I have been browsing through the forums today and it has become abundantly clear that there is a lot of pent up frustration on the part of the people donating their resources to this project.
A number of which appear to be ready to pull the plug as they believe that their efforts are going awry.
If the Admin does not start corresponding with these well meaning individuals soon , I am afraid that not getting an answer to the reason why donators are getting so many aborted work units could seriously damage the effectiveness of this project.
I am sure that there are a lot of experienced users donating time to this project that understand the inherant problems that a project of this size can cause, but for rookies it is a lot more difficult to continue without answers.I feel you must give priority to users questions to ensure continued dedication to this excellent project
I am sure your attention to these problems will be very much appreciated by all participants and put a lot of peoples minds at rest

The more comments to this post, the more effective it will be
Thank you in advance for your comments
Graham [Team Europe}

The work unit is placed within a folder named ud_21380_0.dir created by the executable with the same name.

Within it are the working unit files including random number generator for seeding the rosetta program and caching of results which are processed and sent back to the server. There is a file with a .pdb extension that corresponds to the protein data sequence - however it is inaccessible.

Other .ud extension files created once the first result is returned to the server house the users profile information/settings/cpu time etc...

If anyone has found any other info on the workings of the program that are out of the norm ... post it ... something more than just 'my WU terminated after x%' would be useful as we are all aware of this issue.

ATTN Admin - An Error Log would be very useful, as suggested.

As a researcher myself I understand the scale of the project... which in itself is very ambitious... and would like to continue contributing to its success. However I do not understand how global participation can be achieved and project deemed fit to commence without the following:


- Cross platform support - Linux, Mac OS etc...
- Mutiple CPU support
- Error log reporting
- Multi lingual support


These are simple implementations to a program that should have been available at the time of release seeing little user input is required in the first place. The extra time spent developing the application is insignificant to the amount of processing time already lost due its poor functionality.

I hope those involved can rectify what seems like a 'rushed' job.

Hi Rick,

You say I should get credit for the non converged but I don't
In the statistics I have dropped down from 70h of cpu-time per day to 20h.
Eventhough my 4 devices has been running all the time.
I also dropped down from 2000 to 600 point per day.

I now have 3 devices that only go to approx. 4% and then gets a new work unit.

How many "dud" work units can I expect? and why don't I get the points for CPU time?

I've been getting the same results as the previous posters - the first task completed, then numerous aborts, each time reloading about a 1Mbyte file. I have gotten no points for the last several days, despite virtually continuous crunching. I understand that not all the protein folding problems complete, but not getting any "work units" for the work to find this out doesn't make any sense.

I'm deinstalling the WCG program, even though I think the idea has a lot of promise. Email me when the work-unit situation is fixed, then I'll try it again.

admins, thanks for identifying the problem , my machine is now back above the critical three percent

a log file is still a good idea, in case something like this comes up again.