found 1069097 3:25.1.0-4
thanks

Hi. I know this is unexpected, both for you and for me,
but it is what it is.

After building each of the 17 packages 100 times (excluding the one which
currently ftbfs always, which I reported this afternoon), this is what
I got on machines with 2 CPUs:

https://people.debian.org/~sanvila/build-logs/openstack-dashboard/

I would suggest that you try to fix this problem first:


Found another file with the destination path 
'horizon/lib/angular_schema_form/i18n/angular-locale_en-nf.js'. It will be 
ignored since only the first encountered file is collected. If this is not what 
you want, make sure every static file has a unique path.


In fact, I see things in the postinst which are not very orthodox:

    if [ ! -L /usr/lib/python3/dist-packages/openstack_dashboard/themes ]; then
                ln -sf /usr/share/openstack-dashboard/themes 
/usr/lib/python3/dist-packages/openstack_dashboard/themes
        fi

        if [ ! -L 
/usr/lib/python3/dist-packages/openstack_dashboard/local/local_settings.d ]; 
then
                ln -sf /etc/openstack-dashboard/local_settings.d 
/usr/lib/python3/dist-packages/openstack_dashboard/local/local_settings.d
        fi

(  btw: why are you using "! [" here instead of "[ !" ?  )

        if ! [ -L 
/usr/lib/python3/dist-packages/openstack_dashboard/local/local_settings.py ] ; 
then
                ln -sf /etc/openstack-dashboard/local_settings.py 
/usr/lib/python3/dist-packages/openstack_dashboard/local/local_settings.py
        fi

        if [ ! -L /usr/lib/python3/dist-packages/openstack_dashboard/conf ]; 
then
                ln -sf /etc/openstack-dashboard/policy 
/usr/lib/python3/dist-packages/openstack_dashboard/conf
        fi

Here, the postinst is creating things in /usr/lib which will not be under the 
control of dpkg.
This is not allowed by policy.

I fear that those symlinks may be related to the error message above, 
contributing
to the confusion of the django tools.

I guess that until know are assuming that those errors are harmless, but: what 
if
they are not harmless and they are in fact related with the random error at 
hand?

So: Can you address that before we decide what to try next?

Thanks.

Reply via email to