Just let me know if you were finally able to run the new 4.04 Windows version via BOINC 4.09 or not. A simple "Yes" or "No" post is fine if you don't want to make a big speech! :-) If it looks promising I will email the 1200 or so people who tried with 98/ME/NT4 to give it another try. Thanks!

Hi Carl,

Just "allowed" my CPDN "BOINC" client to download the new v4.04 files and have adjusted my preferences to let CPDN take the lions share of the processing "time-slice" so I can see quite quickly if I still get 2 hours worth of crunching before the WU gets returned with an error, or whether your kind "fix" has worked.

Will reply here in a days or so's time to report back.

Many thanks for taking the time to create the fix - am sure another 1199 people will also be happier now.!

regds,

Timbo

Hokay, tried re-attaching Marion, WinME, to CP-boinc and it now runs okay past the first 144 timestep 'save' point where it always crashed before - nice one. :-)

'Marion' is currently 80% through phase 3 of a THC Slowdown run, so I've left her running classic CPDN for the moment - let me know if you want me to run the CP-boinc model through a trickle point to check that too...

<a href="http://www.nmvs.dsl.pipex.com/"><img src="http://boinc.mundayweb.com/cpdn/stats.php?userID=6&amp;team=off&amp;trans=off"></a>

great well that sounds good. I ran 10 short runs on ME &amp; NT4 so post-processing &amp; checkpoints and uploading should be fine too.

yeah if you're that far on "classic CPDN" then best to just finish up there. But it seems like this fix is working (it ended up being trivial, once you know the "trick" to apply to 1.2 million lines of source code!) I guess I'll "spam" the 1200 people that tried with Win98/ME/NT4 tomorrow --- although I can't yet help the people who tried to attach with Solaris! :-)

A Note for NT4 users -- you will need the latest (final) service pack (6a I think) to run CPDN/BOINC.

&gt; Hi Carl,
&gt;
&gt; Just "allowed" my CPDN "BOINC" client to download the new v4.04 files and have
&gt; adjusted my preferences to let CPDN take the lions share of the processing
&gt; "time-slice" so I can see quite quickly if I still get 2 hours worth of
&gt; crunching before the WU gets returned with an error, or whether your kind
&gt; "fix" has worked.
&gt;
&gt; Will reply here in a days or so's time to report back.
&gt;

Carl,

Have now been running CPDN version 4.04/BOINC v4.09, under Win98SE, for 6 hours of solid processing time and it's still running with no apparent problems. - Only another 676 hours to go to completion !! (according to the "BOINC" client).

so, seems your fix has been VERY GOOD.

Well done

rgds,

Timbo

great, nice to hear "good news" on Win98 machines.

&gt; Have now been running CPDN version 4.04/BOINC v4.09, under Win98SE, for 6 hours of
&gt; solid processing time and it's still running with no apparent problems. - Only
&gt; another 676 hours to go to completion !! (according to the "BOINC" client).
&gt;
&gt; so, seems your fix has been VERY GOOD.
&gt;
&gt; Well done
&gt;
&gt; rgds,
&gt;
&gt; Timbo



Carl,

Further update (FWIW),

Have now been crunching CPDN under BOINC v4.09 and Win 98SE and have completed 24 hours worth of CPU work and NO PROBLEMS, errors or any other hassle have occurred.

Will leave further feedback in a weeks time, but for now the Win98SE fix is still good.

rgds,

Timbo

Is this the fix that has broken Boinc on my XP computer I wonder?

BOINC 4.09 has been running Seti and LHC@Home just fine on this machine since it came out. I got a message from Carl early yesterday that the new version should run, I detached from Seti and attached to Climateprediction to give the new stuff a bigger share of the pie.

Ran fine for about two days... swapping in and out with LHC@Home as expected with each having 50%. Trickled credit once.

WinME, PIV 1.7 GHz, 512 MB
BOINC 4.09.

Then today, upon resuming got this:

-------

Visual Fortran run-time error:

fontl: severe(30) open failure, unit 6, file CONOUT$

Image PC Routine Line
HADSM3um_4.04_Win 008C0547 Unknown Unknown Unknown
HADSM3um_4.04_Win 008AA216 Unknown Unknown Unknown
HADSM3um_4.04_Win 008A8F25 Unknown Unknown Unknown
HADSM3um_4.04_Win 008A9450 Unknown Unknown Unknown
HADSM3um_4.04_Win 0088E828 Unknown Unknown Unknown
HADSM3um_4.04_Win 0040E10A Unknown Unknown Unknown
KERNEL32.DLL BFF7B9E4 Unknown Unknown Unknown
KERNEL32.DLL BFF7B896 Unknown Unknown Unknown
KERNEL32.DLL BFF7A24F Unknown Unknown Unknown
HADSM3um_4.04_Win 006B006E Unknown Unknown Unknown

Stack trace terminated abnormally.

----------

Rebooting got it going again, and has run for an hour or so since. I'll post a followup when(if) something happens one way or another.

Bill, that's odd, how are you doing with hard drive space?

&gt; Bill, that's odd, how are you doing with hard drive space?
&gt;
&gt;
Plenty... 60 GB drive, 30 free on the drive itself.
BOINC allocation (from the pie chart) -- BOINC Free space 2.91 GB.

//added moments after I hit "post" -- it's been running fine since the reboot. Swapped out with LHC(3 hours in, 3 out). //

Hi Carl,
Saw the posts about CPDN now running on Boinc on win 98se, and thought I'd give it a go with the first of 4 machines I'd like to move over from classic. It's been running for two days now, but I have seen no trickles added to the stats even though its connected 24/7.
Model appears to be running ok (no crashes or the like), and I don't see any messages refering to comunication, could there something wrong? or should I leave it a bit longer?
My xp machine running same prefs would have trickeled several times by now.

Colin

Colin/Carl --

My WinME box has trickled about a half dozen times. I did report a Fortran Runtime error in CPDN in a previous post, but it hasn't reoccured. I had a Windows crash earlier today, but that isn't a great surprise.

Bill

&gt; Hi Carl,
&gt; Saw the posts about CPDN now running on Boinc on win 98se, and thought I'd
&gt; give it a go with the first of 4 machines I'd like to move over from classic.
&gt; It's been running for two days now, but I have seen no trickles added to the
&gt; stats even though its connected 24/7.
&gt; Model appears to be running ok (no crashes or the like), and I don't see any
&gt; messages refering to comunication, could there something wrong? or should I
&gt; leave it a bit longer?
&gt; My xp machine running same prefs would have trickeled several times by now.
&gt;
&gt; Colin
&gt;

Carl --

FWIW, my WinME box finally gave up the ghost. I mentioned in a previous post that I had a Windows crash. Those became more frequent. BOINC was running CPDN and LHC@Home. This machine is my primary desktop machine, so I really can't deal with a huge amount of instability.

Anyway, yesterday, I had a crash, and the box refused to boot -- kept getting a "Windows protection fault" on booting. I was able to use a Windows Restore point, and the machine came back up (of course, the last restore was before the latest CPDN upgrade)... it downloaded a new workunit (and seemed to have erased the work in progress) after the restore. Well, again, this is my primary computer, so I told it to stop that nonsense, and detached. The box has been running just fine since with BOINC and LHC@Home (which may, or may not, be significant) since.

I still have a 3GHz hyperthreading WinXP box running CPDN just fine (50-50 with LHC@home), but can't deal with the problems on the WinME box, so that machine says adios... Sorry

Hi,

had this Visual Fortran errors on my Win98SE machine too.
I noticed CPDN is not stopping after timeslicing. When quit BOINC GUI one CPDN task is still running. Guess which. YES its HADSM3um_4.04_Win which reports this error later.

Bernhard

Carl,
I have seen problems twice on my Windows Me machine, but am trying to find out why - I am not sure the the dotNet framework is correctly installed, and so have been investigating.
I have also noticed another thread showing problems with WindowsMe <a href="http://climateapps2.oucs.ox.ac.uk/cpdnboinc/forum_thread.php?id=1029">here</a>

I should say that my Windows NT4 SP6a machine is working at CPDN nicely (although a little slowly, lol)

I should point out that as a SETI alpha tester I am using BOINC v4.11 on all machines.

Dave

Education is not the filling of a pail, but the lighting of a fire — William Butler Yeats

NO

---

Philip - it would be helpful to tell others a little about your machine, in particular the make and speed of processor, and amount of memory.

---

So far nothing major with 4.04, Sometimes the model won't start from a suspend, so a Boinc client restart fixes that. Today i had the CPU time and the progess line freeze on the client, but checking the gfx on the model she was still running fine. I've only completed 8900+ TS so i still have a ways to go. WinME, AMD64 3200+, 512MbRAM. Boinc v4.13