I wonder today(July 19th) why Boinc is computing a workunit, wich is over its report deadline of July 17th. Also it has still workunits with this deadline in queue. Why did'nt Boinc throw them away and stops computing the overdue workunit?

There have been many debates about this over a long period of time through beta about this.

The user really should not be DL and setting his prefs so that quanity of work DL by the host becomes incapable of completing them in that deadline 14 days.

You can just "reset" the project it deletes all local project files and does NOT create a new host during this process.

As a matter of interest what are your prefs and CPU as you have that hidden?

More inportant question is why do you have overdue WUWUs, To big of a cache not running BOINC enough, man this just messes it up for the rest.....

Bah!!! I am in it for the money

> As a matter of interest what are your prefs and CPU as you have that hidden?

The PC mentioned above is a Pentium 1 with 200MHz. My prefs are told to cache 3 to 4 days of work. It is'nt to much in respect of the enormous problems of seti@boinc.

The problem on the mentioned pentium 1 pc arrived cause of transfering the complete boinc install of another client to this old fashioned one. With starting boinc on this pc it down loaded wus without any respect to the estimated completion time. But after i manually performed a new cpu benchmark test it respects its computation estimate.

I asked this question cause I thought boinc should handle this problem of overdue workunits but it seems it did not.

The BOINC client assumes the the work still has value and will run it to completion even though it is past the deadline.

You may still get credit if the resend of the workunit(s) has not been returned yet.

If you terminate the application via task manager it may cause the client to return a comutation error immediately.

John Keck
BOINCing since 2002/12/08