Tomahawk,

If you reset project you lose all ready work, plus what's now been stuck in the transfer tab, which has grown to > 16 cores times 2 results. This number blocks new work fetching. Your choice, and no certainty of a fix that compensates for 5 days of computing to report... fetch work somewhere else, since it's WCG and 1 or 2 other projects that have this issue.

The closing tag issue had past discussion. Look for posts by Chrystal Pellet [IIRC] what the workaround was. I'd start with booting the computer and router to see if this unlodges the stuck hinge.

As for the cc_config.xml, it looks good with the 50 per report set [per reporting cycle]:

11/26/2013 5:14:34 AM | World Community Grid | Reporting 50 completed tasks.

Since it's a proved fix, I've added a paragraph to the Ready To Report and linked to the Ingleside post for reference.

Oh and as for BOINC 7.2.31, that was some testing stop gap release. Downgrade to 7.0.64 as that was the previous stable version.

Got no predicting powers as to what will happen, merely what could happen. Resetting or doing a detach/add cycle might get things back in motion, but it could also happen it doesnt. Resetting & detaching will tell the server to consider all results on the machine as dumped!

You could install a second client, after a backup of the BOINC datadir+subs. Uninstall and point the installer to new Program Files and ProgramData folder, call both .\BOINC2 and see. Hope the servers don't recognise the device, or the same ID will be reused and tasks will be considered as lost and possible resend for recrunch.

Why you get these persistent ''No closing tag in scheduler reply'' is a riddle. Try set that 50 value to 1, do a config read in and same time set work fetch to -No- and also, change the related Device profile project selection to something different. Only CEP2 if it was not doing that. Maybe that will ruffle the featers enough.

Your choices.

Shame, but sometimes there is no gain without pain!