[developers] [pet] PET: Transferring the cm branch to main

[Eric Nichols]

Greetings,

Is PET in a stable state right now? I have started updating my packages on Ubuntu NLP and am wondering which branch to update to.
Removal of redundant dependencies (ECL et al.) is a plus, but doesn't take precedence over stability and compatibility with existing grammars.

Thanks,

Eric Nichols

On Aug 3, 2010, at 9:14 PM, Bernd Kiefer wrote:

> Hi all,
> 
> i've now switched branches, as announced in the previous mail.
> In the repository https://pet.opendfki.de/repos/pet/, there is now:
> 
> - tags/legacy-main     which is the main branch in its last state
> - branches/legacy-cm   which is the cm branch in its last state
> - main                 which is the cm branch in its last state, too
> 
> If you have working copies you'd like to keep (maybe because of
> uncommitted changes), you can use the following `svn switch' statements.
> All commands assume that you have cd'ed into the root of your
> working copy:
> 
> For working copies of pet/main (DON'T MAKE ANY CHANGES HERE!)
> 
> svn switch https://pet.opendfki.de/repos/pet/tags/legacy-main
> 
> For working copies of pet/branches/cm, if you are a developer
> and want to switch to the "new main":
> 
> svn switch https://pet.opendfki.de/repos/pet/main
> 
> For working copies of pet/branches/cm, only if you *absolutely*
> have to stick with the old branch (strongly discouraged)
> 
> svn switch https://pet.opendfki.de/repos/pet/branches/legacy-cm
> 
> If you encounter any problems, please let me know.
> 
> Best,
> 	Bernd
> 
> On 30.07.2010 11:34, Bernd Kiefer wrote:
>> Hi Stephan,
>> 
>> i see your point. So instead of two legacy branches, i'll add a tag
>> (indicating read-only use) to the repository, reflecting the current
>> state of main (for people that have working copies with non-committed
>> changes), and a legacy branch for cm, to support people who need time
>> to move to the new "main".
>> 
>> I've triple-checked the current diffs between main and cm, and i'm now
>> quite confident that nothing is lost by replacing main with cm.
>> 
>> Can we all agree on that?
>> 
>> Best,
>> Bernd
>> 
>> On 29.07.2010 23:34, Stephan Oepen wrote:
>>> hi bernd,
>>> 
>>> do we actually need two legacy branches? i suspect the point in having a
>>> separate branch available, reflecting the code base as of now (i.e.
>>> prior to spring cleaning and refactoring), would be creating the
>>> infrastructure for future revisions on the legacy branch, for example
>>> back-porting critical bug fixes, if need be. it would seem tempting to
>>> me to restrict such legacy maintenance to a single branch, i.e. the
>>> result of merging ‘cm’ and ‘main’ (which yields a proper union of
>>> functionality, for all i can tell). on this model, you could first unify
>>> ‘cm’ and ‘main’, and we would allow users some grace period to confirm
>>> they are happy with the result. before actual spring cleaning and
>>> refactoring is to be committed to the trunk, say in late august, the
>>> unified code could be tagged as ‘legacy’ (or even 1.0); future
>>> maintenance on this branch (or better tag, in my view) could then be
>>> viewed as 1.0.1 and so on, vs. new development on the trunk would work
>>> towards version 1.1 (or even 2.0, if you prefer :-). does this all sound
>>> plausible to you? finally, anyone just wanting to build a binary
>>> reflecting ‘cm’ or ‘main’ as of today can easily do so, as these
>>> revisions remain accessible through SVN.
>>> 
>>> all best, oe
>>> 
>>> 
>>> On 29. juli 2010, at 16.30, Bernd Kiefer <kiefer at dfki.de> wrote:
>>> 
>>>> Hi all,
>>>> 
>>>> by the beginning of next week, i will carry out the first step on
>>>> the PetRoadMap, namely the merging of the cm and the main branch.
>>>> 
>>>> The last merge has been performed by Peter, and i've looked at the
>>>> remaining diffs.
>>>> 
>>>> The next steps are the following:
>>>> 
>>>> - creating two new branches: legacy-main and legacy-cm, which reflect
>>>> the current state of the main and cm branch, respectively
>>>> - superseding main with cm
>>>> 
>>>> This will be done Monday or Tuesday next week, unless there are
>>>> unforeseeable obstacles, or serious objections, if they reach me in
>>>> time.
>>>> 
>>>> Best,
>>>> Bernd
>>>> 
>>>> 
>>>> --
>>>> ----------------------------------------------------------------------
>>>> Bernd Kiefer DFKI GmbH, Stuhlsatzenhausweg, D-66123 Saarbruecken
>>>> kiefer at dfki.de +49-681/85775-5301 (phone) +49-681/85775-5338 (fax)
>>>> ----------------------------------------------------------------------
>>>> Deutsches Forschungszentrum fuer Kuenstliche Intelligenz GmbH
>>>> Trippstadter Strasse 122, D-67663 Kaiserslautern, Germany
>>>> Geschaeftsfuehrung: Prof. Dr. Dr. h.c. mult. Wolfgang Wahlster (Vor-
>>>> sitzender), Dr. Walter Olthoff
>>>> Vorsitzender des Aufsichtsrats: Prof. Dr. h.c. Hans A. Aukes
>>>> Amtsgericht Kaiserslautern, HRB 2313
>> 
> 
> -- 
> ----------------------------------------------------------------------
> Bernd Kiefer     DFKI GmbH,  Stuhlsatzenhausweg,  D-66123 Saarbruecken
> kiefer at dfki.de   +49-681/85775-5301 (phone)   +49-681/85775-5338 (fax)
> ----------------------------------------------------------------------
> Deutsches Forschungszentrum fuer Kuenstliche Intelligenz GmbH
> Trippstadter Strasse 122, D-67663 Kaiserslautern, Germany
> Geschaeftsfuehrung: Prof. Dr. Dr. h.c. mult. Wolfgang Wahlster (Vor-
>                    sitzender), Dr. Walter Olthoff
> Vorsitzender des Aufsichtsrats: Prof. Dr. h.c. Hans A. Aukes
> Amtsgericht Kaiserslautern, HRB 2313