My work unit of ClimateP\' finished phase5 on August 16th apparently with success. But it had previously crashed at about 4% in March due to graphics problems. I was then told that if I restore my WU from a BOINC save made at about 2.5 % and let it continue to 100% with success, the WU will eventually be marked as a success. This not happening after some weeks, I asked in a forum why, people probably not knowing themselves enough about the system suggested that my WU was actually flawed, but could not show me an error code - instead questioned the numeric validity of the results (which didn\'t convince me). Could someone with access to those who do know transmit my Q. why after 100 days the WU has still error?

The BOINC server software is setup for the bulk of the projects, which don\'t allow a restore. So there\'s no provision in it for changing a \'label\' on the results page; what gets there first stays there.

So, if you have a crash, the results says \"Error\", and this doesn\'t change after re-starting from a backup.
But the researchers don\'t use any of these messages, just the data itself.
You can tell from all the graphs being there, that the data is OK.
See this page, for example. Although it does look a bit odd; perhaps a data point was missed.

But the researchers don\'t use any of these messages, just the data itself.
You can tell from all the graphs being there, that the data is OK.

That\'s fine and what I wanted to know

Although it does look a bit odd; perhaps a data point was missed.

What exactly do you mean by the last sentence (after the semi-colon) ?

---

I meant that the drop in the lines to below the base of the graph may be because of missing data. i.e. some of your zips didn\'t make it back to the server.
But it could also have been caused by the actual values, or a problem with the graphing program.
The researchers will know when they look at everything for those models.