[CP2K:502] Re: cp2k release goals

[Teodoro Laino]

Hi Axel, hi all,

that's a very good point and you will have my full support for it!
I fully understand that one of the weakness of cp2k is the lack of a  
proper release, that's why
I always was/am for starting releasing the code.

Discussing many times with other developers we always stopped the  
"release" topic because
we all agreed that somehow we had not enough man power to keep a  
developer tree and a
release tree.

But (there's always a but...) if there is anyone out of there willing  
to offer some man-power
in order to FULLY keep the release tree, this would be the keystone  
in order to solve the big issue
related with releasing cp2k.
I Mean with "keeping a release tree" all the kind of "dirty" job  
connected with it (possible bug
fixes, testing on several architectures, etc)? Of course, I think you  
will have some kind of help from
developers (at least from me), but keep in mind that the person  
willing to keep the release should
put most of the work on it.

Regarding where branching the release w.r.t. the developer tree, with  
the present implementation speed
it would be anyway difficult to wait for possible implementation, so  
what I suggest is to fix a date for a release,
branch the release tree, fully "clean" that tree (i.e. take out  
features not fully working and leave only the ones
tested). All other pending implementations will be for the next one  
(we should also decide the period (1 year?) for
keeping a standard release speed).

In any case I would be for setting a set of rules on how to release.  
I'm not an expert about that but a sort of discipline
is absolutely mandatory in my opinion (I'm trying to push it harder  
also for the developers' tree).

Anyway once there will be one offering himself as volunteer we can  
really start to plan everything.
Hope releasing cp2k will be soon a goal!

Ciao,
teo

On 4 Jan 2008, at 12:59, Jörg Saßmannshausen wrote:

>
> Dear all,
>
> Happy New Year as well!
>
> I second Alex' suggestion and I volunteer for the different platforms.
> I can offer:
> AMD: Opteron and AMDx2 (debian etch, 64 bit)
> Intel: Xeon (debian etch, 32bit, 2.80GHz) and some old PIII (debian  
> etch,
> Coppermine, 866 MHz).
> I got access to the latest Intel Compiler (10.x) and also some  
> Portland
> compilers (if needed).
> Most of the machines are clustered with either GbE oder 100 Mbit  
> and we are
> using MPICH2 as a MPI (yes, I know about OpenMPI ;-) ).
> Maybe that helps a bit and encourages others to work on the project  
> as well.
>
> All the best from Graz
>
> Jörg
>
>> == 1 of 1 ==
>> Date: Thurs, Jan 3 2008 7:44 am
>> From: Axel
>>
>>
>> happy new year everybody!
>>
>> since the beginning of a new year is also a time for
>> making plans and resolutions, i'd like to revive a subject
>> that has been discussed quite a bit, but is important to
>> many cp2k _users_: a cp2k release.
>>
>> IMO cp2k has gained significant maturity over the last
>> year(s) and despite plenty of new and exciting developments,
>> many cp2k users simply need a point of reference version that
>> (super)computing center staff can support for them and where
>> one can write pre-and post-processing tools for without having
>> to change the code every other day or so.
>>
>> in order set the bar not too high, i'd like to simply start
>> a discussion on what features are ready to be used for
>> production or which features people would (realistically!!)
>> like to see in a release and which needs most urgent fixing.
>>
>> once this is settled, the next step would be to gang up and
>> collect and run proper release tests on as many different
>> platforms as possible and finally find people that are willing
>> to maintain a release branch (i.e. carry over bugfixes from
>> the development, maintain backward compatibility).
>>
>> what do you all think?
>>
>> cheers,
>>     axel.
>
> -- 
> *************************************************************
> Jörg Saßmannshausen
> Institut für chemische Technologie organischer Stoffe
> TU-Graz
> Stremayrgasse 16
> 8010 Graz
> Austria
>
> phone: +43 (0)316 873 8954
> fax: +43 (0)316 873 4959
> homepage: http://sassy.formativ.net/
>
> Please avoid sending me Word or PowerPoint attachments.
> See http://www.gnu.org/philosophy/no-word-attachments.html
>
> >