Bug#888162: [pkg-cryptsetup-devel] Bug#888162: Bug#888162: cryptsetup: `loopaesOpen --key-file=-` doesn't read the key from stdin but tries to open key file "./-"

2018-02-10 Thread Guilhem Moulin 
On Sat, 10 Feb 2018 at 23:17:34 +0100, Cyril Brulebois wrote:
> Guilhem Moulin  (2018-02-10):
>> Should we?  I was refraining from uploading 2.2.1 due to the following
>> note in the transition page
>> 
>>“Please avoid uploads unrelated to this transition, they would
>>likely delay it and require supplementary work from the release
>>managers.” — https://tracker.debian.org/pkg/cryptsetup
> 
> This is mainly targeted at packages which are included in the reverse
> dependencies graph: if they in turn get uploaded frequently, that tends
> to lengthen delays for the actual migration of all related packages.
> More on that in my last paragraph.
> […]
> cryptsetup-bin is a binary package built from the cryptsetup source
> package, so its RC bug count is included in cryptsetup's RC bug count,
> and having an extra bug compared to what's in testing means it's a
> regression = a blocker for the migration.

I see, thanks for the poke and the explanation.

> Back to the question at end: of course, fixing a regression in your
> package is something you want, and actually need to do. It's totally
> related to getting your package in shape, so please follow Michael's
> advice on this.

Uploading right after hitting save :-)

-- 
Guilhem.


signature.asc
Description: PGP signature

Bug#888162: [pkg-cryptsetup-devel] Bug#888162: Bug#888162: cryptsetup: `loopaesOpen --key-file=-` doesn't read the key from stdin but tries to open key file "./-"

2018-02-10 Thread Cyril Brulebois 
Hi Guilhem,

Guilhem Moulin  (2018-02-10):
> On Sat, 10 Feb 2018 at 09:22:44 +0100, Michael Biebl wrote:
> > On Wed, 24 Jan 2018 14:38:50 +0100 Guilhem Moulin 
> > wrote:
> >> On Wed, 24 Jan 2018 at 09:13:53 +0100, Milan Broz wrote:
> >>> Fixed upstream in
> >>> https://gitlab.com/cryptsetup/cryptsetup/commit/8728ba08e2e056a4c18b55407146eea7ac0043c6
> > 
> > could you please release an updated Debian cryptsetup package with that
> > fix. With the upload of 2.2.0 to unstable you started a library
> > transition which now blocks other packages from entering testing as
> > cryptsetup won't migrate due to this RC bug.
> 
> Should we?  I was refraining from uploading 2.2.1 due to the following
> note in the transition page
> 
> “Please avoid uploads unrelated to this transition, they would
> likely delay it and require supplementary work from the release
> managers.” — https://tracker.debian.org/pkg/cryptsetup

This is mainly targeted at packages which are included in the reverse
dependencies graph: if they in turn get uploaded frequently, that tends
to lengthen delays for the actual migration of all related packages.
More on that in my last paragraph.

> We were waiting for the transition to complete, which is currently
> blocking on #888072 (currently tagged pending); I was expecting a new
> upload of volume-key sooner though.
> 
> In fact this bug #888162 applies to ‘cryptsetup-bin’ not ‘cryptsetup’
> nor ‘libcryptsetup12’.  Reassigning accordingly; hopefully that will
> release most reverse dependencies from entering testing as
> ‘cryptsetup-bin’ only reverse depends on ‘ceph-base’ and ‘tomb’, both of
> which have not been updated since I raised the severity of #888162.

cryptsetup-bin is a binary package built from the cryptsetup source
package, so its RC bug count is included in cryptsetup's RC bug count,
and having an extra bug compared to what's in testing means it's a
regression = a blocker for the migration.

Back to the question at end: of course, fixing a regression in your
package is something you want, and actually need to do. It's totally
related to getting your package in shape, so please follow Michael's
advice on this.


Cheers,
-- 
Cyril Brulebois (k...@debian.org)
D-I release manager -- Release team member -- Freelance Consultant


signature.asc
Description: PGP signature

Bug#888162: [pkg-cryptsetup-devel] Bug#888162: Bug#888162: cryptsetup: `loopaesOpen --key-file=-` doesn't read the key from stdin but tries to open key file "./-"

2018-02-10 Thread Guilhem Moulin 
Control: reassign -1 cryptsetup-bin

Hi Michael,

On Sat, 10 Feb 2018 at 09:22:44 +0100, Michael Biebl wrote:
> On Wed, 24 Jan 2018 14:38:50 +0100 Guilhem Moulin 
> wrote:
>> On Wed, 24 Jan 2018 at 09:13:53 +0100, Milan Broz wrote:
>>> Fixed upstream in
>>> https://gitlab.com/cryptsetup/cryptsetup/commit/8728ba08e2e056a4c18b55407146eea7ac0043c6
> 
> could you please release an updated Debian cryptsetup package with that
> fix. With the upload of 2.2.0 to unstable you started a library
> transition which now blocks other packages from entering testing as
> cryptsetup won't migrate due to this RC bug.

Should we?  I was refraining from uploading 2.2.1 due to the following
note in the transition page

“Please avoid uploads unrelated to this transition, they would
likely delay it and require supplementary work from the release
managers.” — https://tracker.debian.org/pkg/cryptsetup

We were waiting for the transition to complete, which is currently
blocking on #888072 (currently tagged pending); I was expecting a new
upload of volume-key sooner though.

In fact this bug #888162 applies to ‘cryptsetup-bin’ not ‘cryptsetup’
nor ‘libcryptsetup12’.  Reassigning accordingly; hopefully that will
release most reverse dependencies from entering testing as
‘cryptsetup-bin’ only reverse depends on ‘ceph-base’ and ‘tomb’, both of
which have not been updated since I raised the severity of #888162.

Cheers,
-- 
Guilhem.


signature.asc
Description: PGP signature