I (probably as many others) have several Models listed in the Results Table, which never actually made it onto my Systems.

Therefor, these will sit idle and never complete within their year of time.

Would it be possible for CPDN to allow Users to "purge" or at least mark those Models with a "Reject" so they can be removed and be redistributed ?
Another Idea would be something like a manual Request "send to host again" to solve the Problem.

Second :
Some Models actually finished, but the Server still states Outcome as "Unknown" despite having all Trickles.
Would be nice to allow for some manual interaction to at least trigger the Server to look it up again on the Host to recover any possible transmisstion errors.
Something like "Re-Evaluate", that makes the Server signal the Host to re-send all (or just the missing) Information again.

--------
Since CPDN is the only Project working with Models that are exposed to each individual Host for an extremely long time (and consume by far the most CPU Cycles), I'd be happy to see something like above Options to take some care of troubled Models.

Call me a crackpot, but while I don't care too much about an individual 2h SETI WorkUnit, I 'do' care alot about a 400h to 1000h CPDN Model <img src="http://www.falconfly.de/img/grin_blue.gif">

---

Scientific Network : 44800 MHz - 77824 MB - 1970 GB

I completely agree with you, so I'll sign up on this too :-)

&gt; I (probably as many others) have several Models listed in the Results Table,
&gt; which never actually made it onto my Systems.
&gt;
&gt; Therefor, these will sit idle and never complete within their year of time.
&gt;
AFAIK, these wus are reassigned by the servers to other users after a couple of months if the servers have received no trickle.

---

Arnaud

&gt; AFAIK, these wus are reassigned by the servers to other users after a couple
&gt; of months if the servers have received no trickle.

That was also what I heard, but my oldest "ghost" Model is dated 17 Sep 2004, yet has not been distributed to anyone but me.

My point was primarily, that the Results pages somehow look very "unclean", with far less than ~40% of all Results actually trickling correctly, let alone finishing.
It will end up with alot of ugly "No Reply" Markings, I assume the same will happen even for a few that actually finished but somehow didn't correctly end up on the Server.

If there was a way to recover at least those that finished but never reached the "Done" Status, I'd be a happy Camper. (Backup's of all finished Models are available, minus one that lost 60% of its Data during a crash while backing it up :p )

---

Scientific Network : 44800 MHz - 77824 MB - 1970 GB