Closed philipportner closed 3 weeks ago
Good catch Philip. I just saw your issue and was curious about it, so I checked. I think there is no bug. It is probably a miss documentation/communication on our side. If you check the "default" scheduling strategy, it matches the description of MSTATIC.
We need to fix that. I will get to it ASAP. In any case, now it basically means that if a user does not specify a scheduling technique, MSTATIC will be used which is not all bad as it usually outperforms "simple/naive" STATIC.
Good catch Philip. I just saw your issue and was curious about it, so I checked. I think there is no bug. It is probably a miss documentation/communication on our side. If you check the "default" scheduling strategy, it matches the description of MSTATIC.
We need to fix that. I will get to it ASAP. In any case, now it basically means that if a user does not specify a scheduling technique, MSTATIC will be used which is not all bad as it usually outperforms "simple/naive" STATIC.
Oh right, I didn't realize default:
is actually MSTATIC
, I guess simply adding a case MSTATIC
should be fine. Thanks for pointing this out @jhmkorndorfer !
The modified version of static scheduling,
MSTATIC
, is documented, and exists as a CLI argument.However, this is not implemented in the code.