[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Packages failing to be removed from testing [take 2]

Revised analysis of packages trying to be removed from testing that just
won't go:

> trying: -cl-uncommonsql
> skipped: -cl-uncommonsql (300 <- 38)
>     got: 188+0: a-41:a-42:h-41:i-64
>     * i386: cl-local-time-db

Package: cl-uncommonsql
Unstable: 1.2.1-1
Suggestion: don't-remove-from-testing
Analysis: in addition to cl-local-time-db, cl-metadata also has a
 dependency on both of the preceding packages.  The original RC bug
 against cl-uncommonsql, 186521, has since been fixed; there's no longer
 any reason to try to kick this package (and its dependents) out of
 testing, as the updated version should trivially make its way in once
 this block is removed.

> trying: -curl-ssl
> skipped: -curl-ssl (301 <- 37)
>     got: 384+0: a-41:a-42:h-41:i-63:i-41:m-43:m-37:m-36:p-40
>     * i386: php4-curl
>     * powerpc: php4-curl

Package: curl-ssl
Testing: 7.9.5-2
Unstable: none
Suggestion: yank-it-out-by-the-roots
Analysis: PHP is already a mess in testing, and this package needs to go
 -- one more broken dependency won't hurt anything.  Or if it's easier,
 it can languish a bit longer until php4 4.3 makes it in.

> trying: -eel
> skipped: -eel (302 <- 36)
>     got: 44+0: a-44
>     * alpha: galeon-nautilus, libnautilus-dev, nautilus

Suggestion: wait-and-see
Analysis: package superseded by eel2, and has already been removed from
 unstable; just waiting on galeon and nautilus to be up-to-date in

> trying: -gimp1.2
> skipped: -gimp1.2 (303 <- 35)
>     got: 50+0: a-50
>     * alpha: exult-studio, gimp-dimage-color, gimp-python, gimp1.2-quiteinsane, gtkam-gimp, mtink, sane, xpcd-gimp, xsane

Suggestion: wait-and-see
Analysis: Needs removing, but needs the new gimp package to be ready for
 testing first to satisfy all the dependencies on libgimp1.2; and that's
 held up by perl right now, which is waiting on mips{,el} binaries
 (last build attempt on both archs failed).

> trying: -gnome-core
> skipped: -gnome-core (304 <- 34)
>     got: 48+0: a-48
>     * alpha: galeon-nautilus, gnome-admin, gnome-media, ickle-gnome, libpanel-applet-ruby, nautilus, seti-applet

Suggestion: wait-and-see
Analysis: Replaced by meta-gnome2, which is also not yet ready for
 testing; it depends on gnucash, which has known serious portability
 problems.  This could probably use some help from someone of a
 scheme-ish bent.  There seems to be a sentiment that the FTBFS archs
 for gnucash should be dropped from the supported list, but there's
 reason to believe the bugs affect *all* archs even if the PRNG-using
 tests squeak by on some of them.

> trying: -gnome-guile
> skipped: -gnome-guile (305 <- 33)
>     got: 42+0: a-42
>     * alpha: gwave

Package: gwave
Testing: 20020122-1
Unstable: 20020122-1
Bugs: 187829
Suggestion: remove-from-testing
Analysis: Package depended on libguilegtk0, which has been removed from
 unstable and is therefore unmaintained in testing.  The gwave package
 has a 5-month-old grave bug open about this issue, with no sign of

> trying: -id3lib
> skipped: -id3lib (306 <- 32)
>     got: 42+0: a-42
>     * alpha: grip

Suggestion: wait-and-see
Analysis: I previously reported that grip had been rebuilt against
 the new library, but this turns out to *only* be the case on i386.  Bug
 212793 has been filed about this issue.

> trying: -java-virtual-machine-dummy
> skipped: -java-virtual-machine-dummy (307 <- 31)
>     got: 190+0: a-41:a-42:h-41:i-66
>     * i386: jmeter, tomcat4, tomcat4-webapps

Package: java-virtual-machine-dummy
Testing: 0.3
Unstable: none
Suggestion: yank-it-out-by-the-roots
Analysis: These three packages are all in contrib, and the dependency
 should therefore be ignorable. (java-virtual-machine-dummy is also the
 last in a list of dependency-satisfying alternatives for the packages.)

> trying: -jsboard
> skipped: -jsboard (308 <- 30)
>     got: 194+0: a-41:a-42:h-41:i-70
>     * i386: jsboard-theme-aicom-ko, jsboard-theme-debian-ko, jsboard-theme-diary-en, jsboard-theme-diary-ko, jsboard-theme-trash-en, jsboard-theme-trash-ko, jsboard-theme-wizz-ko

Package: jsboard-theme-aicom-ko, jsboard-theme-debian-ko, jsboard-theme-diary-en, jsboard-theme-diary-ko, jsboard-theme-trash-en, jsboard-theme-trash-ko, jsboard-theme-wizz-ko
Testing: 0.5-1
Unstable: 0.5-1
Suggestion: remove-from-testing
Analysis: Depend on a package we want to remove; these packages are not
 useful without jsboard itself.

> trying: -kernel-image-2.2.20-i386
> skipped: -kernel-image-2.2.20-i386 (309 <- 29)
>     got: 190+0: a-41:a-42:h-41:i-66
>     * i386: pcmcia-modules-2.2.20, pcmcia-modules-2.2.20-compact, pcmcia-modules-2.2.20-idepci

Suggestion: wait-and-see
Analysis: Per Joey Hess, these pcmcia modules packages are no longer
 useful; and indeed, the current version of pcmcia-cs doesn't include
 these binary packages.  Should be in sync on all archs within a few
 days, at which point kernel-image-2.2.20-i386 will drop out on its

> trying: -kernel-image-2.4.20-3-i386
> skipped: -kernel-image-2.4.20-3-i386 (310 <- 28)
>     got: 208+0: a-41:a-42:h-41:i-84
>     * i386: alsa-modules-2.4.20-3-386, alsa-modules-2.4.20-3-586tsc, alsa-modules-2.4.20-3-686, alsa-modules-2.4.20-3-686-smp, alsa-modules-2.4.20-3-k6, alsa-modules-2.4.20-3-k7, alsa-modules-2.4.20-3-k7-smp, i2c-2.4.20-3-386, i2c-2.4.20-3-586tsc, i2c-2.4.20-3-686, i2c-2.4.20-3-686-smp, i2c-2.4.20-3-k6, i2c-2.4.20-3-k7, i2c-2.4.20-3-k7-smp, linux-wlan-ng-modules-2.4.20-3-386, linux-wlan-ng-modules-2.4.20-3-586tsc, linux-wlan-ng-modules-2.4.20-3-686, linux-wlan-ng-modules-2.4.20-3-686-smp, linux-wlan-ng-modules-2.4.20-3-k6, linux-wlan-ng-modules-2.4.20-3-k7, linux-wlan-ng-modules-2.4.20-3-k7-smp

No one has indicated that the alsa, i2c, and linux-wlan-ng modules
should also be disposed of.  The alsa-modules-i386 package, at least,
has been updated in sid to match the available kernels, and is now
merely in testing purgatory.  i2c has been updated in unstable, but is
already out-of-date again wrt the kernels it supports.

> trying: -kernel-source-2.2.20
> skipped: -kernel-source-2.2.20 (311 <- 27)
>     got: 188+0: a-41:a-42:h-41:i-64
>     * i386: kernel-patch-2.2.20-powerpc

Package: kernel-patch-2.2.20-powerpc
Testing: 2.2.20-3
Unstable: 2.2.20-3
Suggestion: remove-from-testing, remove-from-unstable
Analysis: Bug 212953 has already been filed requesting removal of this
 package from the archive, since it's no longer needed for the powerpc

> trying: -kgeo
> skipped: -kgeo (312 <- 26)
>     got: 188+0: a-41:a-42:h-41:i-64
>     * i386: junior-kde
> trying: -kstars
> skipped: -kstars (313 <- 25)
>     got: 188+0: a-41:a-42:h-41:i-64
>     * i386: junior-kde

Package: junior-kde
Testing: 1.1
Unstable: 1.1
Suggestion: remove-from-testing
Analysis: The new kdeedu package that provides both kgeo and kstars is
 out of sync in unstable, and has been since Jun 5.  In the meantime,
 the old versions of these packages are helping to hold up the removal
 of various old KDE libs as well.  The junior-kde package can be removed
 from testing to clean this up, or we can wait for kdeedu to be
 installable in testing.

> trying: -libapache-mod-perl
> skipped: -libapache-mod-perl (314 <- 24)
>     got: 46+0: a-46
>     * alpha: axkit, libapache-authensmb, libapache-db-perl, libapache-request-perl, logtrend-visuapache

Suggestion: wait-and-see
Analysis: In unstable, libapache-mod-perl comes from the apache source
 package, which currently has two RC bugs (209276, 207453) that are
 being worked on.

> trying: -libmime-base64-perl
> skipped: -libmime-base64-perl (315 <- 23)
>     got: 42+0: a-42
>     * alpha: sympa

Suggestion: wait-and-see
Analysis: sympa is out of date on hppa due to the glibc problems there.
 There is an RC bug on sympa which may or may not still be applicable --
 and may or may not even affect the version currently in testing.  Have
 followed up to the bug, but otherwise think this is a wait-and-see for
 the moment.

> trying: -librsvg
> skipped: -librsvg (316 <- 22)
>     got: 46+0: a-46
>     * alpha: galeon-nautilus, libeel-dev, libeel0, libnautilus-dev, nautilus

Suggestion: wait-and-see
Analysis: Just as eel is superseded by eel2, this package has been
 superseded by librsvg2, and is waiting for the same reason.

> trying: -libweakref-perl
> skipped: -libweakref-perl (317 <- 21)
>     got: 193+0: a-41:a-42:h-41:i-69
>     * i386: libxml-twig-perl, libxmltv-perl, libxtm-perl, xmltv, xmltv-gui, xmltv-util

Suggestion: wait-and-see
Analysis: Removed from unstable, can't be removed from testing until a
 few more packages are brought up to date (e.g., perl).

> trying: -medusa
> skipped: -medusa (318 <- 20)
>     got: 45+0: a-45
>     * alpha: galeon-nautilus, libnautilus-dev, libnautilus0, nautilus

Suggestion: wait-and-see
Analysis: Removed from unstable, caught up in the whole nautilus chain.

> trying: -orp
> skipped: -orp (319 <- 19)
>     got: 188+0: a-41:a-42:h-41:i-64
>     * i386: orp-classpath

Package: orp-classpath
Testing: 1:0.02.1-3
Unstable: 1:0.02.1-3
Suggestion: remove-from-testing
Analysis:  When requesting orp's removal, Colin wrote, "Note that
 orp-classpath depends on this, but is obviously useless itself with a
 broken orp."

> trying: -pd
> skipped: -pd (320 <- 18)
>     got: 42+0: a-42
>     * alpha: gem

Suggestion: wait-and-see
Analysis: Package renamed to 'puredata' in unstable, and should
 propagate into testing soon; gem, however, is being held out of testing
 by a FTBFS on alpha (and possibly another one on hppa, though the hppa
 buildd doesn't seem to have attempted the latest version of this
 package).  Bug #212003 has been filed for the alpha failure.

> trying: -php4
> skipped: -php4 (321 <- 17)
>     got: 47+0: a-47
>     * alpha: apl, php-cgiwrap, php-imlib, php4-apc, php4-dbase, php4-dbtcp, php4-gd2, php4-mcrypt, snui-server

Package: php4
Suggestion: don't-remove-from-testing
Analysis: Obviously, php4 is resisting removal.  Given that it's not
 coming out, the original motivation for trying to remove it no longer
 applies, and it's better to focus again on getting the newer version of
 the package into testing.

> trying: -pike
> skipped: -pike (322 <- 16)
>     got: 44+0: a-44
>     * alpha: libroxen-floatingcode, libroxen-form, roxen

Suggestion: wait-and-see
Analysis: Now provided by pike7.2 source package in unstable, which is
 waiting only for caudium to be built on all architectures.  As soon as
 s390 packages are uploaded, this can be hinted in.

> trying: -pike7
> skipped: -pike7 (323 <- 15)
>     got: 189+0: a-41:a-42:h-41:i-65
>     * i386: roxen2, roxen2-ssl

Suggestion: wait-and-see
Analysis: In bug 166311, the roxen2 maintainer admits that this package
 is "dead upstream".  I've followed up to point out it is now
 uninstallable in sid, asking whether these packages should be shipped
 with sarge.

> trying: -pspell
> skipped: -pspell (324 <- 14)
>     got: 42+0: a-42
>     * alpha: sylpheed-claws

Suggestion: wait-and-see
Analysis: The sylpheed-claws package has already been rebuilt against
 aspell in sid, and is waiting on the python2.3 transition.  Should
 resolve itself in due course.

> trying: -quicktime4linux
> skipped: -quicktime4linux (325 <- 13)
>     got: 195+0: a-41:a-42:h-41:i-71
>     * i386: fbtv, motv, scantv, streamer, ttv, webcam, xawtv, xawtv-plugins

Suggestion: wait-and-see
Analysis: The list of uninstallable packages seem to all point back to
 xawtv-plugins, which depended on libquicktime4linux0 in testing.  The
 version of xawtv in unstable is, in turn, waiting for the new version
 of libquicktime to be current on all archs before it propagates into
 testing.  The build failures on s390/arm appear to be caused by
 toolchain problems, not package bugs.

> trying: -ruby
> skipped: -ruby (326 <- 12)
>     got: 87+0: a-87
>     * alpha: eruby, libapache-mod-ruby, libdb-ruby, libdb2-ruby, libdb3-ruby, libdb4.1-ruby, libdbd-sqlite-ruby, libeb-ruby, liberuby, liberuby-dev, libexif-ruby, libfam-ruby, libgd-ruby, libgdk-imlib-ruby, libgdk-pixbuf-ruby, libglade-ruby, libgnome-ruby, libgtk-ruby, libiconv-ruby, libintl-gettext-ruby, libkakasi-ruby, libldap-ruby, libmysql-ruby, libpanel-applet-ruby, libpcap-ruby, libpgsql-ruby, libsdl-ruby, libshadow-ruby, libstrscan-ruby, libsufary-ruby, libtmail-ruby, libuconv-ruby, libvorbisfile-ruby, libwidestudio-dev, libwidestudio-ruby, libwrap-ruby, libxml-parser-ruby, libxmms-ruby, libzlib-ruby, mhc-utils, quantlib-ruby, racc-runtime, rubymagick, vim-ruby, widestudio, xmlrpc4r

Suggestion: wait-and-see
Analysis: eruby has been rebuilt against libruby1.8, but has build
 failures on two archs.  libapache-mod-ruby has also been rebuilt, but
 has a build failure on powerpc (and is waiting on apache).  Others
 still need rebuilding against libruby1.8.  Since ruby1.8 is current in
 testing, this looks fairly parallelizable; I'll go through and file
 bugs on any packages that haven't yet been bugged about it.

> trying: -suidmanager
> skipped: -suidmanager (327 <- 11)
>     got: 42+0: a-42
>     * alpha: amavis-ng-milter-helper

Suggestion: wait-and-see
Analysis: The amavis-ng package is uninstallable on two archs due to the
 absence of an arj package there, which keeps the suidmanager-less
 package out of testing.  Bug #213024 has been filed for this.

> trying: -uw-imap-ssl
> skipped: -uw-imap-ssl (328 <- 10)
>     got: 45+0: a-45
>     * alpha: php3-cgi-imap, php3-imap, php4-imap, postman

Suggestion: wait-and-see
Analysis: These packages have been rebuilt against uw-imap, but are
 waiting for that package to propagate into testing.  This appears to be
 another dependency waiting game.

> trying: -xanim
> skipped: -xanim (329 <- 9)
>     got: 188+0: a-41:a-42:h-41:i-64
>     * i386: tkxanim

From <20030909053946.EFDF11C93D@laptop.nahmias.net>:

Package: tkxanim
Testing: 0.43-6
Bugs: 208373
Suggestion: remove-from-testing, remove-from-unstable
  This package contains a Tcl/Tk interface to xanim.  xanim was removed
  from unstable because it was a long time orphaned non-free app that has
  since been replaces/superceded by xine.  As this package is useless
  without xanim, there doesn't seem much of a point keeping it around.
  N.B. tkxanim is the only reverse-dep on xanim; so after it is removed,
  both can/should be removed from testing.

Bug 212076 has been filed on ftp.debian.org requesting this removal.

> trying: -xerces
> skipped: -xerces (330 <- 8)
>     got: 45+0: a-45
>     * alpha: libapache-mod-xslt, libxalan1.2, libxalan1.2-dev, xalan
> trying: -xerces20
> skipped: -xerces20 (331 <- 7)
>     got: 42+0: a-42
>     * alpha: libxml-xerces-perl

These two will hopefully be taken care of tomorrow; requires some
hinting to get xalan, icu, and all three versions of xerces to go
through at once.

Additionally, these packages have already been removed from testing, and
can probably be reconsidered as candidates going the other way:

 haddock ndtpd mirrordir cwcdr ecartis mantis elvis sourcenav
 chicken gdis pydb gnome-pilot gnome-pilot-conduits solfege grap
 motioneye tora docbook-jrefentry fenris thoughttracker
 multi-gnome-terminal ferret gnome-pim netjuke eglade ncps jbofihe
 siege-ssl root-portal dancer-ircd boot-floppies gnews uae
 kernel-patch-ttl realtimebattle j-cvsbook vold objc-poc
 innovation3d-plugins xtend 

Many of these packages will still be held out of testing for other
reasons (RC bugs, out-of-date on some architectures).  

These packages should probably be manually kept out of testing for a
while due to the nature of historic brokenness:

 animals gamix ibm-jdk1.1-installer zope-ldap oops

I've requested removal from unstable for this package:


Steve Langasek
postmodern programmer

Attachment: signature.asc
Description: Digital signature

Reply to: