To make up for the differences between our estimations and the current "real world" situation of the project we adjusted some values for the workunit generation ("flops factor" and "runtime variation"). For newly generated workunits (not the ones that already had tasks in progress) this means:
- the runtime variation is estimated lower, i.e. the difference (in credit and flops estimation) between workunits of the same frequency band is smaller
- flops estimation and thus credit was raised (maximum 33%)
We'll keep an eye on this; usually we get mostly the longer running ones at the beginning of a run, which makes averaging a bit difficult.
A typo in the code of the workunit generator that I modified for these adjustments caused some trouble to the project yesterday (unresponsive http server etc.). Sorry for that, everything should be back to normal again.
A typo in the code of the workunit generator that I modified for these adjustments caused some trouble to the project yesterday (unresponsive http server etc.). Sorry for that, everything should be back to normal again.
I mentioned over in the problems and bug reports that I am unable to upload. After seeing your message here, I retried uploading. Still no go...
To make up for the
)
To make up for the differences between our estimations and the current "real world" situation of the project we adjusted some values for the workunit generation ("flops factor" and "runtime variation"). For newly generated workunits (not the ones that already had tasks in progress) this means:
- the runtime variation is estimated lower, i.e. the difference (in credit and flops estimation) between workunits of the same frequency band is smaller
- flops estimation and thus credit was raised (maximum 33%)
We'll keep an eye on this; usually we get mostly the longer running ones at the beginning of a run, which makes averaging a bit difficult.
A typo in the code of the workunit generator that I modified for these adjustments caused some trouble to the project yesterday (unresponsive http server etc.). Sorry for that, everything should be back to normal again.
BM
BM
RE: A typo in the code of
)
I mentioned over in the problems and bug reports that I am unable to upload. After seeing your message here, I retried uploading. Still no go...
RE: I mentioned over in the
)
There doesn't seem to be a problem with the file upload handler on the server side. I also didn't notice a problem on my interactive machines so far.
BM
BM
RE: RE: I mentioned over
)
I'll follow the advice Richard gave when I am able to later today...