New version Auto-update crash

[dbuske]

I ran the screen saver version. The SS uploaded my latest data, then downloaded the latest auto-update. I rebooted, tried running the screensaver and crash. I got a black dos box that said "Fatal error invalid arguments"

In fact, it crashes just when I bring up the WinXP control panel, by right clicking on the desktop and choosing properties when the folding screensaver is set as the default.

I downloaded the new SS update manually and ran set-up and still the same problem.

[Stardragon]

Can you tell me exactly how you start the program, as well as the precise entry in your error log, starting from the relevant date stamp of starting the software.

[dbuske]

I right click on empty desktop. Choose properties, Then I choose screensaver tab (this is where it crashes) and folding screensaver is the default. Error says in black dos box:

[Null_caption] Fatal Error: [000.000] Missing/Invalid Arguments. For Usage, r
Hit return:

There is no error log in the directory. Unless it has some strange name!

Tried running the install program manually and I get that same error message as above. It does seem to finish install though, but same as above.

I allowed the screensaver to start up on it's own, and got the same error message in the black dos box.

[rmmcdaniell]

I have the same problem on two seperate computers running XP.

[Stardragon]

Please see http://www.blueprint.org/proteinfolding/distributedfolding/distfold_news.html for details on this - there was a small bug which is now fixed. We apologize for any inconvenience.

[dbuske]

It works, but dumbly I deleted the directory, before I got the fixed version. Now it appears I am starting over. Is there any way to get my 230 something generations and 11.46 scores back?

[bwkaz]

Not if you deleted the data files or filelist.txt...

However, you couldn't have gotten all the way to generation 230 on the new client. The client starts over at generation 0 after an autoupdate. The old client should have uploaded all its work, too (unless you were running it in nonet mode?).

[dbuske]

They didn't make it clear that we would all start over with a different protein. So I thought we were on the old one yet. The projects reporting to the general community needs work.

[tpdooley]

We've worked with this particular protein a number of times - with different clients. It makes it possible for them to compare the results of one client to another. We upload all the old client's data and then start working with the new client at generation 0.
The new client is producing points close to 4 times as fast as the old one. And the protein scores are tremendously better, as well. (In just a day and a half, we've broken the 8.8A score from the old client which ran for over a month.)