I recently completed my first run-this uploaded fine but before it did so the next run downloaded and started processing.

That\'s fine as I was happy to continue assisting the project-however if I had wanted to stop there what should I have done? If I had detached after the first run had uploaded presumably this would have caused the project a problem as the new run would be cancelled? Does the project compensate for this by later downloading the same run to another user?

Thanks

---

BOINC version 4.45 has a button for \'no new work\' which can be applied to each project to prevent further downloads. For 4.19, it is possible to set the disk preferences so work will not download.


So many runs have failed, I don\'t think a further few failed runs (very early on in run) can do much more harm. Also, it isn\'t a big problem for CPDN as WU can be generated at will.

<b>Expanded version</b>
If the failed runs are randomly distributed, it shouldn\'t do any harm loosing runs on download or because people have decided to cease doing CPDN. I would be concerned if there was some selection process that cause the distribution not to be randomly distributed.

Loosing models that have got some way into the model is more harmful. First the computer time is not being used efficiently and it can also be disheartening causing loss of crunchers.

HTH

---

Visit BOINC WIKI for help

And join BOINC Synergy for all the news in one place.

Thanks, but my new run appears to be using 4.13, which again I don\'t understand as I thought that had been superseded.

---

Thanks, but my new run appears to be using 4.13, which again I don\'t understand as I thought that had been superseded.

It\'s very confusing, but BOINC itself and the science applications each have their own set of version numbers. So the current version of the CPDN HadSM3 units being sent out is 4.13, even though you may be using BOINC 4.45.

---