[Dune] [Dune-Commit] dune-istl r1027 - releases/1.2.1/istl/paamg

Martin Nolte nolte at mathematik.uni-freiburg.de
Wed Jun 10 15:32:21 CEST 2009


Hi Markus, hi all,

since this mail went to all Dune developers, I will place the answer here. 
Dune 1.2 has been released already. Hence I think we should not make any 
changes to this branch anymore.

You said your students are using Dune 1.2. Either they are using the published 
tar balls or they are using the svn version. In the first case, simply merging 
to 1.2 would not help because we don't build new tarballs all the time. In the 
second case, they can just switch to the current 1.2.1 bug fix branch to get 
the newest patches.

Maybe the real question is whether we should publish the latest bug fixes as 
soon as possible. On this question I would call for the opinion of the 
developers. In any case, it is probably a good idea for everybody to look for 
patches that should be in 1.2.1 and are not, yet.

Yours,

Martin

Markus Blatt wrote:
> Hi,
> 
> vielen Dank. Unsere Studenten benutzen leider noch 1.2 (ohne
> Unternummer 1). Würdest du den patch da bitte auch noch reinmergen.
> 
> Danke!
> 
> Gruß,
> 
> Markus
> 
> On Wed, Jun 10, 2009 at 09:25:39AM +0200, mnolte at dune-project.org wrote:
>> Author: mnolte
>> Date: 2009-06-10 09:25:39 +0200 (Wed, 10 Jun 2009)
>> New Revision: 1027
>>
>> Modified:
>>    releases/1.2.1/istl/paamg/transfer.hh
>> Log:
>> merge patch 1026 from trunk
>>
>>
>> Modified: releases/1.2.1/istl/paamg/transfer.hh
>> ===================================================================
>> --- releases/1.2.1/istl/paamg/transfer.hh	2009-06-09 22:03:48 UTC (rev 1026)
>> +++ releases/1.2.1/istl/paamg/transfer.hh	2009-06-10 07:25:39 UTC (rev 1027)
>> @@ -134,6 +134,9 @@
>>  									     ParallelInformation<T>& comm)
>>      {
>>        Transfer<V,BlockVector<B>,SequentialInformation>::restrict(aggregates, coarse, fine, SequentialInformation());
>> +      // We need this here to avoid it in the smoothers on the coarse level.
>> +      // There (in the preconditioner d is const.
>> +      comm.project(coarse);
>>      }
>>  
>>      template<class V, class B, class T1, class T2>
>> @@ -151,8 +154,9 @@
>>  									     OwnerOverlapCopyCommunication<T1,T2>& comm)
>>      {
>>        Transfer<V,BlockVector<B>,SequentialInformation>::restrict(aggregates, coarse, fine, SequentialInformation());
>> -      comm.copyOwnerToAll(coarse,coarse);
>> -      //comm.project(coarse);
>> +      // We need this here to avoid it in the smoothers on the coarse level.
>> +      // There (in the preconditioner d is const.
>> +      comm.project(coarse);
>>      }
>>  #endif
>>  	/** @} */
>>
>>
>> _______________________________________________
>> Dune-Commit mailing list
>> Dune-Commit at dune-project.org
>> http://lists.dune-project.org/mailman/listinfo/dune-commit
>>

-- 
Martin Nolte <nolte at mathematik.uni-freiburg.de>

Universität Freiburg                                   phone: +49-761-203-5642
Abteilung für angewandte Mathematik                    fax:   +49-761-203-5632
Hermann-Herder-Straße 10
79104 Freiburg, Germany




More information about the Dune mailing list