On Wed, 23 Jul 2008, John Peterson wrote:

> On Wed, Jul 23, 2008 at 8:52 AM, Derek Gaston <[EMAIL PROTECTED]> wrote:
>>
>> Sorry for the long winded reply... the short of it is: just use a global
>> variable ;-)
>
> Ouch.  Yeah this needs to not be the "official" way of doing this.
> Unless there is a technical problem with changing the signature of the
> jacobian and residual functions (?) I would second the notion of
> having it take (probably a pointer to, so it could be NULL by default)
> an EquationSystems object.

In that case you might still have to have some global to "remember"
which system in the EquationSystems you were solving, if you had two
loosely coupled systems and wanted to use PETSc on both.  Better to
pass a pointer to a System object, and get the EquationSystems object
from that when necessary.

But anyway, I agree with John's main point: we ought to be passing
more information in to *residual/*jacobian/*matvec.  Using the
existing interface is awkward enough that I suspect even those people
whose code will be broken by an API change will thank us.
---
Roy

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Libmesh-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/libmesh-users

Reply via email to