The mailing lists on Launchpad are now officially retired.
Please everyone subscribe to the new mailing list here:
http://fenicsproject.org/mailman/listinfo/fenics
The address for the new list is simply
fen...@fenicsproject.org
Some comments:
- We will have one single list for all
On Thu, Apr 11, 2013 at 11:26:23AM +0100, Patrick Farrell wrote:
On 11/04/13 11:24, Corrado Maurini wrote:
I sent a pull request on bitbucket.
Out of curiosity -- I made a pull request yesterday for dolfin,
which I follow on bitbucket. I didn't get any emails notifying
me of the pull
I have added those meshes to the download script now.
--
Anders
On Wed, Apr 10, 2013 at 03:43:49PM +0200, Johan Hake wrote:
Locally only SubMesh and PETScSNESSolver fails. That is because meshes
in the unit test directory has been removed. The relevant files need to
be uploaded to the data
As previously announced, the git conversion will take place the coming
weekend. For this purpose, the repositories will be frozen at 2pm on
Friday afternoon CET. Everyone at Simula will be eating cake at that
time anyway... :-)
The freezing itself just means that everything that happens to the
On Thu, Mar 28, 2013 at 10:06:11PM +, Florian Rathgeber wrote:
On 27/03/13 21:40, Anders Logg wrote:
On Wed, Mar 27, 2013 at 03:23:53PM +, Florian Rathgeber wrote:
On 26/03/13 22:32, Anders Logg wrote:
On Tue, Mar 26, 2013 at 09:57:00PM +, Florian Rathgeber
wrote: ok. So from
On Sat, Mar 30, 2013 at 10:10:47AM +, Florian Rathgeber wrote:
On 30/03/13 07:55, Anders Logg wrote:
On Thu, Mar 28, 2013 at 10:06:11PM +, Florian Rathgeber wrote:
On 27/03/13 21:40, Anders Logg wrote:
On Wed, Mar 27, 2013 at 03:23:53PM +, Florian Rathgeber wrote:
On 26/03/13
On Thu, Mar 28, 2013 at 02:46:49AM +0100, Roland Siegbert wrote:
On 03/27/2013 10:40 PM, Anders Logg wrote:
On Wed, Mar 27, 2013 at 03:23:53PM +, Florian Rathgeber wrote:
On 26/03/13 22:32, Anders Logg wrote:
Yes, I still haven't found a satisfactory solution that is safe in all
cases
On Thu, Mar 28, 2013 at 09:03:50AM +0100, Roland Siegbert wrote:
On Thu, Mar 28, 2013 at 7:35 AM, Anders Logg l...@simula.no wrote:
And still no one with a feature branch has claimed that it would be a
problem to start with a fresh clone. The only thing we would lose
would be the history
On Thu, Mar 28, 2013 at 09:28:13AM +, Florian Rathgeber wrote:
On 28/03/13 06:35, Anders Logg wrote:
On Thu, Mar 28, 2013 at 02:46:49AM +0100, Roland Siegbert wrote:
On 03/27/2013 10:40 PM, Anders Logg wrote:
On Wed, Mar 27, 2013 at 03:23:53PM +, Florian Rathgeber wrote:
On 26/03
On Wed, Mar 27, 2013 at 03:23:53PM +, Florian Rathgeber wrote:
On 26/03/13 22:32, Anders Logg wrote:
On Tue, Mar 26, 2013 at 09:57:00PM +, Florian Rathgeber wrote:
On 26/03/13 17:02, Anders Logg wrote:
On Tue, Mar 26, 2013 at 04:23:30PM +, Florian Rathgeber
wrote:
On 26/03
On Tue, Mar 26, 2013 at 09:31:57AM +0800, Garth N. Wells wrote:
On 26 March 2013 06:45, Anders Logg l...@simula.no wrote:
In reference to previous discussions on the FEniCS mailing list, are
there any objections to the proposal posted earlier today regarding a
freeze for branch merging
On Tue, Mar 26, 2013 at 06:39:25AM +0100, Marie E. Rognes wrote:
This week and coming weekend (including Monday) is Easter holiday in
Norway in general and at Simula in particular. Please postpone
deadlines by (at least) a week. No other objections.
ok: freeze 5 April, new repos 8 April.
--
.
Does that mean that you have branches that you want to merge? Unless
there are branches to be merged, I don't see the benefit of a delay.
Garth
On 03/25/2013 11:45 PM, Anders Logg wrote:
In reference to previous discussions on the FEniCS mailing list, are
there any
want to merge?
Unless there are branches to be merged, I don't see the benefit
of a delay.
Garth
-- Marie
On 03/25/2013 11:45 PM, Anders Logg wrote:
In reference to previous discussions on the FEniCS mailing
list, are there any objections to the proposal posted earlier
On Tue, Mar 26, 2013 at 09:57:00PM +, Florian Rathgeber wrote:
On 26/03/13 17:02, Anders Logg wrote:
On Tue, Mar 26, 2013 at 04:23:30PM +, Florian Rathgeber wrote:
On 26/03/13 16:11, Anders Logg wrote:
On Tue, Mar 26, 2013 at 04:04:27PM +, Florian Rathgeber wrote:
On 26/03/13
In reference to previous discussions on the FEniCS mailing list, are
there any objections to the proposal posted earlier today regarding a
freeze for branch merging before the pending git conversion?
In detail, the suggestion is to
- Set a window for Friday this week to merge in development
On Thu, Mar 21, 2013 at 02:49:48PM +0100, Johan Hake wrote:
Interesting!
Will you add a hook for autogenerate these in the cmake script or
something a user need to do once to bootstrap a code checkout?
Yes, something will be needed. Johannes suggested to add it to the
build system. Perhaps it
On Fri, Mar 15, 2013 at 04:22:36PM +, Garth N. Wells wrote:
The error reported by the buildbot after revision 7542 is caused by a
bug in CGAL 4.0.2, see
https://bugs.launchpad.net/dolfin/+bug/1134475
Should we require CGAL 4.1?
Fine with me.
Yes.
--
Anders
On Tue, Mar 12, 2013 at 01:21:05AM +0100, Nico Schlömer wrote:
For all you Git lovers out there:
I just set up a (one-way) mirror of Dolfin's public repo on GitHub:
https://github.com/nschloe/dolfin/.
--Nico
Interesting. I was curious and did some timings:
$ time bzr branch lp:dolfin
On Tue, Mar 12, 2013 at 01:02:52PM +0100, Marie E. Rognes wrote:
Can we have a green buildbot please? There's some GenericFile
stuff breaking the demo builds.
Also, there is the failing build because of ExodusFile issues...
Which one? I see only a failed build of the Cahn-Hilliard demo due to
On Tue, Mar 12, 2013 at 01:08:40PM +0100, Martin Sandve Alnæs wrote:
I'm guessing merge commit -r7521 by Dr Anders Logg has something to do
with that :)
Dr. Logg??? Prof. Logg! :-)
Are you sure? I don't think that changeset touched GenericFile and the
Cahn-Hilliard demo builds fine here. I
On Tue, Mar 12, 2013 at 01:08:40PM +0100, Martin Sandve Alnæs wrote:
I'm guessing merge commit -r7521 by Dr Anders Logg has something to do
with that :)
Martin
Can we have a green buildbot please? There's some GenericFile stuff
breaking the demo builds.
Also
On Tue, Mar 12, 2013 at 01:16:57PM +0100, Johan Hake wrote:
Dr. Logg??? Prof. Logg! :-)
Are you sure? I don't think that changeset touched GenericFile and the
Cahn-Hilliard demo builds fine here. I haven't pulled the latest
changesets but will do so now to try.
That is caused by my merge.
On Tue, Mar 12, 2013 at 01:34:46PM +0100, Marie E. Rognes wrote:
Which one? I see only a failed build of the Cahn-Hilliard demo due to
GenericFile issues (due to MeshFunction output merged by Johan?).
I think you've fixed it already (I was referring to this one:
On Tue, Mar 12, 2013 at 01:50:24PM +0100, Nico Schlömer wrote:
I guess the bug was originally caused by my commit about Exodus file
writing, and the fact that the automated testing runs stricter compile
flags than I (and apparently Anders) did on locally.
We could have CMAKE_CXX_FLAGS or the
Date: 12 March 2013 12:20
Subject: [Branch ~dolfin-core/dolfin/trunk] Rev 7526: Tiny bug fix in
ExodusFile.cpp
To: Garth Wells gn...@cam.ac.uk
revno: 7526
committer: Anders Logg logg@glaurung
branch nick: dev
timestamp: Tue 2013-03
On Tue, Mar 12, 2013 at 01:04:49PM +, Garth N. Wells wrote:
Because I didn't spot it when I reviewed the code. I'll replace it
with size_t if it can be done easily.
It will break builds because DOLFIN does not provide a uint typedef.
Trilinos does (outside of a namespace), which I assume
branches with git.
Martin
Den 12. mars 2013 08:17 skrev Anders Logg l...@simula.no følgende:
___
Mailing list: https://launchpad.net/~dolfin
Post to : dolfin@lists.launchpad.net
Unsubscribe : https://launchpad.net/~dolfin
More help : https
On Mon, Mar 11, 2013 at 10:52:05AM +0100, Marie E. Rognes wrote:
Is dolfin.mesh.netgen.py still active or is it considered replaced
by the dolfin/generation/CSG* functionality?
I has been replaced and should be removed.
--
Anders
___
Mailing list:
On Mon, Mar 11, 2013 at 10:52:05AM +0100, Marie E. Rognes wrote:
Is dolfin.mesh.netgen.py still active or is it considered replaced
by the dolfin/generation/CSG* functionality?
But someone should port the LEGO class first.
--
Anders
___
Mailing
On Mon, Mar 11, 2013 at 11:30:21AM +0100, Marie E. Rognes wrote:
But someone should port the LEGO class first.
:-)
The LEGO class seems to be in existence already in CSGGeometries3D.h:
// A standard LEGO brick starting at the point x with (n0, n1)
// knobs and height n2. The height should
On Fri, Mar 08, 2013 at 07:55:19AM +, Garth N. Wells wrote:
No, if there is only one subclasse there's no need to keep it.
I assume the interface is unchanged?
The DirichletBC interface will be unchanged.
ok.
--
Anders
___
Mailing list:
What is the error? Does it fail for both PETSc and uBLAS?
--
Anders
On 7 March 2013 01:44, Garth N. Wells gn...@cam.ac.uk wrote:
On 6 March 2013 23:32, Anders Logg l...@simula.no wrote:
The LinearOperator unit test under la is failing on all the buildbots.
This
is strange as it works
Yes, it's been piecewise constant. :-)
--
Anders
On 7 March 2013 12:59, Garth N. Wells gn...@cam.ac.uk wrote:
On 5 March 2013 21:19, Anders Logg l...@simula.no wrote:
On Tue, Mar 05, 2013 at 05:01:01PM +, Garth N. Wells wrote:
I was thinking about keeping the ones used in the demos
Actually, it breaks for me too. It's the C++ version that breaks. The
Python version works fine. Should be easy to fix.
On 7 March 2013 00:32, Anders Logg l...@simula.no wrote:
The LinearOperator unit test under la is failing on all the buildbots.
This is strange as it works fine for me
On Thu, Mar 07, 2013 at 11:26:44PM +, Garth N. Wells wrote:
With the recent periodic bc clean up, the class BoundaryCondition now
has only one subclass (DirichletBC), which makes it largely redundant.
Any objections to BoundaryCondition being removed?
No, if there is only one subclasse
I might have to take back what I said before about there being no issues
with merging and the append-revisions-only tag, but for a different reason
than the one that came up in the discussion last time.
The merge itself goes fine, but there's a practical issue with merging in
a separate
is now in sync again.
--
Anders
On 6 March 2013 19:02, Martin Sandve Alnæs marti...@simula.no wrote:
cd ../dev
bzr pull ../merge
Martin
On 6 March 2013 17:42, Anders Logg l...@simula.no wrote:
I might have to take back what I said before about there being no issues
with merging
cd dev
bzr merge-into trunk
--
Anders
On 6 March 2013 20:55, Martin Sandve Alnæs marti...@simula.no wrote:
That's what I do if I want to play it safe. A bit inconvenient but it
works for me every time.
Martin
On 6 March 2013 19:13, Anders Logg l...@simula.no wrote:
Hmm... That actually
The LinearOperator unit test under la is failing on all the buildbots. This
is strange as it works fine for me. Is it failing for everyone else?
It was enabled today after I discovered it had been disabled (or maybe
never added to the top level list of unit tests).
--
Anders
On Tue, Mar 05, 2013 at 08:10:41AM +, Garth N. Wells wrote:
I'm pretty sure I did that (forced build of UFC).
I saw the same error as the buildbots yesterday. I think a UFL update
did the trick (but I can't be sure, with so many updates).
Garth
Looks ok now. Some buildbots fail with HDF5
I'd like to remove the meshes (data/meshes) from the repository and
replace them with a web page where those meshes and some more
interesting meshes can be downloaded. ok?
--
Anders
___
Mailing list: https://launchpad.net/~dolfin
Post to :
I was thinking about keeping the ones used in the demos. The ones I
want to remove is the top level directory, which can be replaced by an
online version which can then be allowed to grow.
--
Anders
On Tue, Mar 05, 2013 at 03:42:03PM +0100, Martin Sandve Alnæs wrote:
Maybe you can add a
The buildbot is failing after the merge today. It's working fine here
so I assume the reason is some problem with the synchronization of the
packages on the buildbots, a need for cleaning out the instant cache
or similar.
Is it working fine for the rest of you?
Is there some way to clean out and
ok. Forced Instant bots now. Will force DOLFIN later when those are
green.
--
Anders
On Mon, Mar 04, 2013 at 11:10:57PM +0100, Martin Sandve Alnæs wrote:
Manually force build all instant bots to clean cache. Then force build
dolfin.
Martin
Den 4. mars 2013 23:01 skrev Anders Logg [1
. mars 2013 23:21 skrev Anders Logg [1]l...@simula.no
følgende:
ok. Forced Instant bots now. Will force DOLFIN later when those are
green.
Manually force build all instant bots to clean cache. Then force
build
dolfin.
Martin
Den 4. mars 2013 23:01
4. mars 2013 23:52 skrev Anders Logg [1]l...@simula.no
følgende:
It seems to have the latest changeset, and the header has the new
tabulate_tensor:
[2]http://bazaar.launchpad.net/~ufc-core/ufc/main/view/head:/src/ufc
/ufc.h
Actually you have not pushed ufc
On Wed, Feb 27, 2013 at 07:52:25AM +0100, Johannes Ring wrote:
On Tue, Feb 26, 2013 at 9:24 PM, Anders Logg l...@simula.no wrote:
Here's an interesting picture:
http://fenicsproject.org/pub/images/code_growth.png
It confirms that the code has matured in the last few years
On Wed, Feb 27, 2013 at 02:11:37PM +0100, Johannes Ring wrote:
On Wed, Feb 27, 2013 at 1:50 PM, Anders Logg l...@simula.no wrote:
On Wed, Feb 27, 2013 at 07:52:25AM +0100, Johannes Ring wrote:
On Tue, Feb 26, 2013 at 9:24 PM, Anders Logg l...@simula.no wrote:
Here's an interesting picture
On Wed, Feb 27, 2013 at 02:32:07PM +0100, Johannes Ring wrote:
On Wed, Feb 27, 2013 at 2:26 PM, Johannes Ring joha...@simula.no wrote:
It will be updated during the next ohloh analysis, which, I think,
usually happens daily.
It looks like it was analyzed instantly.
Great! Now the curve is
I don't think they should count --- real men write C++. ;-)
--
Anders
On Wed, Feb 27, 2013 at 06:24:52PM +0100, Johan Hake wrote:
What is SWIG files classified to?
Johan
On 02/27/2013 02:35 PM, Anders Logg wrote:
On Wed, Feb 27, 2013 at 02:32:07PM +0100, Johannes Ring wrote:
On Wed
On Wed, Feb 27, 2013 at 07:56:59PM +0100, Johan Hake wrote:
On 02/27/2013 07:31 PM, Anders Logg wrote:
I don't think they should count --- real men write C++. ;-)
Ouch, that hurt...
I'm kidding of course. SWIG is as tough as it gets and I'm happy I
don't need to touch it.
--
Anders
On Wed, Feb 27, 2013 at 08:52:35PM +0100, Johan Hake wrote:
On 02/27/2013 08:31 PM, Anders Logg wrote:
On Wed, Feb 27, 2013 at 07:56:59PM +0100, Johan Hake wrote:
On 02/27/2013 07:31 PM, Anders Logg wrote:
I don't think they should count --- real men write C++. ;-)
Ouch, that hurt
Here's an interesting picture:
http://fenicsproject.org/pub/images/code_growth.png
It confirms that the code has matured in the last few years, with an
essentially linear growth since 2010.
--
Anders
___
Mailing list: https://launchpad.net/~dolfin
On Thu, Feb 21, 2013 at 10:22:45PM +0100, Johan Hake wrote:
Not sure - it should work reading old MeshFunction files into a size_t
Meshfunction.
Ok, if no one come up with a strong argument for keeping it I vote for
removing it.
I think the only place where we need unsigned int is the uint
Can't we just make all Expression time-dependent by adding a variable
't' that all expressions have (set to 0 by default).
Then a user can either update t manually for all expressions, or
connect them somehow to an object that handles time.
t = Time()
t.add(e0)
t.add(e1)
t.add(e2)
t.set(0.1) //
On Fri, Feb 15, 2013 at 12:04:35PM +0100, Johan Hake wrote:
Thinking just a little more about this I should be able to do what I
want by enabling the following syntax in the CompiledExpression interface:
time = Constant(t, t=0.0)
time_dep_expression = Expression(sin(t), t=time)
Here we need
On Fri, Feb 15, 2013 at 12:11:01PM +0100, Johan Hake wrote:
It would serve my need, but as Marting says what if a an Expression need
several time scales, aso.
If t here was a scalar Constant and passed to Expressions which needed
it, we would have more or less the same solution. However it
On Fri, Feb 15, 2013 at 12:16:40PM +0100, Johan Hake wrote:
In that case, one could think of some fancy class GlobalConstant that
can be used for any such constant and which keeps track of who uses
the constant so it can be updated in a single place.
I think you miss the point. By using the
On Thu, Feb 07, 2013 at 08:17:00AM +, Garth N. Wells wrote:
I'm about to start cleaning up the assemblers, and fixing some bugs
will necessitate some interface changes. For example we have:
/// Assemble tensor
void assemble(GenericTensor A,
const Form a,
ok. So ordering can be changed.
--
Anders
On Wed, Feb 06, 2013 at 09:35:10PM +0100, Joachim Berdal Haga wrote:
Forgot to CC list.
On 6 February 2013 00:00, Joachim Berdal Haga [1]j...@simula.no
Plotting must be updated, but that should be simple enough. Line 344 of
Looks like a very nice piece of work.
--
Anders
On Wed, Feb 06, 2013 at 07:18:34PM +0100, Kent-Andre Mardal wrote:
Nice!!
Kent
Hi,
I'd like to announce a geometric multigrid solver for FEniCS,
[2]https://launchpad.net/fmg
Some performance comparisons can be found in
On Tue, Feb 05, 2013 at 07:33:37PM +, Garth N. Wells wrote:
For mixed/vector-values functions,
GenericFunction::compute_vertex_values(..) returns the values in a
vector as
[u0, u1, u2, u3, u4, . . . . . . , v0, v1, v2, v3, . . . , w0, w1,
w2, w3, . . . .]
This isn't very convenient
@gmail.com
*To:* Anders Logg l...@simula.no
*Cc:* dolfin-dev dolfin@lists.launchpad.net
*Sent:* Monday, January 28, 2013 10:52 PM
*Subject:* Re: [Dolfin] DofMap.tabulate_vertex_map
On 01/28/2013 11:36 PM, Anders Logg wrote:
On Mon, Jan 28, 2013 at 10:33:15PM +, Garth N. Wells wrote
On Mon, Jan 28, 2013 at 10:10:40PM +0100, Johan Hake wrote:
Hello!
I have now added a method to DofMap, which tabulate a map between
vertices and dofs. It will only work for DofMaps with dofs on vertices.
So basically for any CG 1 FunctionSpaces and will raise error else wise.
Usage:
On Mon, Jan 28, 2013 at 09:31:27PM +, Garth N. Wells wrote:
On 28 January 2013 21:22, Anders Logg l...@simula.no wrote:
On Mon, Jan 28, 2013 at 10:10:40PM +0100, Johan Hake wrote:
Hello!
I have now added a method to DofMap, which tabulate a map between
vertices and dofs
On Mon, Jan 28, 2013 at 10:37:46PM +0100, Johan Hake wrote:
On 01/28/2013 10:22 PM, Anders Logg wrote:
On Mon, Jan 28, 2013 at 10:10:40PM +0100, Johan Hake wrote:
Hello!
I have now added a method to DofMap, which tabulate a map between
vertices and dofs. It will only work for DofMaps
On Mon, Jan 28, 2013 at 11:19:13PM +0100, Johan Hake wrote:
On 01/28/2013 11:08 PM, Anders Logg wrote:
On Mon, Jan 28, 2013 at 10:37:46PM +0100, Johan Hake wrote:
On 01/28/2013 10:22 PM, Anders Logg wrote:
On Mon, Jan 28, 2013 at 10:10:40PM +0100, Johan Hake wrote:
Hello!
I have now
On Mon, Jan 28, 2013 at 10:33:15PM +, Garth N. Wells wrote:
On 28 January 2013 22:27, Anders Logg l...@simula.no wrote:
On Mon, Jan 28, 2013 at 11:19:13PM +0100, Johan Hake wrote:
On 01/28/2013 11:08 PM, Anders Logg wrote:
On Mon, Jan 28, 2013 at 10:37:46PM +0100, Johan Hake wrote
Thanks for the example.
This would be a nice example to include in DOLFIN. The following steps
would be necessary in order to do this:
1. It needs to follow the style of other demos wrt copyright notice
etc.
2. We need a corresponding C++ version which should look as much as
the Python demo as
Agree. Let's keep one buildbot for 1.0 if practical.
--
Anders
On Mon, Jan 21, 2013 at 06:54:15PM +0100, Johan Hake wrote:
I think it make sense to have at least one buildbot running 1.0.x.
Johan
On Jan 21, 2013 5:39 PM, Johannes Ring [1]joha...@simula.no wrote:
I'm about
For those interested, you can follow the progress of the ongoing FEniCS
code sprint here:
https://blueprints.launchpad.net/sprints/fenics-multidomain-2013
--
Anders
___
Mailing list: https://launchpad.net/~dolfin
Post to :
B1;3400;0cOn Sat, Dec 22, 2012 at 05:54:08PM +, Garth N. Wells wrote:
I think the class 'Restriction' should be re-named to make the name
more meaningful and representative of what it is. Even better, the
functionally should be put in SubMesh. Reading the description of
both classes, it's
, Anders Logg [1]l...@simula.no wrote:
On Thu, Dec 06, 2012 at 09:26:09AM +, Garth N. Wells wrote:
On Thu, Dec 6, 2012 at 9:20 AM, Anders Logg [2]l...@simula.no
wrote:
bzr branch lp:dolfin clean
cd clean
bzr merge ../path-to-complexly-merged-branch
bzr
Aha.
--
Anders
On Mon, Dec 10, 2012 at 11:59:23AM +0100, Martin Sandve Alnæs wrote:
I had used bzr init instead of init-repo...
On 10 December 2012 11:40, Johannes Ring [1]joha...@simula.no wrote:
On Mon, Dec 10, 2012 at 10:47 AM, Martin Sandve Alnæs
[2]marti...@simula.no
at 11:20 AM, Anders Logg l...@simula.no wrote:
It looks like the way to set that flag has changed on Launchpad. The
old method using sftp does not seem to work anymore.
Instead, I tried the following:
bzr config -d bzr+ssh://bazaar.launchpad.net/+branch/dolfin/
append_revisions_only=true
On Thu, Dec 06, 2012 at 09:26:09AM +, Garth N. Wells wrote:
On Thu, Dec 6, 2012 at 9:20 AM, Anders Logg l...@simula.no wrote:
bzr branch lp:dolfin clean
cd clean
bzr merge ../path-to-complexly-merged-branch
bzr commit
bzr push lp:dolfin
The 'clean' branch is not necessarily
On Thu, Dec 06, 2012 at 10:41:04AM +0100, Anders Logg wrote:
On Thu, Dec 06, 2012 at 09:26:09AM +, Garth N. Wells wrote:
On Thu, Dec 6, 2012 at 9:20 AM, Anders Logg l...@simula.no wrote:
bzr branch lp:dolfin clean
cd clean
bzr merge ../path-to-complexly-merged-branch
bzr commit
in
their repositories.
Martin
On 6 December 2012 10:50, Anders Logg [1]l...@simula.no wrote:
On Thu, Dec 06, 2012 at 10:41:04AM +0100, Anders Logg wrote:
On Thu, Dec 06, 2012 at 09:26:09AM +, Garth N. Wells wrote:
On Thu, Dec 6, 2012 at 9:20 AM, Anders Logg [2]l
300.0 seconds
ConnectionReset reading response for 'Branch.set_last_revision_info',
retrying
--
Anders
On Wed, Dec 05, 2012 at 01:35:47AM +0100, Anders Logg wrote:
On Tue, Dec 04, 2012 at 10:06:58PM +, Garth N. Wells wrote:
On Tue, Dec 4, 2012 at 2:39 PM, Anders Logg l...@simula.no wrote
The buildbots were green yesterday but some of them are failing
again. The problem seems to be a an error appearing at random for
some demos using PETSc in combination with Hypre. Johannes and I
discussed it and the problem is likely a partly broken OpenMPI package
in Ubuntu Precise (12.04).
What
Garth, could you live with setting the append_revisions_only flag?
At least for now.
I'm not trying to make trouble for individual developers, but it will
help to not break the buildbot again.
--
Anders
___
Mailing list: https://launchpad.net/~dolfin
On Tue, Dec 04, 2012 at 10:06:58PM +, Garth N. Wells wrote:
On Tue, Dec 4, 2012 at 2:39 PM, Anders Logg l...@simula.no wrote:
Garth, could you live with setting the append_revisions_only flag?
At least for now.
Yes, if someone can tell me how to get a clean branch without pulling
I want to bring up the removed revisions issue again...
To prevent getting removed revisions changesets into trunk, it's
possible to set a simple configuration flag on the lp branch. There
was quite a bit of resistance to adding this flag last time we
discussed it, and the arguments in favor were
...
It's a solved problem if we set that flag.
--
Anders
Have you tried to fix this on the buildbot side Johannes?
Johan
On 12/03/2012 09:58 AM, Anders Logg wrote:
I want to bring up the removed revisions issue again...
To prevent getting removed revisions changesets into trunk, it's
branch is bound to lp:dolfin.
This seems easier than having an extra scratch branch, but I assume
both are equivalent when it comes to changeset ordering in trunk?
--
Anders
Martin
On 3 December 2012 10:32, Anders Logg [1]l...@simula.no wrote:
On Mon, Dec 03, 2012 at 09:22:55AM
.
On 3 December 2012 11:18, Anders Logg [1]l...@simula.no wrote:
On Mon, Dec 03, 2012 at 11:12:02AM +0100, Martin Sandve Alnæs wrote:
The most important reason in my opinion is to preserve the
ordering of
states the trunk has been in, such that bugs can be tracked down
Most buildbots are now green, except for timeouts on the following buildbots:
http://fenicsproject.org:8010/builders/dolfin-trunk-full-mpich/builds/176/steps/make%20run_unittests/logs/stdio
On Mon, Dec 03, 2012 at 02:41:21PM +0100, Johannes Ring wrote:
On Mon, Dec 3, 2012 at 1:24 PM, Anders Logg l...@simula.no wrote:
Most buildbots are now green, except for timeouts on the following
buildbots:
http://fenicsproject.org:8010/builders/dolfin-trunk-full-mpich/builds/176/steps
On Mon, Dec 03, 2012 at 02:41:21PM +0100, Johannes Ring wrote:
On Mon, Dec 3, 2012 at 1:24 PM, Anders Logg l...@simula.no wrote:
Most buildbots are now green, except for timeouts on the following
buildbots:
http://fenicsproject.org:8010/builders/dolfin-trunk-full-mpich/builds/176/steps
On Mon, Dec 03, 2012 at 10:41:57PM +, Garth N. Wells wrote:
Any objections to slowly removing the binary file output format in
favour of HDF5? An object can be sent into and be retrieved from a
HDF5 file by a string. For large vectors (10^6), the HDF5 output is
almost 3 times faster than
Why is the type called 'sizet' and not 'size_t' in the Python
interface for templated MeshFunctions?
--
Anders
___
Mailing list: https://launchpad.net/~dolfin
Post to : dolfin@lists.launchpad.net
Unsubscribe : https://launchpad.net/~dolfin
More
On Thu, Nov 22, 2012 at 10:41:10AM +, Garth N. Wells wrote:
On Thu, Nov 22, 2012 at 10:22 AM, Anders Logg l...@simula.no wrote:
Why is the type called 'sizet' and not 'size_t' in the Python
interface for templated MeshFunctions?
The appropriate name was raised as a question when
to use uintp, which is the NumPy equivalent, but then we should
consider
using float_ or float64 instead of double, which is unintuitive.
Johan
On 11/22/2012 12:00 PM, Anders Logg wrote:
On Thu, Nov 22, 2012 at 10:41:10AM +, Garth N. Wells wrote:
On Thu, Nov 22
On Thu, Nov 22, 2012 at 07:42:48PM +, Garth N. Wells wrote:
On Thu, Nov 22, 2012 at 7:19 PM, Anders Logg l...@simula.no wrote:
ok, so we all agree on size_t?
Fine with me. We might be able to have uint print an informative
error message from within the SWIG interface.
Sounds good.
I
I have added experimental support for restricted function spaces. It
is currently only implemented for cell-based restrictions but the plan
is to extend to subspaces on boundaries etc.
The implementation involves renumbering the dofs on the subsets of the
cells included in the restriction and
:41AM +0100, Joachim Berdal Haga wrote:
Ok. To be certain qt is used with dolfin, you can check the presence of
f.x. ctrl-w in the help text - or just try it.
Den 19. nov. 2012 23:49 skrev Anders Logg [1]l...@simula.no
følgende:
Thanks for the clarification. The plot-qt demo
On Wed, Nov 21, 2012 at 11:14:50AM +, Garth N. Wells wrote:
On Wed, Nov 21, 2012 at 11:01 AM, Anders Logg l...@simula.no wrote:
The problem seems to be I have QT installed but not QVTK. How should
QVTK be installed? I have built VTK myself using Dorsal. Does it help
to rebuild VTK now
On Mon, Nov 19, 2012 at 08:13:10AM +, Garth N. Wells wrote:
Then configure the backend to support 64 bit integers.
So why do we need size_t? Shouldn't we rather use the LA int type for
everything that needs to be big (and dolfin::uint (typedef for
unsigned int as before) for all other
On Mon, Nov 19, 2012 at 09:04:29AM +, Garth N. Wells wrote:
On Mon, Nov 19, 2012 at 8:41 AM, Anders Logg l...@simula.no wrote:
On Mon, Nov 19, 2012 at 08:13:10AM +, Garth N. Wells wrote:
Then configure the backend to support 64 bit integers.
So why do we need size_t? Shouldn't
1 - 100 of 1739 matches
Mail list logo
