Hi,

as announced in bug #1012103[1] I'm hereby asking for removal of this
package.  I do absolutely share the sentiment of Sascha Silbe that it is
a shame to see this package go.  However, if there is no commitment to
fix the issues of this package inside Debian it does not make any sense
to pretend it is maintained.

I'd be happy if someone educated in Java and sufficient time resources
might prove my assumption wrong, tag this bug `moreinfo` and fix bug
#1012103.

Kind regards
    Andreas.

[1] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1012103#32

Hi Andreas, I vaguely remember that Kotlin requires some parts of the
Android ecosystem. I haven't check, but if
android-platform-external-doclava is in the dependency graph it should
be preserved in Sid where it can be built with openjdk-8.

Emmanuel Bourg

Hi Emmanuel,

Am Tue, Dec 09, 2025 at 12:40:17PM +0100 schrieb Emmanuel Bourg:

Thank you for you response.  There is only

udd=> select source, build_depends from sources where release = 'sid' and build_depends like '%doclava-aosp%';
        source        |                                                                                              build_depends
----------------------+---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
 android-framework-23 | aapt, aidl, android-logtags-tools, debhelper (>= 11), default-jdk-headless | default-jdk (>= 2:1.7), dh-exec, doclava-aosp, gradle-debian-helper (>= 1.4~), maven-repo-helper, p7zip-full, proguard-cli
(1 Zeile)


Given that android-framework-23 builds three binary package there is


udd=> select distinct source, build_depends from sources where release = 'sid' and build_depends like '%android-sdk-platform-23%' or build_depends like  '%libandroid-23-java%' or build_depends like '%libandroid-uiautomator-23-java%' ;

  source  |                                                                                                                                                                                    build_depends
----------+-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
 libscout | debhelper (>= 10~), bash-completion, default-jdk-headless (>= 2:1.7) | default-jdk (>= 2:1.7), gradle-debian-helper, javahelper, libandroid-23-java, libantlr4-runtime-java, libcava-java, libcommons-cli-java (>= 1.2), libcommons-io-java (>= 2.6), libgoogle-gson-java (>= 2.8.0), libguava-java, liblogback-java, libsemver-java, libslf4j-java, libwala-java (>= 1.5.2), unzip
(1 Zeile)


android-framework-23 and libscout are affected by the RC buggy
android-platform-external-doclava but I can't see any other broken
(Build-)Depends.  If android-platform-external-doclava is a *hard
Build-Depends (and can't be dropped) I guess these packages should be
removed as well.

And yes, I see their practical relevance and it would be great to have
some way to keep them.

Kind regards
    Andreas.

There is a reverse dependency that needs to be taken care of:


Checking reverse dependencies...
# Broken Build-Depends:
android-framework-23: doclava-aosp


In case it matters, this needs to be addressed first. Please remove the
moreinfo tag once that is done.

   Thorsten

Hi android-framework-23 maintainers,

I filed

   #1122250 RM: android-platform-external-doclava -- RoQA; Not in stable nor old-stable due to RC bug

but Thorsten pointed out the dependency to android-framework-23.  Could
you please give some statement whether you plan to fix the open RC bugs
for android-framework-23 in some way?  Due to open RC bugs this package
has missed two releases and I wonder whether it might be time to ask for
removal of this source package as well.

Kind regards
   Andreas.

Am Tue, Jan 20, 2026 at 06:11:17PM +0000 schrieb Thorsten Alteholz: