dolfin team mailing list archive
-
dolfin team
-
Mailing list archive
-
Message #21447
Re: [noreply@xxxxxxxxxxxxx: [Branch ~dolfin-core/dolfin/rognes] Rev 5663: Add init to child after mesh refinement (why is this needed?!) and]
-
To:
Anders Logg <logg@xxxxxxxxx>
-
From:
"Garth N. Wells" <gnw20@xxxxxxxxx>
-
Date:
Sun, 13 Feb 2011 21:23:36 +0000
-
Cc:
dolfin@xxxxxxxxxxxxxxxxxxx
-
In-reply-to:
<20110213212105.GA30024@eowyn>
-
User-agent:
Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.2.13) Gecko/20101208 Thunderbird/3.1.7
On 13/02/11 21:21, Anders Logg wrote:
> On Sun, Feb 13, 2011 at 09:17:18PM +0000, Garth N. Wells wrote:
>>
>>
>> On 13/02/11 21:02, Anders Logg wrote:
>>> mesh.init() computes all connectivity and should almost never be
>>> called, at least not from inside DOLFIN.
>>>
>>> Instead, the proper init function for the connectivity that is
>>> actually used should be called, like mesh.init(1, 2) or similar.
>>>
>>> Note that calling init functions is not needed when using iterators
>>> since the connectivity is then automatically initialized.
>>>
>>> The mesh that comes out of refine()
>>
>> Which you were going to change to 'adapt.
>
> Yes, but this depends on a merge with Marie's branch, which depends on
> a release of UFC, which depends on you commenting on the proposed
> changes of UFC. ;-)
>
Mark all the proposals as Blueprints for 1.6 - then I can see them all
in one place and add comments.
Garth
> --
> Anders
Follow ups
References