But, I run a cache of 0 (!) days and saw, that 5.01 seemed to be out for several hours. So, I expected to see my clients crunch with 5.01 and checked them again, seeing they got fresh 5.00 ...

Humm 5.00 was good enough for me ...
u may finish them to not lose the credits of that ones partially crunched,
however I will abort all my 5.00 WUs that did not started crunching yet,
so, I can start testing the new version 5.01 asap !

ps: Why u did not reported that 5.00 crashed into the 5.00 bugs thread ?
with a link to each result. this way-> http://url-of-the-crashed-result

Anyway, would be nice Installing the debugger -or- at least upgrading u boinc
[url] http://ralph.bakerlab.org/forum_thread.php?id=166[/url]
*Reead all thread

Carlos, I was on the way to post the errors in 5.00-thread.

But, I run a cache of 0 (!) days and saw, that 5.01 seemed to be out for several hours. So, I expected to see my clients crunch with 5.01 and checked them again, seeing they got fresh 5.00 ...

Sorry, that I didn't upgrade my clients from 5.4.3 to 5.4.4, but since 5.4.4 was announced, I didn't have the time to upgrade :-(

Regarding your post with debugger: I read your post about it, but found it sound very complicated ...

HM, only to your info: If they only forgot to change the vers-number, it looks not really good. I saw several WUs crashing, similar like this:

http://ralph.bakerlab.org/workunit.php?wuid=81221

My hosts run cc 5.4.3, but it seems not to have to do with the cc, as I saw in result-files from other users 5.2.13 ...

Shuold we abort 5.00 WUs ?

Just discovered, that a lot of this 5.00 WUs crashed and wanted to post info as i saw 5.01 ...

Feet1st, The "no finished file" message means Boinc.exe (the daemon), lost contact with Boincmgr.exe (the manager). The move to a network drive may have something to do with it. I assume the manager in on a local puter and the daemon is on the network drive.

tony

I have a WU: HBLR_1.0_1ogw_377_23_1 using rosetta_beta version 499, which only gets to 1.30 % and then goes back to 1.13 % after switching to another project.

Result
Workunit

What should I do now?

It would be nice if there was a little button in the message editor that said "make a link".

So, the question is here, should these be aborted, especially given the fact that 5.00 is out?

You'll need to copy over symsrv.yes as well, otherwise it'll lock up at a EULA dialog MS wants you to agree too before using the symbol server.

:)

----- Rom

Over the last couple of weeks I've been adding features to the BOINC Windows Debugger, which is based on Microsoft's DBGHELP.DLL and Stackwalker.

Many of these changes required a newer version of DBGHELP.DLL than what came with BOINC 5.2.x. which is why I have been suggesting to people to upgrade to 5.4.x.

Target CPU run time 2 hours
Actual CPU Time 7 hours 20 minutes
Done 1.34% HBLR_1.0_2tif_375_169_2
Measured integer speed 4622.17 million ops/sec ´
http://ralph.bakerlab.org/result.php?resultid=85155

What I should do ? % seems to be increasing ...

Damm Seems that this bug, was reintroduced by 4.99

March 16, 2006
We updated the windows application version to 4.93. The new version has a fix for the "leave in memory when switching apps" bug that we would like to test. If you are running on Windows, please set your user preference to remove applications from memory when switching between projects and let us know if you are still experiencing errors in this thread.

After this
Date Location Project ID Message
4/12/2006 11:22:42 PM carlos.cp3 ralph@home 245 Pausing result HBLR_1.0_2tif_375_169_2 (removed from memory)

It reverted back to
done 1.09%
and the CPU time increased to 8 hours 5 minutes

Jeez, this will give a lot of cobblestones if it runs to the end -:!

And it will run ... I am now adding more 1 GB of swap space,
and modifying my preferences to Left applications in memory while preempted
Physical RAM is costly $
*Perhaps someone wants to donate me 1 GB ram ddr pc2100 ?

I hope that bugs, once fixed, remains fixed forever
*It is a nonsense otherwise