[Users] meeting minutes for 2016-06-27

Steven R. Brandt sbrandt at cct.lsu.edu
Mon Jun 27 11:11:40 CDT 2016


The pull request for the auto synchronization is here:

https://bitbucket.org/eschnett/carpet/pull-requests/13/presync/diff

These changes helped us resolve a problem with the GPU ML_BSSN, so I'd 
like to not lose them. Thanks.

Cheers,

Steve


On 06/27/2016 10:40 AM, Roland Haas wrote:
> Present: Frank, Eloisa, Matt, Roland, Steve, Zach, Erik,
>
> Using slack to replace the IRC channel:
> * Frank inverstigated slack and used it in a different project and has
>    had good experience so far. A test slack project has been created:
>    https://einsteintoolkit.slack.com if you would to test this please
>    send an email an Frank Loeffler knarf at cct.lsu.edu until we have a
>    more permanent solution (in case we decide to use slack). Please
>    provide your email address and name to appear in slack.
>
> Making grid function storage in HydroBase optional:
> * currently there are a number of grid functions that are not always
>    being used or could be recomputed
> * there is a desire to make storage for those optional with the default
>    being to have storage
> * Roland feels that this can be cumbersome if all client thorns require
>    logic to check for storage rather than being able to rely on a
>    complete set of variables always being present. This applies in
>    particular to eps which can only be recomputed if the EOS is known
>
> Making GRHydro_Initdata independent of HydroBase:
> * agree that this is good idea, would make it useful as ID generator
>    for other ID thorns eg for shock tubes
> * some parts are unfortunately tied to GRHydro and would have to be
>    moved out of the thorn
>
> Reducing number of checkpointed variables in McLachlan:
> * RHS variables are fine to not be checkpointed. There should be a
>    helper thorn ML_BSSN_Helper that adjust the checkpoint tags at runtime
> * constraints need to be checkpointed since they require prolongation
>    to compute which cannot be re-created after recovery
>
> Carpet-preSync by Steve, a pull request exists:
> * an attempt to move the start of the SYNC forward so that a grid
>    function that is computed only in the interior later needed
>    everywhere SYNCing happens automatically and early
>
> TeamSpeak or Mumble as Google Hangout replacement:
> * both use equivalent technology
> * community is split about which one is better
> * hosting at LSU may be possible as IT is interested in this
> * external hosters tend to offer both for little money, will need grant
>    money for this
>
> Yours,
> Roland
>
>
>
> _______________________________________________
> Users mailing list
> Users at einsteintoolkit.org
> http://lists.einsteintoolkit.org/mailman/listinfo/users
>
>
> _______________________________________________
> Users mailing list
> Users at einsteintoolkit.org
> http://lists.einsteintoolkit.org/mailman/listinfo/users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.einsteintoolkit.org/pipermail/users/attachments/20160627/9f6ba59c/attachment.html 


More information about the Users mailing list