There are a number of things we are working to get done before we open up the BOINC eOn project, including various platform binaries and multiple project options server side. On top of this, we want to migrate the statistics from the original eOn system for current and past contributors. In spite of the delays, I am happy to report that we have more (and more skilled) hands aboard the project and have made significant progress with the extra help.
It's great to see the project going again! Any idea when the old stats will be integrated with the new?
Old stats were based on number of workunits done independently of the time and size of them but BOINC credit system is not. How will you incorporate the old stats into the new one? Maybe by just setting on BOINC a wu one point.
Well, the "old" eOn had WU's that varied tremendously as far as required processing power/elapsed time (0.01 s to several minutes) required, and all of the WU's counted for one point. So unless there is a way to incorporate that fact in the conversion to the new scores, maybe the old points should not be converted...
It will also be a hassle, like pointed out in the previous post : for SETI, seperate stats are supplied. For WCG, the scores are (old points) + (BOINC points *7), rendering the XML stats useless as a sole base for project points, and also making points hard to compare between WCG and other projects.
Yes, I have been working old-style eOn, and would like to see some of my efforts there coming up with points in new-style eOn. But it might just be a hassle you can easily avoid
Seems that all that has to be done is assign an "average" value to the old point system as compared with the new and then apply it. Shouldn't be that hard to do.
Is it possible to have two metrics... boinc credits and wus completed. That way wus completed could be combined with old stats. I think something similar is done at WCG... they have boinc credits and they have their own metrics that is combined with their older pre-boinc stats.
graeme wrote:Without a sensible conversion factor or corresponding user names between the projects, we don't have any immediate plans to convert the old stats.
I think the important part is simply converting the team stats. Can't imagine that a conversion factor should be that that difficult, just pick whatever you think is fair and go with it. As offered before, provide a factor and I'll help with the conversion if you're too busy.