stephenbrooks.orgForumMuon1GeneralMuon V4.21 crashed
Username: Password:
Search site:
Subscribe to thread via RSS
Phippi
2002-08-07 23:14:44
I installed the new client on two computer (Win98 and Win 2000) and both clients crashed within 6 hours.  On the Win98 I had a protection failure of a dll-file and the Win2000 system crashed without a failure message.

Regards,
Daniel
Pascal
2002-08-07 23:22:05
I also had a client crash after 40 minutes on my first machine.  It was an exception failure.  I don't know the file which caused that.

But I think, Stephen will correct this and bring some more features within the following versions (4.22 and so on..) cool
I will have a look at the running time for getting a benchmark of the client.  wink

___________________________
1: Athlon TB-C, 1.2 GC/s, 256 MB DDR-RAM, Erazor x², ADSL-Flatrate, NIC Intel, Win 98 SE Mainboard MSI-6380 Rev.  1
2: Pentium III, 600 MC/s, 256 MB RAM, NIC Intel, Win 98 SE
Bluumi [SwissTeam.NET]
2002-08-07 23:29:59
Same by my Clients (NT / W2K / W98se)
All Crashed.

mad

Good i don't install it in life evoriments big grin

But the SwissTeam is falling back wink
Skygod
2002-08-08 00:32:51
I had the same trouble, crashed on the 2 machines I had installed it on.  Tries to send while offline creating text files containing data.  While online will send but not delete results.txt.

I am back to running 4.2 til I hear some good reports.
rustymax
2002-08-08 00:45:36
My client crashed, too.
I think the client crashed after5 houres.
Pascal
2002-08-08 01:19:16
It's not the problem of the time in general.  My client crashed a few times after several minutes.  Actually it's running since approx.  20 minutes.

___________________________
1: Athlon TB-C, 1.2 GC/s, 256 MB DDR-RAM, Erazor x², ADSL-Flatrate, NIC Intel, Win 98 SE Mainboard MSI-6380 Rev.  1
2: Pentium III, 600 MC/s, 256 MB RAM, NIC Intel, Win 98 SE
gogomaus
2002-08-08 01:25:14
same to me.  After 95 min and 10 runs (very poor results while using a copy of my good result.dat from v42) the prog tried to connect to internet, though autosend was set NO within config.txt.  Hopefully T-Online and WinMe stopped connection after 15 min as has been configured.  Task manager cancelled background.exe due to some dll failure.
I´m back to version 4.2 again.
Hope probs can be managed soon.
Stephen Brooks
2002-08-08 02:28:44
Damn.  Sorry about that... I knew it was a bad idea to upload it without testing the new results management and sending features, but I'd spent the whole day getting the _simulation_ to work, so lost patience.

I've removed v4.21 from the site (although the muon page still has a dud link to it: I can't change that from this computer) so for now you should use http://stephenbrooks.org/muon1v42.zip

Later on today I should be able to upload a real v4.21 (one that works) and I'll post here when I do.  The old v4.21 produced corrupt results and had memory leaks and god-knows-what.  I didn't upload the signal file (the one that brings up message boxes saying "upgrade now!") because I wasn't totally sure of it, which is fortunate.  Another fortunate thing is that the stats database currently doesn't recognise v4.21 as a version, so none of the corrupt results you may have uploaded have been added to it: they're still sitting on the server along with about 12 test upload files I've done.  I'll delete them before I upload the correct version.

--[same to me.  After 95 min and 10 runs (very poor results while using a copy of my good result.dat from v42) the prog tried to connect to internet, though autosend was set NO within config.txt.]--

I'm testing the program with autosend set to "NO" right now to make sure this works correctly.  As for the "very poor results" there's a reason for that: I've made the program only read from results of the same version number as it is, so when people accidentally use a results.dat with v4.13 results in it, the new version is not corrupted by it.  Of course v4.2 and v4.21 are actually compatible, so now I've made it reckon this by looking at the first four characters "v4.2" of the version-identifier string so all v4.2x are treated as alike.

Well results.txt has got to 259KB without it trying to send, so hopefully this version takes proper notice of what is in the config file.  I'm not waiting 8 hours for these things: I've set up the inner simulation loop to only do one timestep per run, so I get to see what happens to the results files in 'timelapse'.

Actually I think now I've tested that all the results-bugs seem to be ironed out (there was a stupid one with forgetting to remove a carriage-return in one place).  I now need to compile the new package.


"As every 11-year-old kid knows, if you concentrate enough Van-der-Graff generators and expensive special effects in one place, you create a spiral space-time whirly thing, AND an interesting plotline"
Stephen Brooks
2002-08-08 02:42:05
Try it now... I can't really make any promises but I have done a bit of testing with the sending stuff and it seems OK.  When a couple of people have managed to send results successfully with this new version I'll send the signal for all the clients to ask their users to upgrade (and then activate this version in the stats D.
If it's still messed up the version 4.2 package is still available at the URL above, but I'm more confident of the 4.21 one now.


"As every 11-year-old kid knows, if you concentrate enough Van-der-Graff generators and expensive special effects in one place, you create a spiral space-time whirly thing, AND an interesting plotline"
Skygod
2002-08-08 04:14:44
My last result update contained a lot of v4.2 and only a few v4.21, any chance of retrieving my good results before deleting?........please smile
Stephen Brooks
2002-08-08 07:06:28
They're gone now frown


"As every 11-year-old kid knows, if you concentrate enough Van-der-Graff generators and expensive special effects in one place, you create a spiral space-time whirly thing, AND an interesting plotline"
Pascal
2002-08-08 07:38:53
Version 4.21 - Background client is running since several hours without problems.  I'm using my old results.dat from version 4.2 (best results
near 0.6 %).
My best results now are just near 0.03% (32 new results produced).

___________________________
1: Athlon TB-C, 1.2 GC/s, 256 MB DDR-RAM, Erazor x², ADSL-Flatrate, NIC Intel, Win 98 SE Mainboard MSI-6380 Rev.  1
2: Pentium III, 600 MC/s, 256 MB RAM, NIC Intel, Win 98 SE
Stephen Brooks
2002-08-09 02:24:37
Are you all using old results.dat files with the new v4.21 client?  It shouldn't cause problems, but there are STILL differences between the files I upload to the server with my copy of it and the files other people upload.  Either they're all using the older version, or something strange is happening.  I will try now uploading a version as [v4.21a] so I can distinguish it from the other corrupt version.


"As every 11-year-old kid knows, if you concentrate enough Van-der-Graff generators and expensive special effects in one place, you create a spiral space-time whirly thing, AND an interesting plotline"
Pascal
2002-08-09 03:28:57
I copied the old result.dat (from version 4.2) into the actual version.  Before changing to a new version I sent all the new results per manualsend to the server. 
Either you say, we can use the old results.dat file with a new version or not.

People feel a bit dissappointed about the different versions 4.21, and so they graded down to version 4.2 again.  And because of that you get results from version 4.2

___________________________
1: Athlon TB-C, 1.2 GC/s, 256 MB DDR-RAM, Erazor x², ADSL-Flatrate, NIC Intel, Win 98 SE Mainboard MSI-6380 Rev.  1
2: Pentium III, 600 MC/s, 256 MB RAM, NIC Intel, Win 98 SE
Bluumi [SwissTeam.NET]
2002-08-09 07:10:59
I don't upgrade any Client.

Perhaps at home .... But no at work....

If the new 4.21 crash ... i don't want go over weekend to restart any Clients...

After Weekend, and if here are all happy, i upgrade ...
Stephen Brooks
2002-08-09 07:22:28
It looks like v4.21 is working properly now.  It's been irritating having these bugs, but this version has sorted out a few other things like people 'losing' blocks of results.  That should not happen so often any more.  Also the multithread version should work properly on multi-processor machines.

Me sorting out those bugs was part of the reason why this version in particular was so difficult to get running - it wasn't just a simulation upgrade, but a sending-system upgrade too, which is much more difficult to test.


"As every 11-year-old kid knows, if you concentrate enough Van-der-Graff generators and expensive special effects in one place, you create a spiral space-time whirly thing, AND an interesting plotline"
HETTATLONGUN
2002-08-09 21:12:35
I had three crashes with muon1v4.2 and one crash with muon1v4.21a. I'll start them then leave them alone.  Sooner or later the computer freezes and I have to press the reset button.  No error messages because the monitor is off.

Running muon1_background each time.  Haven't tried it running the GUI version.

Ahh, morning....Nothing like a good cigar, a cup of coffee and the sound of hot shell casings hitting the deck....HETT[TEP]
Michal Hajicek
2002-08-10 03:50:53
Back to 4.2
Muon1 v 4.21a: no crash, but poor results, version check in results.dat not corrected yet, I think.
[ARS]odessit
2002-08-10 06:55:26
confused
have not got any credit for 4.21a shown in stats.
I've been crunching with XP2000+ and P3m-866
user.txt shows [ARS]Odessit, but I am at the same # of results / place.
Stephen Brooks
2002-08-10 16:34:57
I think we've found out on another thread that the stats bug is now corrected - is there still no evidence of the program reading the v4.2 results from results.dat?


"As every 11-year-old kid knows, if you concentrate enough Van-der-Graff generators and expensive special effects in one place, you create a spiral space-time whirly thing, AND an interesting plotline"
HETTATLONGUN
2002-08-11 06:31:23
I am using the results.dat file from muon1v4.2 with muon1v4.21 and muon1v4.21a without a problem, I think.  I just moved results.dat and user.txt from muon1v4.2 to muon1v4.21, then to muon1v4.21a.

The result file uploaded and is shown in my stats.

Ahh, morning....Nothing like a good cigar, a cup of coffee and the sound of hot shell casings hitting the deck....HETT[TEP]
HETTATLONGUN
2002-08-12 19:43:26
I have been using muon1v421b for the past 14 hours.  It hasn't crashed, yet.  Gonna run it for another day or two, if it'll go that long.

Ahh, morning....Nothing like a good cigar, a cup of coffee and the sound of hot shell casings hitting the deck....HETT[TEP]
: contact : - - -
E-mail: sbstrudel characterstephenbrooks.orgTwitter: stephenjbrooksMastodon: strudel charactersjbstrudel charactermstdn.io RSS feed

Site has had 25164071 accesses.