Hi,
I guess that the main problem/discussion here is the release schedule:
Amber/AmberTools is scheduled to be released every year while
Amber/pmemd is every two years.
It is most probably a little late for scheduling a Amber/pmemd release
in 2015. It could be stated that starting from Amber16, pmemd releases
will be synchronized with the free stuff, but that would also mean that
people should pay for Amber every year. I don't think that it would be a
wise move.
This says, if we decide for saying AmberTools15 is Amber15, we need to
find a way to make people understand that on even years, Amber has a
major release (including the non free stuff), and that on odd years, it
corresponds more to a minor release (new free stuff only).
I would tend to think that what Daniel suggests should be considered:
what people will get by paying the Amber licence fee in 2015 will be the
2014 version of pmemd (+updates). Downloading the free stuff will be the
2015 version of Amber(Tools). In some ways, the directory tree should
show that. But the risk here is that, at the end, you don't sell
something called Amber but Pmemd...
Gerald.
P.S.: Dave, I'm sorry for mistakenly creating the confusion with my
slides...
On 02/21/2015 12:04 AM, Ross Walker wrote:
>>
>>> On Feb 20, 2015, at 5:23 PM, Ross Walker <ross.rosswalker.co.uk> wrote:
>>>
>>> I believe the 'thing' people pay for should still be called AMBER - Lots of companies have approval for purchasing AMBER - if we suddenly change the name of what they are buying it will generate paperwork etc.
>>>
>>> So I think we should just simply have a free component of AMBER and a premium component. Simple solution would be to get rid of this weird AmberTools Fork - which I never liked in the first place since it just complicated build procedures, meant for a while we had multiple manuals and had to explain things in the tutorials etc and instead just finally merge everything back under AMBER - call it AMBER 15, roll in all the updates so far to pmemd and make the current git tree version of pmemd the AMBER 15 pmemd - move all the AmberTools out of it's own folder and back into the main source folder and then just have it be a standalone thing - consisting of two tar balls - Amber15.tar.bz2 and Amber15Pro.tar.bz2 or something like that (maybe have that include the Amber15 tar ball so we don't have this confusing thing of having to download two different things)
>>
>> I would be on board with the naming convention. But I’m a strong -1 on moving everything into $AMBERHOME/src.
>>
>> This is a *massive* change, and would probably rain down merge conflicts galore on anybody either (a) maintaining their own development branch or (b) that hasn’t pushed their changes to the git repo in several commits.
>>
>> Not to mention the retooling it would require of update_amber, which would somehow have to get tested in advance of the release (and it is *not* easy to test a new system like that).
>>
>> Moving sander was hard enough, and I think there are more places where paths are hard-coded than you realize, and I think you’re underestimating the challenge this will pose.
>
> Ah okay - if this is hard we could defer it.
>
> Would be nice to consider going to a single tar ball for installation though - will defer on the names for now.
>
>> I also don’t think pmemd-master is ready to become the next pmemd release. There’s a lot of pmemd.amoeba-related work and new stuff in pmemd (MC barostat move sets and SGLD on the GPUs) that aren’t ready for release yet.
>>
>
> We could just make it AMBER 14 with patches then - we have a few patches to add soon - two patches - one which I believe Perri shipped to you for release and a second for fixing an NTT=2 bug but we need a week or two to iron out the bugs on that.
>
> All the best
> Ross
>
>
>
> _______________________________________________
> AMBER-Developers mailing list
> AMBER-Developers.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber-developers
>
--
____________________________________________________________________________
Prof. Gerald MONARD
SRSMC, Université de Lorraine, CNRS
Boulevard des Aiguillettes B.P. 70239
F-54506 Vandoeuvre-les-Nancy, FRANCE
e-mail : Gerald.Monard.univ-lorraine.fr
tel. : +33 (0)383.684.381
fax : +33 (0)383.684.371
web : http://www.monard.info
____________________________________________________________________________
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Fri Feb 20 2015 - 15:30:03 PST
