stephenbrooks.orgForumMuon1Generalv4.32e
Username: Password:
Search site:
Subscribe to thread via RSS
Stephen Brooks
2003-10-22 08:23:29
I'm pretty sure this is a good build: v4.32e RAR, ZIP.
Try it out... it won't go "gold" on the official download page until Thursday evening.  I'm hoping some of those who have had problems with 4.32d will be able to try this one in the same situations.

It doesn't make any sense: that's why they call it "virtual"
Boots
2003-10-22 13:37:29
No User.txt generated again!

Life the Universe and Everything!
[OCAU] badger
2003-10-22 18:50:15
Shivering with anticipation, I rushed off to put the new version on my k6-400 box.  Fingers trembling I double clicked on muon1_cmdline.bat.  I watched in terror as the dreaded dialog box appeared, its white cross in red circle glaring out at me like an evil eye:

"This program has performed an illegal operation and will be shut down.
If the problem persists, contact the program vendor.

MUON1 executed an invalid instruction in
module MUON1.EXE at 0167:004211a1.
Registers:
EAX=00000001 CS=0167 EIP=004211a1 EFLGS=00010212
EBX=00000000 SS=016f ESP=0068f7f4 EBP=0068f808
ECX=00560593 DS=016f ESI=00000000 FS=344f
EDX=7fc1d7e8 ES=016f EDI=7fc1d7e8 GS=0000
Bytes at CS:EIP:
df e9 dd d8 75 07 c6 05 cc fe 44 00 01 68 e0 ed
Stack dump:
8170afad 817134f0 00580000 00000000 6e000020 0068fdb8 00426705 004501ae 8170af88 0043b200 0044edac 004501bc 0044eda8 004501c3 8170afad 817134f0 "

damn.

www.BadgerMotorsport.tk Proudly sponsored by GRX-Computers
Sir Avanor
2003-10-22 19:38:02
I had that same trouble.  It's easy to fix.  Just create a user.txt file with your user name in it.  Things ought to work fine then, at least they did when I installed 4.32d. I am guessing the same solution will work.

He who liketh long words.
thormodr
2003-10-23 03:07:48
Nope.  Tried that... It's something special wrt the K6 processor.
Stephen Brooks
2003-10-23 03:10:01
This was precisely the reason I didn't put the program up on the main site this evening.  Big Grin I've just fixed a bug early on in the initialisation part of the program, that _may_ have been causing this problem (it's in about the right region of the code).
If you try downloading the links above again, to get v4.32e build 2, does that help?

It doesn't make any sense: that's why they call it "virtual"
[DPC]Stephan202
2003-10-23 05:26:50
I just tried version e on these two configurations:

AMD K6 166Mhz, 48MB EDO, Windows 98 SE
Intel Pentium 150Mhz, 64MB EDO, Windows 98 SE

On both computers the client caused an error right after I launched it.
Seems like it's an EDO or Win98SE problem rather than a K6 problem...

---
Dutch Power Cow.
MOOH!
rluce
2003-10-23 07:19:40
Stephen

I found that if i install v4.32c first and run it then install v4.32d or e into the same directory the program runs fine.  I deleted the old manual send and used the new version of manual send and that also works fine.

user: rluce
[OCAU] badger
2003-10-23 16:33:18
no joy with v432e build 2 either.

www.BadgerMotorsport.tk Proudly sponsored by GRX-Computers
[OCAU] badger
2003-10-23 16:36:32
BTW stephen I can emphathise with you - I have been working on a project for a year, every time I think it is together I find another bug in it, finally got it finished yeasterday!

www.BadgerMotorsport.tk Proudly sponsored by GRX-Computers
Boots
2003-10-23 20:03:35
Morning, Downloaded this morning V4.32eb2. Install in new folder.  Started OK requested user name (produced User.txt).  Produced Queue.txt OK (For first five runs) was then removed by software.  Have run Muon1.exe - Muon1_Cmdline.bat and all seems fine.
Thanks Boots

Win98SE 2.2 P4 Intel 756Meg 80 Gig

Life the Universe and Everything!
SanderT
2003-10-23 21:52:41
Hi there,
The rechecks of results seems to be working, but by the result in the
result.txt is the Mpts 1time and in the resutlts.dat is Mpts ~5times.

results.txt :
tantalumrodz=500;tantalumrodr=185;s1l=998;s1f=999;d1l=000;s2l=992;s2r=799;
s2f=998;d2l=000;s3l=530;s3r=997;s3f=993;d3l=000;s4l=448;s4r=998;s4f=918;
d4l=000;s5l=570;s5r=999;s5f=994;d5l=006;s6l=995;s6r=998;s6f=977;d6l=000;
s7l=999;s7r=997;s7f=998;d7l=000;s8l=995;s8r=995;s8f=997;d8l=000;s9l=997;
s9r=999;s9f=999;d9l=000;s10l=998;s10r=998;s10f=994;d10l=000;s11l=999;
s11r=997;s11f=995;d11l=000;s12l=996;s12r=996;s12f=998;d12l=000;s13l=998;
s13r=999;s13f=999;d13l=000;s14l=999;s14r=952;s14f=997;d14l=000;s15l=991;
s15r=999;s15f=998;d15l=000;s16l=998;s16r=995;s16f=998;d16l=000;s17l=999;
s17r=999;s17f=998;d17l=000;s18l=995;s18r=999;s18f=999;d18l=000;s19l=998;
s19r=999;s19f=999;d19l=000;s20l=999;s20r=996;s20f=997;d20l=034;s21l=997;
s21r=998;s21f=997;d21l=000;s22l=998;s22r=997;s22f=998;d22l=000;s23l=999;
s23r=995;s23f=994;d23l=000;s24l=998;s24r=999;s24f=996;d24l=000;s25l=999;
s25r=993;s25f=999;d25l=001;s26l=999;s26r=993;s26f=999;d26l=273;s27l=997;
s27r=999;s27f=998;d27l=000;s28l=997;s28r=629;s28f=999;d28l=000;sfinall=000;
sfinalf=999;#runs=5;
9.562535 (322.6 Mpts) [v4.32e] {A98CB54B}

results.dat :
tantalumrodz=500;tantalumrodr=185;s1l=998;s1f=999;d1l=000;s2l=992;s2r=799;
s2f=998;d2l=000;s3l=530;s3r=997;s3f=993;d3l=000;s4l=448;s4r=998;s4f=918;
d4l=000;s5l=570;s5r=999;s5f=994;d5l=006;s6l=995;s6r=998;s6f=977;d6l=000;
s7l=999;s7r=997;s7f=998;d7l=000;s8l=995;s8r=995;s8f=997;d8l=000;s9l=997;
s9r=999;s9f=999;d9l=000;s10l=998;s10r=998;s10f=994;d10l=000;s11l=999;
s11r=997;s11f=995;d11l=000;s12l=996;s12r=996;s12f=998;d12l=000;s13l=998;
s13r=999;s13f=999;d13l=000;s14l=999;s14r=952;s14f=997;d14l=000;s15l=991;
s15r=999;s15f=998;d15l=000;s16l=998;s16r=995;s16f=998;d16l=000;s17l=999;
s17r=999;s17f=998;d17l=000;s18l=995;s18r=999;s18f=999;d18l=000;s19l=998;
s19r=999;s19f=999;d19l=000;s20l=999;s20r=996;s20f=997;d20l=034;s21l=997;
s21r=998;s21f=997;d21l=000;s22l=998;s22r=997;s22f=998;d22l=000;s23l=999;
s23r=995;s23f=994;d23l=000;s24l=998;s24r=999;s24f=996;d24l=000;s25l=999;
s25r=993;s25f=999;d25l=001;s26l=999;s26r=993;s26f=999;d26l=273;s27l=997;
s27r=999;s27f=998;d27l=000;s28l=997;s28r=629;s28f=999;d28l=000;sfinall=000;
sfinalf=999;#runs=5;
9.569707 (1611.5 Mpts) [v4.32e] {BF270C64}

In results.txt/dat is only this entry not all five saperated.
I'm using v432e commandline(downloaded 20:25 German time); winXP; AMD Athlon XP 2800 Barton.

hope that helps for Gold version (-;

Sander

[This message was edited by Stephen Brooks on 2003-Oct-24 at 11:34.]
Stephen Brooks
2003-10-24 03:34:14
OK the version now linked to at the top of this thread is 4.32e build 3 (which will probably be the final build).  SanderT's error looks like it's saving the proper result to results.dat (which is the primary file muon1 works from) but then saving only the last result to results.txt - unfortunately still with the "#runs=5" tag.  I'd recommend you stop using that build as soon as possible and upgrade to the new one - also delete results.txt and copy results.dat to be results.txt, so you send only the correctly checked results.

Badger might want to try build 3 as well - I've changed the code so it simply doesn't go where it used to in the part where it was crashing.  Due to the nature of the crash I expect you'll just get it crashing later on, but on the other hand you might be in luck.

It doesn't make any sense: that's why they call it "virtual"
BOHICASETI
2003-10-24 10:24:47
Something that I noticed with this latest build (4.32e build 3) is that it doesn't correctly read the config.txt file.  It seems that it was compiled to just look for config and not config.txt cause it doesn't use my settings in my config.txt and it also creates its own config file with no extension (.txt) to it.  I edited the new config (no extension) file and gave it 2 threads like the original (I have a dual processor system) and noticed that it read from it correctly.  Probably just a simple fix and I doubt you will get any erronious results back from this minor bug.

Friends help friends move....
REAL friends help friends move dead bodies.
SanderT
2003-10-24 10:46:05
Hi,
the autocli.sav seems to be writen right.  But if I stop and start muon,
it doesn't start at the autosaved state.  Muon makes a new genome.

bohicaseti's workaround works.

mfg
Sander
Stephen Brooks
2003-10-24 12:53:45
OK, looks like I better not make that version public just yet.  I should be able to do that Monday evening instead though.  I think the error is just a trivial not-writing-the-right-filename, so renaming config.txt to config ought to work if you prefer the new version.

It doesn't make any sense: that's why they call it "virtual"
[OCAU] badger
2003-10-26 15:15:42
quote:
Originally posted by SanderT:
Hi,
the autocli.sav seems to be writen right.  But if I stop and start muon,
it doesn't start at the autosaved state.  Muon makes a new genome.

bohicaseti's workaround works.

mfg
Sander

actually I have noticed this for a few versions, 432d I think does it too.

I just noticed the different results in results.dat and results.txt also.  I will get the next build and see if it works on my k6...

www.BadgerMotorsport.tk Proudly sponsored by GRX-Computers
[OCAU] badger
2003-10-26 16:05:12
and the survey says: do dum !

MUON1 executed an invalid instruction in
module MUON1.EXE at 0167:004262b0.
Registers:
EAX=00000001 CS=0167 EIP=004262b0 EFLGS=00010202
EBX=00000000 SS=016f ESP=0068f808 EBP=0068f81c
ECX=0056059f DS=016f ESI=00000000 FS=3247
EDX=7fc1d7e8 ES=016f EDI=7fc1d7e8 GS=0000
Bytes at CS:EIP:
df e9 dd d8 75 07 c6 05 bc e7 44 00 01 68 d0 d6
Stack dump:
8170d441 8170aad4 00580000 00000000 6c000020 0068fdb8 00426a56 0044ea7e 8170d41c 0044c13c 0044eaac 0043c0b8 0044ea89 0044d69c 0044ea9c 0044d698

not at the same spot as before.

as for the results.txt vs results.dat, I seem to recall seeing this sort of thing in previous versions, (c,d?) I remember seeing results which should have been rechecked because they were the highest so far, but they were only valued at 3xx mpts instead of 15xx

www.BadgerMotorsport.tk Proudly sponsored by GRX-Computers
[OCAU] badger
2003-10-26 17:20:21
further checking reveals the following:
1. v432b and v432d don't appear to have the results.txt error ( I didn't actually run v432c, but it should be safe).
2. from 109 v432e results I had here, 20 were rechecked.  results.txt versions showed the first sim only, with 1/5 of the mpts.

The total score the 109 was 23439.5 mpt (txt) and 49253.7 mpts (dat) with best results of 9.670644 and 9.5872.19 respectively.  Fortunately these results haven't been submitted yet. 
I did send some over the weekend, but I don't think those machines have v432e yet.

3. my config.txt works fine
4. autosave is no worries today, although i notice if autosave file is called Auto.sav = not work, autocli.sav = works

www.BadgerMotorsport.tk Proudly sponsored by GRX-Computers

[This message was edited by [OCAU] badger on 2003-Oct-27 at 1:53.]

[This message was edited by [OCAU] badger on 2003-Oct-27 at 1:54.]
SanderT
2003-10-26 21:48:07
Hi,
The results.dat vs results.txt problem is not solved with
4.32e build 24.10.2003 12:47.

So I will check se results.txt befor I send the results.  Roll Eyes

Sander
Stephen Brooks
2003-10-27 05:08:03
OK, I've debugged and recompiled again.  I changed the way I handled the results-saving since the old way wasn't actually such a great way of doing it.  Also I've changed "config" to "config.txt" at one point.  Hopefully this will be the real v4.32e.

It doesn't make any sense: that's why they call it "virtual"
Boots
2003-10-27 05:47:36
Stephen.  What do you call this update v4.32eb4?  Running now.  I believe not necessary to send old stuff now!
Boots

Life the Universe and Everything!
Stephen Brooks
2003-10-27 06:48:15
That designation sounds appropriate.

It doesn't make any sense: that's why they call it "virtual"
John Kitchen
2003-10-27 07:57:09
I'd suggest that each variant should have a unique identification, it's not good that there are now some number of different programs all claiming to be 4.32e

Some discipline is required, I think!

(What a novel idea Big Grin)
Stephen Brooks
2003-10-27 09:19:37
I think I'd need to alternate letters - so maybe in future a,c,e... would be 'beta' versions but b,d,f... would be the released versions.
In fact it's not as bad as it looks on the whole project, as I only update the main page plus the thing that makes those "version update" alerts appear, when this conversation-over-bugs is over.  So as it's happened, just a few of the more keen people have ended up downloading the betas, and the majority of users will only be alerted to the 'good' version.  But probably next time I'll make the "beta" status over here more explicit.

It doesn't make any sense: that's why they call it "virtual"
[OCAU] badger
2003-10-27 17:01:18
I just downloaded it again to get build 4, (using the link on the first page of this thread) but the muon1.exe has a build date of 24/10 at 12:47, the same as build 3...

www.BadgerMotorsport.tk Proudly sponsored by GRX-Computers
MetalRush
2003-10-27 17:22:04
Stephen, I found this rule "-1.#IND00 (0.0 Mpts) [v4.32e] {96AA4DD9} " in my results.txt (running WIN-XP, P4-2.4HT, 2 threads).
Does it sound familiar to you?
I saw that a queue file was created, containing that same information.
Since this rule was also, randomly, inserted in my results.txt file I deleted it and hope the results will get back to 'normal'.
If I am not mistaken I am using version 4.32eb2 Wink

Keep smiling... it makes people wonder what you're up to Big Grin
Boots
2003-10-27 17:40:04
MetalRush, See reply in (Just a Question - Boots) re 432b2 from Stephen to me!
Boots

Badger check your Hotmail.
Boots

Life the Universe and Everything!
[OCAU] badger
2003-10-27 18:37:23
ok got V432e4, seems to be running ok, but it doesn't see the save file on startup (which defeats the purpose of having one) I put the save from build 3 (which also does this) in the install dir, and it missed it, so I renamed the file with a .bak then I waited until the client did its first save of the first genome, then stopped it, and restarted, it again didn't realise there was a save file.  I tried renaming it to auto.sav instead, no joy.


boots: thanks!  I just realised that v432e is on the front page for download.  I didn't think Stephen would have officially released it yet...

ps which team are you in?

www.BadgerMotorsport.tk Proudly sponsored by GRX-Computers

[This message was edited by [OCAU] badger on 2003-Oct-28 at 2:49.]
Boots
2003-10-27 20:10:20
badger.
I'm not in team!  Did not no we had any, also how do you pick one?
Boots

Life the Universe and Everything!
[OCAU] badger
2003-10-27 20:33:01
further to the autosave issue, I have noticed that it doesn't actually print a message saying it is searching for a saved file (commandline version obviously) as it has done in past versions.

boots: You can join us at overclockers australia, just change your name in user.txt to "[OCAU] boots" or "boots [OCAU]" and post in the name change thread asking Stephen to change your already submitted results to that name.

You can check out the OCAU muon site at http://muon.mega-reviews.com/
and of course come and join us at the forums http://forums.overclockers.com.au (you will need to register to view or post).  Front page is http://www.overclockers.com.au

www.BadgerMotorsport.tk Proudly sponsored by GRX-Computers
Boots
2003-10-27 21:27:29
Stephen, I believe we are correct that Muon1 does not use autocli.sav, when one restarts.  I observed that it's been written at each 4 min time step.  I though you had got it right this time!  This is more fun, than watching TV!  What's the next Muon1 going to called?
Boots

badger, shall do, and thanks......

Life the Universe and Everything!

[This message was edited by Boots on 2003-Oct-28 at 12:18.]
John Kitchen
2003-10-28 03:45:23
[QUOTE]Originally posted by Boots:
What's the next Muon1 going to called?
[QUOTE]

Same name, just randomly generated functions!  Wink
Stephen Brooks
2003-10-28 11:28:43
quote:
Originally posted by Boots:
What's the next Muon1 going to called?
Muon1 v4.32f***ing annoying.

But seriously.  I'll take your advice and call it 'f' when I fix this autosaving bug.  If it's working fine otherwise then I suppose that's at least pyrrhic partial victory.

Incidentally there's an item on the muon to-do plan that says "Only alert user of new version once it's N days old" where N is 7 by default.  That would mean you'd only get bugged by the client for versions that I leave unchanged for a week.

It doesn't make any sense: that's why they call it "virtual"
MetalRush
2003-10-28 15:06:52
quote:
Originally posted by Boots:
MetalRush, See reply in (Just a Question - Boots) re 432b2 from Stephen to me!
Boots

Badger check your Hotmail.
Boots

Life the Universe and Everything!

Eeeh Boots, I read your question.  But I am using the 'e' version, your question/complaint was about the 'b'/'b2' version.
Look at that line I quoted... it really says [v4.32e]

@ Stephen,
I am nagging again and/or getting old, (too) lame and too sorry for an excuse Big Grin but can you check your logs please.
Seems the results from the following date/time didn't get counted:
20031027-203324 70 results sent to stephenbrooks.org
20031027-203417 36 results sent to stephenbrooks.org
20031027-203907 140 results sent to stephenbrooks.org

There was a beautiful 9.71 in it... Frown

EDIT: could you also check this file pls:
20031029-084125 100 results sent to stephenbrooks.org
It contained several 'weird' entries like I stated above.
I will download a fresh copy of the 4.32e version and hope this will not occur anymore.  Thing is that my system created a queue file (again) with that weird entry (which I believe is the cause of this).  Funny enough, my other systems (running the same version) do not have this 'problem'.

EDIT2: Ok, results are updated again.  Thing i forgot to mention in my first 'EDIT' is that i removed the weird entries in that returned results.txt file.  Just wondering if it was/is useable.

Keep smiling... it makes people wonder what you're up to Big Grin

[This message was edited by MetalRush on 2003-Oct-29 at 8:45.]

[This message was edited by MetalRush on 2003-Oct-29 at 15:45.]
: contact : - - -
E-mail: sbstrudel characterstephenbrooks.orgTwitter: stephenjbrooksMastodon: strudel charactersjbstrudel charactermstdn.io

Site has had 16985861 accesses.