[dpkg] 08/08: document 1.18.10.0~reproducible1

Mattia Rizzolo mattia at debian.org
Wed Sep 21 14:43:46 UTC 2016


On Wed, Sep 21, 2016 at 01:13:42PM +0000, Jérémy Bobbio wrote:
> Mattia Rizzolo:
> > The reason we like to have a static string in there (and in jenkins we
> > use the host architecture), is that otherwise diffoscope would be noisy
> > about the difference in the name if we compared the .changes files […].
> 
> Would it still be the case after Satyam's work which resulted in
> https://anonscm.debian.org/cgit/reproducible/diffoscope.git/commit/?id=c61f41c5128094508d6fff8b746cecc3841f6b27
> ?
> 
> This change has been shipped as part of diffoscope/60 from what I can see.

yes:

mattia at chase ~/pbuilder/result/sid/amd64 % diffoscope b1/libjs-jquery-scrollto_2.1.2+dfsg-3_all.changes b2/libjs-jquery-scrollto_2.1.2+dfsg-3_all.changes
--- b1/libjs-jquery-scrollto_2.1.2+dfsg-3_all.changes
+++ b2/libjs-jquery-scrollto_2.1.2+dfsg-3_all.changes
├── Files
│ @@ -1,3 +1,3 @@
│  
│ - 20e7077ef60cd2df6d4b6a21139f6c3d 4492 web extra libjs-jquery-scrollto_2.1.2+dfsg-3_20160921T143043z-20e7077e.buildinfo
│ + 20e7077ef60cd2df6d4b6a21139f6c3d 4492 web extra libjs-jquery-scrollto_2.1.2+dfsg-3_20160921T143138z-20e7077e.buildinfo
│   28bd68174b41c58b73d56cd8746d3c88 8308 web extra libjs-jquery-scrollto_2.1.2+dfsg-3_all.deb
│   --- libjs-jquery-scrollto_2.1.2+dfsg-3_20160921T143043z-20e7077e.buildinfo
├── +++ libjs-jquery-scrollto_2.1.2+dfsg-3_20160921T143138z-20e7077e.buildinfo
│┄ Files similar despite different names (difference score: 0)
╵
1 mattia at chase ~/pbuilder/result/sid/amd64 %                                                                                                                         :(


And this is a case for a reproducible build.
I also tried to stick in an extra field in b2/.buildinfo and (still with
the different filename) it reported that difference too (which is not
supposed to do).

And btw, I also tried to stick in an additional file in the second
.buildinfo, and renaming them, which is supposedly enough to have
diffoscope skip that difference, but:

mattia at chase ~/pbuilder/result/sid/amd64 % diffoscope b1/libjs-jquery-scrollto_2.1.2+dfsg-3_all.changes b2/libjs-jquery-scrollto_2.1.2+dfsg-3_all.changes
--- b1/libjs-jquery-scrollto_2.1.2+dfsg-3_all.changes
+++ b2/libjs-jquery-scrollto_2.1.2+dfsg-3_all.changes
├── Files
│ @@ -1,3 +1,3 @@
│  
│ - 20e7077ef60cd2df6d4b6a21139f6c3d 4492 web extra libjs-jquery-scrollto_2.1.2+dfsg-3_amd64.buildinfo
│ + 845e31a2b41665575b4c3af404a3f130 4492 web extra libjs-jquery-scrollto_2.1.2+dfsg-3_amd64.buildinfo
│   28bd68174b41c58b73d56cd8746d3c88 8308 web extra libjs-jquery-scrollto_2.1.2+dfsg-3_all.deb
├── libjs-jquery-scrollto_2.1.2+dfsg-3_amd64.buildinfo
│   ├── Env
│   │ @@ -0,0 +1 @@
│   │ +ciao
│   ╵
╵


now, I'm not sure whether I should take care of also updating all
hashes, etc while messing around with .changes for diffoscope to work
correctly, but it is not working as I would expect.  I remember also
trying the day I saw the commit coming, but noticed something was fishy.
I always meant to check what was going on, but I'm already quite limited
on my time available.

Not to mention that this is not totally applicable right away as we can't
compare .changes until it stops recursing over files twice (one through
.changes and one trhough .changes->.buildinfo).  I should probably file
a bug for this though.  Until now it was really "oral knowledge" of very
few (and maybe of even fewer who remembers seeing such diffs nearly an
year ago).


-- 
regards,
                        Mattia Rizzolo

GPG Key: 66AE 2B4A FCCF 3F52 DA18  4D18 4B04 3FCD B944 4540      .''`.
more about me:  https://mapreri.org                             : :'  :
Launchpad user: https://launchpad.net/~mapreri                  `. `'`
Debian QA page: https://qa.debian.org/developer.php?login=mattia  `-
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 801 bytes
Desc: not available
URL: <http://lists.alioth.debian.org/pipermail/reproducible-builds/attachments/20160921/8a82937b/attachment.sig>


More information about the Reproducible-builds mailing list