proposal for common configuration elements for ITAPS services

[seol at scorec.rpi.edu]

> On Mon, 2010-09-20 at 21:18, seol at scorec.rpi.edu wrote:
>
>
>>
>> Switching to have --enable-i*** by default would work for SCOREC S/W.
>> However, we didn't have --enable-ixxx by default
>> on purpose since users cannot take full advantage of SCOREC S/W functionalities through ITAPS.
>
> So that may be a reason to recommend use of FMDB internal interfaces
> over ITAPS, depending on the circumstances. But, I don't think its a
> good reason NOT to build the ITAPS interface to begin with.
>
> Out of curiosity, whats an example of something FMDB can do that is NOT
> available through iMesh/iMeshP? And, what was/is the rationale for not
> making it available through iMesh/iMeshP? This is just a curiosity.

These are the things which I can quickly think of.

- direction of n-order entity with respect to n+1 order upward entity
- mesh verification
- migrating both entity and entity set with tag data
- setting weights of entity/entity set for Zoltan load balancer
- general purpose iterator (e.g. sharable by mesh/geom/field)
- general purpose entity set (e.g. sharable by mesh/geom/field)

Seegyoung

>
> Mark
>
>
> --
> Mark C. Miller, Lawrence Livermore National Laboratory
> ================!!LLNL BUSINESS ONLY!!================
> miller86 at llnl.gov      urgent: miller86 at pager.llnl.gov
> T:8-6 (925)-423-5901    M/W/Th:7-12,2-7 (530)-753-8511
>
>