hmmm,

Not really sure why you have done it this way. It would mean each stats site would have to write a custom parser just for this project. Quite a bit of work.

Why didn't you just use the per-app credit system that's been recently added? (BitcoinUtopia are using it successfully). Then leave csg as the main export in standard xml output and have the per-app credit in the user_work and team_work files done by the main server code.

Just my 2 cents.

I'm conflicted on what to do with these stats over at free-dc. So, interested to get opinions.

I can

a) add the csg stats, also parsing the subproject data, but leave the deprecated projects out there also (which would mean duplicate credits for those that crunched prior to csg)
b) add the csg stats and remove the data from the deprecated projects.
c) not add csg (not really an option as I try to include every BOINC project regardless)

Reasons for a) are that I also track what is called MM's which are metadata quantifiers for things like how many 100k's, 1M's users have across projects. If I remove the deprecated projects, users will lose some of these.

One other option I could consider is to do b) but then include subproject data in the MM calculation, meaning CSG could over up to 4 targets. (one overall and on each for the 3 subprojects).

Bok

I display it and track it in various places.

http://stats.free-dc.org/stats.php?page=user&proj=bit&name=25

Scroll down a little to see subproject breakdown, can then click on the subproject to see top lists etc.

Some other projects have been doing it for a while with their own code. PrimeGrid, Yoyo@Home and NFS@Home. Makes it easier when it's standardized though. ( I would personally like to see the user_work and team_work described in the db_dump.xml)

I have been putting off merging my SubsetSum accounts until the dust has settled. I will admit that I am a glutton for projects so I am leaning towards option (a), but willing to go anyway.

I'm conflicted on what to do with these stats over at free-dc. So, interested to get opinions.

I can

a) add the csg stats, also parsing the subproject data, but leave the deprecated projects out there also (which would mean duplicate credits for those that crunched prior to csg)
b) add the csg stats and remove the data from the deprecated projects.
c) not add csg (not really an option as I try to include every BOINC project regardless)

Reasons for a) are that I also track what is called MM's which are metadata quantifiers for things like how many 100k's, 1M's users have across projects. If I remove the deprecated projects, users will lose some of these.

One other option I could consider is to do b) but then include subproject data in the MM calculation, meaning CSG could over up to 4 targets. (one overall and on each for the 3 subprojects).

Bok

How about a modified version of a)
Add CSG and parse the (now) sub-projects data as you suggest but add say an * after the CSG listing to indicate that the extra data for MMs is now listed under the CSG project?
That way you wont have the duplicate credit problem you allude to.

Would that be workable?

Just a thought.

Conan

For my proposal I am suggesting that the 3 Citizen Science Grid sub-projects (SubsetSum, DNA, Wildlife) be shown under the Citizen Science Grid listing, as you do for Primegrid, yoyo and others, and removed from the main listing as individual projects to stop the duplication of credits.

I am also proposing that some sort of indicator be placed next to the Citizen Science Grid listing on the main page with a reference to where those 3 projects now are and the fact they are still counted towards your MM data.

That indicator could be anything, the usual one is a * .

So that Citizen Science Grid * is the listing and the * is also placed at the top or bottom of the listing with its information about Citizen Science Grid and its sub-projects of SubsetSum, DNA , Wildlife.

I hope that makes a bit more sense, it is hard to put into words, well at least for me anyway.

Conan

I don't think either of those *fix* the issue unfortunately.

But I think what I'm going to do (though still thinking about it) is to have my main credit MM's include subprojects also. This will mean I'll deprecate the wildlife,subsetsum and dna stats (they are in the subproject data for csg anyway) and there can be a lot more entries eligible for the MM's as it will also include the various Primegrid subprojects, yoyo@home projects etc.

I do need a few days to fully think about it and it needs a table change to include cpid on my subproject table (which unfortunately is not on the user_work table, but that's a minor issue as I can link it in).

For the meantime, i'll leave the other projects in as well so there is some duplicate credit, but this will eventually go away..

Could we simply have the 3 projects as it was before?:

DNA@Home
SubsetSum@Home
Wildlife@home

Nothing more, nothing less

Could we simply have the 3 projects as it was before?:

DNA@Home
SubsetSum@Home
Wildlife@home

Nothing more, nothing less

Could we simply have the 3 projects as it was before?:

DNA@Home
SubsetSum@Home
Wildlife@home

Nothing more, nothing less

Could we simply have the 3 projects as it was before?:

DNA@Home
SubsetSum@Home
Wildlife@home

Nothing more, nothing less

Could we simply have the 3 projects as it was before?:

DNA@Home
SubsetSum@Home
Wildlife@home

Nothing more, nothing less

Could we simply have the 3 projects as it was before?:

DNA@Home
SubsetSum@Home
Wildlife@home

Nothing more, nothing less

Stats are not being exported for several days now.

Please have a look at it when you can.

Thanks