completely different points. Rather than just go with the one we find
first (from the first binary ref), make sure that the source will
match up with the earliest binary record, measured across all arches.
As noticed on the multi-arch DVD; make sure that source:kde-i18n makes
it into the package list before amd64:kde-i18n-ar.
2006-12-21 03:45:01 +00:00
74e9050525
rearrange to make sure is set before we set
2006-12-20 11:02:59 +00:00
4bb6ee2bbe
Always create an empty mkisofs_dirs file for each image; stop warnings about missing files later when we source them in
2006-12-19 17:15:16 +00:00
61c1630ab1
Cosmetic fix - list source in too in debug
2006-12-19 15:57:15 +00:00
b600a32f13
* Update documentation to at least vaguely match how things work.
...
* Added support for hook scripts in a few places around the CD
creation process, to help people customise their CDs. Quite a
lot of refactoring needed in make_disc_trees.pl needed to do that.
* Removed more cruft:
+ old defs in CONF.sh
+ old defs and rules in the Makefile
+ tools/add_secured
2006-12-19 01:58:41 +00:00
7b6988d66d
Make list2cds more verbose in log output, especially when dependencies have failed so a package can't be added
2006-12-18 13:40:53 +00:00
05905f9e37
Better estimation for HFS media (m68k/ppc)
2006-12-12 02:10:53 +00:00
95ffabb409
* Don't put ppc boot files on discs >1.
...
* Add more hfs options to m68k discs, to match ppc.
2006-12-12 02:09:34 +00:00
ad1906c27b
Even more accurate size estimation; reduce how much we need to run mkisofs
2006-12-11 17:52:25 +00:00
8b318a9693
Tweak the size-checking threshold to make things faster
2006-12-11 15:50:49 +00:00
277fc58f3d
When rolling back, cope with package names that are not regexp friendly
2006-12-11 15:47:18 +00:00
e1116e39e4
typo fix
2006-12-10 22:37:38 +00:00
96608804fc
Remove the old shell version of make_disc_trees
...
Missed changing one of the invocations of add_packages - now fixed
2006-12-10 22:36:08 +00:00
52d0a9f08b
Round off the DVD size in blocks
2006-12-10 22:23:51 +00:00
cf079be95b
Re-add the build date to .disk/info, lost in the merge
2006-12-10 21:27:06 +00:00
5b26423869
Minor tweaks for businesscard CDs
2006-12-10 20:12:46 +00:00
7780369fd4
* When we've finished making a CD tree in make_disc_trees.pl,
...
re-compress the gzipped Sources and Packages. Workaround for bug
#402482 in busybox gunzip code.
2006-12-10 20:00:37 +00:00
Steve McIntyre
Raphaël Hertzog
Joey Hess
Frans Pop