Debian Bug report logs - #829362
ruby-ronn: please make the output reproducible

Toggle useless messages

Message #5 received at submit@bugs.debian.org (full text, mbox, reply):

From: Chris Lamb <lamby@debian.org>

To: submit@bugs.debian.org

Subject: ruby-ronn: please make the output reproducible

Date: Sat, 02 Jul 2016 21:08:02 +0200

[Message part 1 (text/plain, inline)]

Source: ruby-ronn
Version: 0.7.3-4
Severity: wishlist
Tags: patch
User: reproducible-builds@lists.alioth.debian.org
Usertags: toolchain timestamps
X-Debbugs-Cc: reproducible-builds@lists.alioth.debian.org

Hi,

Whilst working on the "reproducible builds" effort [0], we noticed
that ruby-ronn generates (other) non-reproducible output.

Patch attached.

 [0] https://wiki.debian.org/ReproducibleBuilds


Regards,

-- 
      ,''`.
     : :'  :     Chris Lamb
     `. `'`      lamby@debian.org / chris-lamb.co.uk
       `-

[ruby-ronn.diff.txt (application/octet-stream, attachment)]

Message #10 received at 829362@bugs.debian.org (full text, mbox, reply):

From: Chris Lamb <lamby@debian.org>

To: 829362@bugs.debian.org

Subject: Re: ruby-ronn: please make the output reproducible

Date: Sat, 02 Jul 2016 21:12:31 +0200

forwarded 829362 https://github.com/rtomayko/ronn/pull/98
thanks

Hi,

I've forwarded this issue upstream here:

  https://github.com/rtomayko/ronn/pull/98


Regards,

-- 
      ,''`.
     : :'  :     Chris Lamb
     `. `'`      lamby@debian.org / chris-lamb.co.uk
       `-

Message #21 received at 829362@bugs.debian.org (full text, mbox, reply):

From: Chris Lamb <lamby@debian.org>

To: 829362@bugs.debian.org

Subject: Re: Bug#829362: ruby-ronn: please make the output reproducible

Date: Fri, 02 Sep 2016 11:04:07 +0100

Dear Maintainer,

> Source: ruby-ronn
> Version: 0.7.3-4
> Tags: patch

There hasn't seem to be any update here in a while and this package
is affecting the reproducibility status of many packages.

Would you consider applying this patch and uploading? :)


Regards,

-- 
      ,''`.
     : :'  :     Chris Lamb
     `. `'`      lamby@debian.org / chris-lamb.co.uk
       `-

Message #26 received at 829362@bugs.debian.org (full text, mbox, reply):

From: Antonio Terceiro <terceiro@debian.org>

To: Chris Lamb <lamby@debian.org>, 829362@bugs.debian.org

Subject: Re: Bug#829362: ruby-ronn: please make the output reproducible

Date: Fri, 2 Sep 2016 15:54:32 -0300

[Message part 1 (text/plain, inline)]

Hi,

On Fri, Sep 02, 2016 at 11:04:07AM +0100, Chris Lamb wrote:
> Dear Maintainer,
> 
> > Source: ruby-ronn
> > Version: 0.7.3-4
> > Tags: patch
> 
> There hasn't seem to be any update here in a while and this package
> is affecting the reproducibility status of many packages.
> 
> Would you consider applying this patch and uploading? :)

Another reproducibility patch has been applied to close #807201, and
this one does not apply on top of 0.7.3-4; in fact it looks like a
modified version of that one.

So do we want to replace the earlier patch with this one?

[signature.asc (application/pgp-signature, inline)]

Message #31 received at 829362@bugs.debian.org (full text, mbox, reply):

From: Chris Lamb <lamby@debian.org>

To: Antonio Terceiro <terceiro@debian.org>, 829362@bugs.debian.org

Subject: Re: Bug#829362: ruby-ronn: please make the output reproducible

Date: Fri, 02 Sep 2016 20:03:06 +0100

Antonio Terceiro wrote:

> So do we want to replace the earlier patch with this one?

Hm, the patch to #829362 (which modifies the patch) is kinda ugly and
almost impossible to mentally parse. Sorry about that.

So, considering this cleaner patch:

 https://github.com/rtomayko/ronn/pull/98/files

… I believe we need all of these changes so, yes, we should replace
it. The following hunk demonstrates that it absorbs the previous patch:

 https://github.com/rtomayko/ronn/pull/98/files#diff-a87e9e80c3a7693175aa8c08bdb48452L70

Thanks!


Regards,

-- 
      ,''`.
     : :'  :     Chris Lamb
     `. `'`      lamby@debian.org / chris-lamb.co.uk
       `-

Message #36 received at 829362@bugs.debian.org (full text, mbox, reply):

From: Antonio Terceiro <terceiro@debian.org>

To: Chris Lamb <lamby@debian.org>, 829362@bugs.debian.org

Subject: Re: Bug#829362: ruby-ronn: please make the output reproducible

Date: Sat, 3 Sep 2016 13:26:03 -0300

[Message part 1 (text/plain, inline)]

On Fri, Sep 02, 2016 at 08:03:06PM +0100, Chris Lamb wrote:
> Antonio Terceiro wrote:
> 
> > So do we want to replace the earlier patch with this one?
> 
> Hm, the patch to #829362 (which modifies the patch) is kinda ugly and
> almost impossible to mentally parse. Sorry about that.
> 
> So, considering this cleaner patch:
> 
>  https://github.com/rtomayko/ronn/pull/98/files
> 
> … I believe we need all of these changes so, yes, we should replace
> it. The following hunk demonstrates that it absorbs the previous patch:
> 
>  https://github.com/rtomayko/ronn/pull/98/files#diff-a87e9e80c3a7693175aa8c08bdb48452L70

I tried that, but with this patch the build failed. I rewrote it into a
much less invasive version:

--- a/lib/ronn/document.rb
+++ b/lib/ronn/document.rb
@@ -165,6 +165,7 @@ module Ronn
     # the current time.
     def date
       return @date if @date
+      return Time.at(ENV['SOURCE_DATE_EPOCH'].to_i).gmtime if ENV['SOURCE_DATE_EPOCH']
       return File.mtime(path) if File.exist?(path)
       Time.now
     end

IMO all the other changes are not really warranted.

With the resulting ruby-ronn binary, I tried both ruby-coveralls and
foodcritic, which are listed in the issue page¹, and they became reproducible.

¹ https://tests.reproducible-builds.org/debian/issues/unstable/timestamps_in_manpages_generated_by_ronn_issue.html,

[signature.asc (application/pgp-signature, inline)]

Message #41 received at 829362@bugs.debian.org (full text, mbox, reply):

From: Chris Lamb <lamby@debian.org>

To: Antonio Terceiro <terceiro@debian.org>, 829362@bugs.debian.org

Subject: Re: Bug#829362: ruby-ronn: please make the output reproducible

Date: Sat, 03 Sep 2016 17:32:38 +0100

> I tried that, but with this patch the build failed.

Please elaborate?

> IMO all the other changes are not really warranted.

Oh? Could you comment on why you believe this to be not necessary?

 https://github.com/rtomayko/ronn/pull/98/files#diff-a87e9e80c3a7693175aa8c08bdb48452L70

(It seems self-evidently not reproducible, and I almost certainly added
it to the patch to fix a particular package...)


Regards,

-- 
      ,''`.
     : :'  :     Chris Lamb
     `. `'`      lamby@debian.org / chris-lamb.co.uk
       `-

Message #46 received at 829362-close@bugs.debian.org (full text, mbox, reply):

From: Antonio Terceiro <terceiro@debian.org>

To: 829362-close@bugs.debian.org

Subject: Bug#829362: fixed in ruby-ronn 0.7.3-5

Date: Sat, 03 Sep 2016 17:27:23 +0000

Source: ruby-ronn
Source-Version: 0.7.3-5

We believe that the bug you reported is fixed in the latest version of
ruby-ronn, which is due to be installed in the Debian FTP archive.

A summary of the changes between this version and the previous one is
attached.

Thank you for reporting the bug, which will now be closed.  If you
have further comments please address them to 829362@bugs.debian.org,
and the maintainer will reopen the bug report if appropriate.

Debian distribution maintenance software
pp.
Antonio Terceiro <terceiro@debian.org> (supplier of updated ruby-ronn package)

(This message was generated automatically at their request; if you
believe that there is a problem with it please contact the archive
administrators by mailing ftpmaster@ftp-master.debian.org)


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

Format: 1.8
Date: Sat, 03 Sep 2016 12:08:42 -0300
Source: ruby-ronn
Binary: ruby-ronn
Architecture: source
Version: 0.7.3-5
Distribution: unstable
Urgency: medium
Maintainer: Debian Ruby Extras Maintainers <pkg-ruby-extras-maintainers@lists.alioth.debian.org>
Changed-By: Antonio Terceiro <terceiro@debian.org>
Description:
 ruby-ronn  - Builds manuals from Markdown
Closes: 829362
Changes:
 ruby-ronn (0.7.3-5) unstable; urgency=medium
 .
   * Team upload.
   * Replace existing output reproducibility patch with a new, less invasive
     one. Thanks to Chris Lamb for the original patch (Closes: #829362)
   * Refresh packaging with `dh-maske-ruby -w`.
Checksums-Sha1:
 a4decc6d825f12dfa3286c2809ceabf9fc6732d9 2108 ruby-ronn_0.7.3-5.dsc
 92945a783e30b6678af194c2d30583383068f28b 4724 ruby-ronn_0.7.3-5.debian.tar.xz
Checksums-Sha256:
 a17b2039abd5c540ea2a7d4e7e900c96f264485b357c447b4dc8280690bcffbe 2108 ruby-ronn_0.7.3-5.dsc
 03bb258ccd17a49688fbae492d092f329197ba6f3cf7e60da6933871f140e390 4724 ruby-ronn_0.7.3-5.debian.tar.xz
Files:
 1610007ecec4c2dbbf7168d865d8d558 2108 ruby optional ruby-ronn_0.7.3-5.dsc
 a1983f9e8936a5119c194ea7bf77b015 4724 ruby optional ruby-ronn_0.7.3-5.debian.tar.xz

-----BEGIN PGP SIGNATURE-----
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=py2s
-----END PGP SIGNATURE-----

Message #51 received at 829362@bugs.debian.org (full text, mbox, reply):

From: Chris Lamb <lamby@debian.org>

To: Antonio Terceiro <terceiro@debian.org>

Cc: 829362@bugs.debian.org

Subject: Re: Bug#829362: ruby-ronn: please make the output reproducible

Date: Sun, 04 Sep 2016 16:28:59 +0100

[Re-adding 829362@bugs.debian.org to CC, and quoting in full]

Antonio Terceiro wrote:

> On Sat, Sep 03, 2016 at 05:32:38PM +0100, Chris Lamb wrote:
> > > I tried that, but with this patch the build failed.
> > 
> > Please elaborate?
> 
> Tests failed. I don't have the logs anymore, but they did.

If you happen to apply it again, please retain them.

> > > IMO all the other changes are not really warranted.
> > 
> > Oh? Could you comment on why you believe this to be not necessary?
> > 
> >  https://github.com/rtomayko/ronn/pull/98/files#diff-a87e9e80c3a7693175aa8c08bdb48452L70
> > 
> > (It seems self-evidently not reproducible, and I almost certainly added
> > it to the patch to fix a particular package...)
> 
> ruby-ronn does not do locales:
> 
> $ cat debian/coveralls.mkd | ronn | grep 2016
> .TH "COVERALLS" "1" "September 2016" "" ""
> $ cat debian/coveralls.mkd | LANG=pt_BR.UFT-8 ronn | grep 2016
> .TH "COVERALLS" "1" "September 2016" "" ""

"UFT-8"? Again, it seems highly unlikely I patched this without good
reason.


Regards,

-- 
      ,''`.
     : :'  :     Chris Lamb
     `. `'`      lamby@debian.org / chris-lamb.co.uk
       `-

Message #56 received at 829362@bugs.debian.org (full text, mbox, reply):

From: Antonio Terceiro <terceiro@debian.org>

To: Chris Lamb <lamby@debian.org>, 829362@bugs.debian.org

Subject: Re: Bug#829362: ruby-ronn: please make the output reproducible

Date: Sun, 4 Sep 2016 17:33:16 -0300

[Message part 1 (text/plain, inline)]

On Sun, Sep 04, 2016 at 04:28:59PM +0100, Chris Lamb wrote:
> [Re-adding 829362@bugs.debian.org to CC, and quoting in full]
> 
> Antonio Terceiro wrote:
> 
> > On Sat, Sep 03, 2016 at 05:32:38PM +0100, Chris Lamb wrote:
> > > > I tried that, but with this patch the build failed.
> > > 
> > > Please elaborate?
> > 
> > Tests failed. I don't have the logs anymore, but they did.
> 
> If you happen to apply it again, please retain them.
> 
> > > > IMO all the other changes are not really warranted.
> > > 
> > > Oh? Could you comment on why you believe this to be not necessary?
> > > 
> > >  https://github.com/rtomayko/ronn/pull/98/files#diff-a87e9e80c3a7693175aa8c08bdb48452L70
> > > 
> > > (It seems self-evidently not reproducible, and I almost certainly added
> > > it to the patch to fix a particular package...)
> > 
> > ruby-ronn does not do locales:
> > 
> > $ cat debian/coveralls.mkd | ronn | grep 2016
> > .TH "COVERALLS" "1" "September 2016" "" ""
> > $ cat debian/coveralls.mkd | LANG=pt_BR.UFT-8 ronn | grep 2016
> > .TH "COVERALLS" "1" "September 2016" "" ""
> 
> "UFT-8"? Again, it seems highly unlikely I patched this without good
> reason.

yeah that was a typo, but again, ronn does not do locales (at all), so your
changes to remove the usage of %B don't accomplish anything. maybe if I copy
and paste the locale names you will believe me :-)

$ cat debian/coveralls.mkd | LANG=C ronn | grep 2016
.TH "COVERALLS" "1" "September 2016" "" ""
$ cat debian/coveralls.mkd | LANG=pt_BR.UTF-8 ronn | grep 2016
.TH "COVERALLS" "1" "September 2016" "" ""
$ cat debian/coveralls.mkd | LANG=fr_FR.UTF-8 ronn | grep 2016
.TH "COVERALLS" "1" "September 2016" "" ""

[signature.asc (application/pgp-signature, inline)]

Message #61 received at 829362@bugs.debian.org (full text, mbox, reply):

From: Chris Lamb <lamby@debian.org>

To: Antonio Terceiro <terceiro@debian.org>, 829362@bugs.debian.org

Subject: Re: Bug#829362: ruby-ronn: please make the output reproducible

Date: Sun, 04 Sep 2016 22:19:35 +0100

> yeah that was a typo, but again, ronn does not do locales (at all)

Looking into this, it appear's that ruby's strftime is locale independent,
rather than something ronn-specific. This the opposite behaviour to every
other strftime that I've come across, hence the confusion.

However, you are still missing:

 - date.strftime
 + date.utc.strftime

… for when building close to month changes with respect to TZ.


Regards,

-- 
      ,''`.
     : :'  :     Chris Lamb
     `. `'`      lamby@debian.org / chris-lamb.co.uk
       `-

Message #66 received at 829362@bugs.debian.org (full text, mbox, reply):

From: Antonio Terceiro <terceiro@debian.org>

To: Chris Lamb <lamby@debian.org>

Cc: 829362@bugs.debian.org

Subject: Re: Bug#829362: ruby-ronn: please make the output reproducible

Date: Sun, 4 Sep 2016 20:30:39 -0300

[Message part 1 (text/plain, inline)]

On Sun, Sep 04, 2016 at 10:19:35PM +0100, Chris Lamb wrote:
> > yeah that was a typo, but again, ronn does not do locales (at all)
> 
> Looking into this, it appear's that ruby's strftime is locale independent,
> rather than something ronn-specific. This the opposite behaviour to every
> other strftime that I've come across, hence the confusion.

Last time I checked strftime in any language will only format according
to the locale defined by environment variables if and only if the
program calling strftime has initialized the locale system with
setlocale().

> However, you are still missing:
> 
>  - date.strftime
>  + date.utc.strftime

one of the things that was making the test crash was this. in Ruby, `Date`
objects do not have a `utc` method.

> … for when building close to month changes with respect to TZ.

let's see

> --- a/lib/ronn/document.rb
> +++ b/lib/ronn/document.rb
> @@ -165,6 +165,7 @@ module Ronn
>      # the current time.
>      def date
>        return @date if @date

if the date has been passed explicitly in the command line (with `ron --date
...`), it will be a fixed date and thus will not vary according to $TZ.

ronn parses the date with

options['date'] &&= Date.strptime(options['date'], '%Y-%m-%d')

so let's do that

$ TZ=UTC ruby -rdate -e 'puts Date.strptime("2016-10-01", "%Y-%m-%d")'
2016-10-01
$ TZ=Asia/Tokyo ruby -rdate -e 'puts Date.strptime("2016-10-01", "%Y-%m-%d")'
2016-10-01

Also, Date objects in Ruby do not store either hours or timezones. Only Time
objcts do that.

> +      return Time.at(ENV['SOURCE_DATE_EPOCH'].to_i).gmtime if ENV['SOURCE_DATE_EPOCH']

otherwise, if SOURCE_DATE_EPOCH is passed, the date is _already_ in UTC (`utc`
and `gmtime` are synonyms in the ruby date class).

>        return File.mtime(path) if File.exist?(path)
>        Time.now
>      end

I believe the patch as it currently is solves any reproducibility issues. if
you can find a concrete and reproducible case where it does not, please report
a new bug.

[signature.asc (application/pgp-signature, inline)]

Message #71 received at 829362@bugs.debian.org (full text, mbox, reply):

From: Antonio Terceiro <terceiro@debian.org>

To: Chris Lamb <lamby@debian.org>

Cc: 829362@bugs.debian.org

Subject: Re: Bug#829362: ruby-ronn: please make the output reproducible

Date: Sun, 4 Sep 2016 20:31:56 -0300

[Message part 1 (text/plain, inline)]

On Sun, Sep 04, 2016 at 08:30:39PM -0300, Antonio Terceiro wrote:
> > +      return Time.at(ENV['SOURCE_DATE_EPOCH'].to_i).gmtime if ENV['SOURCE_DATE_EPOCH']
> 
> otherwise, if SOURCE_DATE_EPOCH is passed, the date is _already_ in UTC (`utc`
> and `gmtime` are synonyms in the ruby date class).

s/ruby date class/ruby Time class/

[signature.asc (application/pgp-signature, inline)]

Send a report that this bug log contains spam.