I have decided to try release candidate 1 of Linux Mint 8 Helena 64bit on the older pc. It's large wcg score was all that had saved it from being played with before now . I am unhappy that the Koala won't run Einstein, QMC, or Climate. Mint 7, built on the Jackalope does (Jackalope doesn't) so I'll try the new ubuntu variant and see if it has been improved in whatever mysterious way they have previously. What the Mint does to make it so that those projects don't fail to find their output files is beyond my ken. I disliked the colour scheme of the mint 7 (too dark, low contrast, hard to see) is why I don't just put 7 in.I tried Koala cuz I had hoped ubuntu had fixed their own problem.

the Quad is still recycling that task, but if i like the Lint I'll probably put it on both my quads in a day or two , so If you want state files or something, before the erring host vanishes speak up :-)

Hey, Thanks again.
I have tried installing koala the recommended way i have troubles getting the 'setimapview' thingy in (from another box) to read the state file.
I couldn't get Mint 8 to install, (device read errors of hd. )
fedora 12 from the package mgr nor from the berkeley .sh would run qmc, even mint 7 (which is running qmc on another box) the only thing i can think of that distinguishes those 2 3000+ sempron amd's is the one that does work has had boinc upgraded from older versions.

I have not posted on this topic @ any other project (rejected tasks) and the "can't find output file" question i have seen in many forums but no solutions i have followed work (ldd the app and reinstall all dependencies or modifying ownerships and permissions (for instance))

I had not come accross the unsupported problem in any other project and i am weeks behind in my reading of the boinc mailing lists, I am going to drop back to my teams forum (we have a few linux guru) perhaps one of them has some more ideas for me to try.

I will stop pestering you now , thanks for your time and attention.
I will drop back to mention any successes, or other setups that show the same behavior as seen here.