Re: [Reproducible-builds] SOURCE_DATE_EPOCH specification document
Hi, On Freitag, 11. September 2015, Chris Lamb wrote: > > Just one little thing: there is no tag "1.0". > Whoops. Pushed. Thanks. > > Should we start a 1.1 branch now? > If you think that would be useful, sure; it would provide a place for > non-trivial changes to go. I guess taht can wait for actual commits, but I think it's good we agreed on this road now. > If it helps, I'm pretending that all cosmetic/typo changes since 1.0 > were actually part of 1.0. Whilst a bit naughty, it does mean that the > version we publish on the web a) has these corrections, and b) is an > actual released version, rather than being labelled as "pre-release" > which would not encourage adoption (the entire point of the spec to > begin with). or we could just release them as 1.0.1, 1.0.2, ...? But then, I also think that's too much for pure typo fixes. OTOH, this would be where an 1.0 branch would come handy... (I'm now also thinking about tje job building + publishing the spec...) cheers, Holger signature.asc Description: This is a digitally signed message part. ___ Reproducible-builds mailing list Reproducible-builds@lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds
Re: [Reproducible-builds] SOURCE_DATE_EPOCH specification document
> Just one little thing: there is no tag "1.0". Whoops. Pushed. > Should we start a 1.1 branch now? If you think that would be useful, sure; it would provide a place for non-trivial changes to go. If it helps, I'm pretending that all cosmetic/typo changes since 1.0 were actually part of 1.0. Whilst a bit naughty, it does mean that the version we publish on the web a) has these corrections, and b) is an actual released version, rather than being labelled as "pre-release" which would not encourage adoption (the entire point of the spec to begin with). Hope that makes sense. Regards, -- ,''`. : :' : Chris Lamb `. `'` la...@debian.org / chris-lamb.co.uk `- ___ Reproducible-builds mailing list Reproducible-builds@lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds
Re: [Reproducible-builds] SOURCE_DATE_EPOCH specification document
Hi Chris, On Mittwoch, 2. September 2015, Chris Lamb wrote: > I declare this version 1.0: > https://reproducible-builds.org/specs/source-date-epoch/ \o/ Thanks a lot for all your (plural!) work on this! Just one little thing: there is no tag "1.0". Also I slightly wonder how / when to increase the version number... IOW: is only e8ba01d2984 = 1.0 or is 35e38e6a23 = 1.0 as well? Should we start a 1.1 branch now? cheers, Holger signature.asc Description: This is a digitally signed message part. ___ Reproducible-builds mailing list Reproducible-builds@lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds
Re: [Reproducible-builds] SOURCE_DATE_EPOCH specification document
> I'll wait a few days, proof-read a more few times and then declare it > published. I declare this version 1.0: https://reproducible-builds.org/specs/source-date-epoch/ I completely gutted what we had previously as I felt the justifications were -- at best -- a distraction. A sincere thank-you to all; even if your own literal words do not remain in the document, it was extremely vital to the process. Regards, -- ,''`. : :' : Chris Lamb `. `'` la...@debian.org / chris-lamb.co.uk `- ___ Reproducible-builds mailing list Reproducible-builds@lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds
Re: [Reproducible-builds] SOURCE_DATE_EPOCH specification document
On 27/08/15 15:57, Ximin Luo wrote: > > In the meantime I'll push some other minor edits too. > I ended up making some more changes, rewording some things, adding considerations of more corner cases on implementors' sides, and more precise language from RFC 2119. Feel free to edit on top of it. Also the two points in my previous email still apply, I'm just confirming here that I *haven't* handled them in my recent commits. X -- GPG: 4096R/1318EFAC5FBBDBCE git://github.com/infinity0/pubkeys.git signature.asc Description: OpenPGP digital signature ___ Reproducible-builds mailing list Reproducible-builds@lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds
Re: [Reproducible-builds] SOURCE_DATE_EPOCH specification document
On 27/08/15 15:32, Chris Lamb wrote: > Chris Lamb wrote: > >> I plan to dump the rest of what's in my head today (depending on the >> level of sun) > > Sincere apologies that this took longer than promised, but I think I've > reached a good point: > > https://reproducible.debian.net/specs/source-date-epoch-spec/ > > I'll wait a few days, proof-read a more few times and then declare it > published. > > In the meantime, suggestions and contributions are highly welcome. Feel > free to commit "rough" changes directly to the repo; I am very happy to > "fold in" improvements in this way. (If you are not already a member of > our Alioth group, this would be the excuse you were waiting for..) > Hi, very nice work. I would add in a paragraph about why it's not feasible to keep embedded timestamps and simply have a "diff" program ignore these. I've encountered this proposal before by lazy upstreams that don't want to acknowledge that the problem (and hence solution) lies on *their* side. Something like (from the current wiki page): .. it is not feasible to develop a diff algorithm to ignore "build dates" in arbitrary data formats, and fundamentally impossible in the case of Turing-complete data formats such as executables, since the real behaviour of the result could easily change based on a piece of data embedded in the file, even if the data is itself static or immutable. > exposed through the system's usual environment mechanism unsure what "environment mechanism" means - hardware clock? syscall? In the meantime I'll push some other minor edits too. X -- GPG: 4096R/1318EFAC5FBBDBCE git://github.com/infinity0/pubkeys.git ___ Reproducible-builds mailing list Reproducible-builds@lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds
Re: [Reproducible-builds] SOURCE_DATE_EPOCH specification document
Chris Lamb wrote: > I plan to dump the rest of what's in my head today (depending on the > level of sun) Sincere apologies that this took longer than promised, but I think I've reached a good point: https://reproducible.debian.net/specs/source-date-epoch-spec/ I'll wait a few days, proof-read a more few times and then declare it published. In the meantime, suggestions and contributions are highly welcome. Feel free to commit "rough" changes directly to the repo; I am very happy to "fold in" improvements in this way. (If you are not already a member of our Alioth group, this would be the excuse you were waiting for..) Regards, -- ,''`. : :' : Chris Lamb `. `'` la...@debian.org / chris-lamb.co.uk `- ___ Reproducible-builds mailing list Reproducible-builds@lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds
Re: [Reproducible-builds] SOURCE_DATE_EPOCH specification document
On Wed, Aug 26, 2015 at 10:22:42PM +0200, Chris Lamb wrote:
> Mattia wrote:
>
> > > After a hint from doko, I've started work on an official-looking spec
> > > for SOURCE_DATE_EPOCH
> >
> > Lamby pointed out that this is not something debian-related, so it would
> > be great to have it outsite a debian.{org,net} site.
> >
> > OTOH also the howto is not strictly debian-related, and both documents
> > are related to reproducibly.
> >
> > My personal proposal is to merge the two documents and move them to a new
> > more meaningful url (under rb.d.n).
>
> I believe this would handicap adoption of SOURCE_DATE_EPOCH in the wider
> software community.
>
> One thing we can learn from FreeDesktop.org, XMPP or even
> industry-format specifications such as CD-DA is that even though there
> is "obviously" a single actor heavily involved in the process, having a
> highly-focused descriptive document promotes adoption by abstracting and
> temporarily removing the vendor's influence from the equation.
>
> As a secondary, practical point, we ideally want to publish the
> SOURCE_DATE_EPOCH specification once and never change it, whilst the
> HOWTO could always be improved upon based on feedback and new ideas.
> This entirely disparate attitude towards changesets suggests that they
> should remain separate.
On a second though, I entirely agree with you.
I think I missed the point of the specification during my first thought.
BTW, thanks for taking the time to write this down :)
--
regards,
Mattia Rizzolo
GPG Key: 66AE 2B4A FCCF 3F52 DA18 4D18 4B04 3FCD B944 4540 .''`.
more about me: http://mapreri.org : :' :
Launchpad user: https://launchpad.net/~mapreri `. `'`
Debian QA page: https://qa.debian.org/developer.php?login=mattia `-
signature.asc
Description: Digital signature
___
Reproducible-builds mailing list
Reproducible-builds@lists.alioth.debian.org
http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds
Re: [Reproducible-builds] SOURCE_DATE_EPOCH specification document
Chris Lamb:
> Mattia wrote:
> > > After a hint from doko, I've started work on an official-looking spec
> > > for SOURCE_DATE_EPOCH
> >
> > Lamby pointed out that this is not something debian-related, so it would
> > be great to have it outsite a debian.{org,net} site.
> >
> > OTOH also the howto is not strictly debian-related, and both documents
> > are related to reproducibly.
> >
> > My personal proposal is to merge the two documents and move them to a new
> > more meaningful url (under rb.d.n).
>
> I believe this would handicap adoption of SOURCE_DATE_EPOCH in the wider
> software community.
Strongly seconded. The HOWTO is meant to be a general document about all
things related to “reproducible builds”. It's an “how to”: it's intent
is to describe several way to achieve reproducible builds.
It is way different than a specification which mention the one true way
on how things should work on specific behaviors.
--
Lunar.''`.
lu...@debian.org: :Ⓐ : # apt-get install anarchism
`. `'`
`-
___
Reproducible-builds mailing list
Reproducible-builds@lists.alioth.debian.org
http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds
Re: [Reproducible-builds] SOURCE_DATE_EPOCH specification document
Mattia wrote:
> > After a hint from doko, I've started work on an official-looking spec
> > for SOURCE_DATE_EPOCH
>
> Lamby pointed out that this is not something debian-related, so it would
> be great to have it outsite a debian.{org,net} site.
>
> OTOH also the howto is not strictly debian-related, and both documents
> are related to reproducibly.
>
> My personal proposal is to merge the two documents and move them to a new
> more meaningful url (under rb.d.n).
I believe this would handicap adoption of SOURCE_DATE_EPOCH in the wider
software community.
One thing we can learn from FreeDesktop.org, XMPP or even
industry-format specifications such as CD-DA is that even though there
is "obviously" a single actor heavily involved in the process, having a
highly-focused descriptive document promotes adoption by abstracting and
temporarily removing the vendor's influence from the equation.
As a secondary, practical point, we ideally want to publish the
SOURCE_DATE_EPOCH specification once and never change it, whilst the
HOWTO could always be improved upon based on feedback and new ideas.
This entirely disparate attitude towards changesets suggests that they
should remain separate.
Regards,
--
,''`.
: :' : Chris Lamb
`. `'` la...@debian.org / chris-lamb.co.uk
`-
___
Reproducible-builds mailing list
Reproducible-builds@lists.alioth.debian.org
http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds
Re: [Reproducible-builds] SOURCE_DATE_EPOCH specification document
On Sat, Aug 22, 2015 at 01:15:05PM +0100, Chris Lamb wrote:
> Hi all,
>
> After a hint from doko, I've started work on an official-looking spec
> for SOURCE_DATE_EPOCH and pushed it to source-date-epoch-spec.git on
> Alioth. I've loosely based it on the look and feel of FreeDesktop.org
> specs in case they are familiar to you.
>
> I plan to dump the rest of what's in my head today (depending on the
> level of sun) but other contributions are, of course, welcome.
great.
So, where should we push it?
Lamby pointed out that this is not something debian-related, so it would be
great to have it outsite a debian.{org,net} site.
OTOH also the howto is not strictly debian-related, and both documents are
related to reproducibly.
My personal proposal is to merge the two documents and move them to a new more
meaningful url (under rb.d.n).
--
regards,
Mattia Rizzolo
GPG Key: 66AE 2B4A FCCF 3F52 DA18 4D18 4B04 3FCD B944 4540 .''`.
more about me: http://mapreri.org : :' :
Launchpad user: https://launchpad.net/~mapreri `. `'`
Debian QA page: https://qa.debian.org/developer.php?login=mattia `-
signature.asc
Description: Digital signature
___
Reproducible-builds mailing list
Reproducible-builds@lists.alioth.debian.org
http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds
