Re: Better Lintian checks

[Bastien ROUCARIES]

On Tue, Jan 26, 2016 at 7:45 AM, Sebastiaan Couwenberg
 wrote:
> On 25-01-16 19:47, Bastien Roucaries wrote:
>> I expect more problems  detected in  the next few days
>
> I guess the autorejects for Multi-Arch: no are among those problems.
> The use of Multi-Arch: no was suggested by
>
>  https://lintian.debian.org/tags/old-style-config-script-multiarch-path.html
>
> The reject message refers to #768353 which was fixed in November 2014,
> but the fix is most likely not available in stable yet, and hence still
> an issue on the Debian infrastructure?
>
> If "Multi-Arch: no support in Debian is broken" is still true, the
> suggested fix for old-style-config-script-multiarch-path is not an
> actual solution.
>
> Should the autoreject be removed, or should lintian suggest a different fix?

This tag is not on autoreject list.

My warning was about
https://lintian.debian.org/tags/license-problem-json-evil.html,
https://lintian.debian.org/tags/license-problem-cc-by-nc-sa.html,
 https://lintian.debian.org/tags/license-problem-non-free-img-lenna.html

About you remark long term solution is here
https://lintian.debian.org/tags/old-style-config-script.html

Bastien



> Kind Regards,
>
> Bas
>
> --
>  GPG Key ID: 4096R/6750F10AE88D4AF1
> Fingerprint: 8182 DE41 7056 408D 6146  50D1 6750 F10A E88D 4AF1
>

Re: Better Lintian checks

[Jakub Wilk]

* Sebastiaan Couwenberg , 2016-01-26, 07:45:

The use of Multi-Arch: no was suggested by

https://lintian.debian.org/tags/old-style-config-script-multiarch-path.html


The wording is unfortunate.

You should not add "Multi-Arch: no" to the control file, but instead 
remove the field, because "no" is the default.


And most of the time changing multi-archness isn't the correct course of 
action anyway...



The reject message


Quoting the reject message would have been helpful...
I guess you're referring to this:
https://anonscm.debian.org/cgit/mirror/dak.git/tree/daklib/checks.py?id=c51e71bbd9c2#n392

refers to #768353 which was fixed in November 2014, 
but the fix is most likely not available in stable yet,


The version graph says it's fixed in "dose3/3.3~beta1-3 (stable)".

--
Jakub Wilk

Re: Better Lintian checks

[Bas Couwenberg]

On 2016-01-26 12:10, Jakub Wilk wrote:

* Sebastiaan Couwenberg , 2016-01-26, 07:45:

The use of Multi-Arch: no was suggested by

https://lintian.debian.org/tags/old-style-config-script-multiarch-path.html


The wording is unfortunate.

You should not add "Multi-Arch: no" to the control file, but instead
remove the field, because "no" is the default.


The gdal package didn't have any Multi-Arch control fields.


And most of the time changing multi-archness isn't the correct course
of action anyway...


Yeah, Bastiens reference to the suggestion for the general 
old-style-config-script tag should be the way forward.


I just don't have time to fix all reverse dependencies, so I'd like to 
keep including gdal-config for the time being.



The reject message


Quoting the reject message would have been helpful...
I guess you're referring to this:
https://anonscm.debian.org/cgit/mirror/dak.git/tree/daklib/checks.py?id=c51e71bbd9c2#n392


I quoted most of the reject message except the bug number.

The reject in question is here:

http://lists.alioth.debian.org/pipermail/pkg-grass-devel/2016-January/042758.html

The speed at which it arrived suggests that it's an auto reject, despite 
Bastien saying that it's not.


refers to #768353 which was fixed in November 2014, but the fix is 
most likely not available in stable yet,


The version graph says it's fixed in "dose3/3.3~beta1-3 (stable)".


Thanks for the feedback, unfortunately no solution yet.

Kind Regards,

Bas

Re: Better Lintian checks

[Jakub Wilk]

* Bas Couwenberg , 2016-01-26, 12:17:

The use of Multi-Arch: no was suggested by

https://lintian.debian.org/tags/old-style-config-script-multiarch-path.html


The wording is unfortunate.

You should not add "Multi-Arch: no" to the control file, but instead 
remove the field, because "no" is the default.


The gdal package didn't have any Multi-Arch control fields.


This will be fixed in the next Lintian release:
https://anonscm.debian.org/cgit/lintian/lintian.git/commit/?id=20ad1f7fbe12

In the mean time, please add an override for the tag.

--
Jakub Wilk

Re: Better Lintian checks

[Bas Couwenberg]

On 2016-01-26 14:41, Jakub Wilk wrote:

* Bas Couwenberg , 2016-01-26, 12:17:

The use of Multi-Arch: no was suggested by

https://lintian.debian.org/tags/old-style-config-script-multiarch-path.html


The wording is unfortunate.

You should not add "Multi-Arch: no" to the control file, but instead 
remove the field, because "no" is the default.


The gdal package didn't have any Multi-Arch control fields.


This will be fixed in the next Lintian release:
https://anonscm.debian.org/cgit/lintian/lintian.git/commit/?id=20ad1f7fbe12


Excellent, thanks!


In the mean time, please add an override for the tag.


Will do.

Kind Regards,

Bas

Re: Better Lintian checks

[Sascha Steinbiss]

Hi Bastien,

> Newer unstable Lintian version check now package testsuite for licence
> problems/missing source

Just FYI, the potential 'source-is-missing' false positive mentioned at
the beginning of #798900 [1] still exists, regarding the automatic
flagging of JS files with lines > 512 chars as minified although they
may not be.
The long line in question is also present in the source in my separately
packaged version of DataTables, which is built from upstream's source
repo exactly to address this borderline case.
See [2].

This just popped back on my radar since the recent rewording of the
lintian message broke my override ;)

Cheers
Sascha

[1] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=798900
[2] https://github.com/DataTables/DataTablesSrc/blob/master/js/DataTables.js


-- 
 The Wellcome Trust Sanger Institute is operated by Genome Research 
 Limited, a charity registered in England with number 1021457 and a 
 company registered in England with number 2742969, whose registered 
 office is 215 Euston Road, London, NW1 2BE. 

Re: Better Lintian checks

[Sebastiaan Couwenberg]

On 25-01-16 19:47, Bastien Roucaries wrote:
> I expect more problems  detected in  the next few days

I guess the autorejects for Multi-Arch: no are among those problems.
The use of Multi-Arch: no was suggested by

 https://lintian.debian.org/tags/old-style-config-script-multiarch-path.html

The reject message refers to #768353 which was fixed in November 2014,
but the fix is most likely not available in stable yet, and hence still
an issue on the Debian infrastructure?

If "Multi-Arch: no support in Debian is broken" is still true, the
suggested fix for old-style-config-script-multiarch-path is not an
actual solution.

Should the autoreject be removed, or should lintian suggest a different fix?

Kind Regards,

Bas

-- 
 GPG Key ID: 4096R/6750F10AE88D4AF1
Fingerprint: 8182 DE41 7056 408D 6146  50D1 6750 F10A E88D 4AF1