@ SWL551- .....and you think LaurV's avatar is disturbing! That "head scratching" avatar sure made me creep out.

[QUOTE=kladner;336795]@ SWL551- .....and you think LaurV's avatar is disturbing! That "head scratching" avatar sure made me creep out.[/QUOTE]
head scratching? when did I have that avatar. And laurvz totally disturbs me.... Still!

[QUOTE=swl551;336796]head scratching? when did I have that avatar. And laurvz totally disturbs me.... Still![/QUOTE]

Actually, I should have said "brain scratching.":yucky:

2.5.2 released
 

VERSION 2.5.2
1. Fixed bug where a STALL process alarm would fire even if the process was stopped as part of the scheduler.


Additional tweaks that came out of #1
2. A scheduled START will enable monitoring (if process monitoring is turned on in main configuration)
3. From the process menu a SEND CTRL-C event will disable monitoring (if process monitoring is turned on in main configuration)
4. From the process menu a START process event will enable monitoring (if process monitoring is turned on in main configuration)

Get from [URL]http://mersenneforum.org/misfit/[/URL]

[QUOTE=swl551;336890]VERSION 2.5.2
1. Fixed bug where a STALL process alarm would fire even if the process was stopped as part of the scheduler.


Additional tweaks that came out of #1
2. A scheduled START will enable monitoring (if process monitoring is turned on in main configuration)
3. From the process menu a SEND CTRL-C event will disable monitoring (if process monitoring is turned on in main configuration)
4. From the process menu a START process event will enable monitoring (if process monitoring is turned on in main configuration)

Get from [URL]http://mersenneforum.org/misfit/[/URL][/QUOTE]
Thanks.

I'll install and let you know how it works.

[QUOTE=cardogab7341;336893]Thanks.

I'll install and let you know how it works.[/QUOTE]


Drum role.....................

Not sure how to set up batch files
 

I have everything automated except for the startup of MFAKTC instances. I have four of them, and I would like them to start with the window colors, sizes and positions that arrange them nicely on the screen.

Right now the only way I know to do this is with shortcuts (.lnk files) which allow for the specification (via the LAYOUT tab) of window size, position, background and foreground color. I have the four .lnk files, which point to the appropriate MFAKTC instance, in the Windows Startup directory.

However, this starts them all nearly simultaneously and for some reason causes problems with the EVGA Precision utility which sets the speed of the GPUs. If I start the instances manually with a couple of seconds delay, I don't have this problem.

I know MISFIT can start the instances with a delay. Can I just create a .bat file in each MFAKTC directory that references a .lnk file to pick up the attributes I want? Or should this be done in some better way?

[QUOTE=Chuck;339196]I have everything automated except for the startup of MFAKTC instances. I have four of them, and I would like them to start with the window colors, sizes and positions that arrange them nicely on the screen.

Right now the only way I know to do this is with shortcuts (.lnk files) which allow for the specification (via the LAYOUT tab) of window size, position, background and foreground color. I have the four .lnk files, which point to the appropriate MFAKTC instance, in the Windows Startup directory.

However, this starts them all nearly simultaneously and for some reason causes problems with the EVGA Precision utility which sets the speed of the GPUs. If I start the instances manually with a couple of seconds delay, I don't have this problem.

I know MISFIT can start the instances with a delay. Can I just create a .bat file in each MFAKTC directory that references a .lnk file to pick up the attributes I want? Or should this be done in some better way?[/QUOTE]
Chuck,
Just configure MISFIT to startup the .LNK files. No need for bat files if you already have working shortcuts.

Scott

Maybe a bug, maybe not. MISFIT's work calculator does not accept anything less than 80 in "System GHzDs Output"

[QUOTE=kracker;339204]Maybe a bug, maybe not. MISFIT's work calculator does not accept anything less than 80 in "System GHzDs Output"[/QUOTE]
yo! Kracker.
Not a bug. It is the designated lower limit for the control. I didn't think anything lower was possible on a modern gpu.

[QUOTE=swl551;339205]yo! Kracker.
Not a bug. It is the designated lower limit for the control. I didn't think anything lower was possible on a modern gpu.[/QUOTE]

Ah, I see. it was just that I ran into that when I only had my integrated to run on...(~40 GHz) I see, thanks :smile:

[QUOTE=swl551;337339]Drum role.....................[/QUOTE]

I've been running with the new feature enabled for a few weeks now with no problems.

Thanks for the update.

[QUOTE=swl551;339199]Chuck,
Just configure MISFIT to startup the .LNK files. No need for bat files if you already have working shortcuts. Scott[/QUOTE]

Unfortunately that doesn't work. The references to the .lnk files are converted to the names pointed to by the link, and the cmd windows open to the default settings.

[QUOTE=Chuck;339210]Unfortunately that doesn't work. The references to the .lnk files are converted to the names pointed to by the link, and the cmd windows open to the default settings.[/QUOTE]
You can reset the defaults and Windows will remember them. I ALWAYS use the START command which allows me to add a title to the window making it easier to identify.,

[url]http://support.microsoft.com/kb/126410[/url]

[QUOTE=swl551;339240]You can reset the defaults and Windows will remember them. I ALWAYS use the START command which allows me to add a title to the window making it easier to identify.,

[URL]http://support.microsoft.com/kb/126410[/URL][/QUOTE]

Thank you for trying to help me but I can't get it to work. If I manually type in the \path\mfaktc-win-32.exe-shortcut.lnk in a command window the window opens as desired, but if I put the same thing in a .bat file instead of the program running a file explorer window opens in the specified directory.

It's easier just to click the four shortcuts on the desktop to get the instances started.

[QUOTE=Chuck;339245]Thank you for trying to help me but I can't get it to work. If I manually type in the \path\mfaktc-win-32.exe-shortcut.lnk in a command window the window opens as desired, but if I put the same thing in a .bat file instead of the program running a file explorer window opens in the specified directory.

It's easier just to click the four shortcuts on the desktop to get the instances started.[/QUOTE]

Update — I added the /D switch to the .bat files and that got them to work. I also renamed the links to mfaktc.lnk etc for ease in accessing them.

Things still did not work exactly as I expected. I thought MISFIT was going to automatically start the batch files when it was started, but it does not. When I click the "Start Batch" the windows do open to the positions and settings desired, so that part works.

I can see that automatically trying to start the batch could cause problems (like knowing if they are already running). I believe I can complete this project by adding the four .bat files to the Windows Startup folder, adding a TIMEOUT command to three of the files, specifying, say, a 10, 20 and 30 second delay to stagger the starts.

Attached is the window arrangement I have been working to produce.

Chuck you may find the SpeedController batch system useful
[url]http://mersenneforum.org/showthread.php?t=17915[/url]

or Mini Geek's system [url]http://mersenneforum.org/showthread.php?t=18088[/url]

[QUOTE=swl551;339270]Chuck you may find the SpeedController batch system useful
[URL]http://mersenneforum.org/showthread.php?t=17915[/URL]

or Mini Geek's system [URL]http://mersenneforum.org/showthread.php?t=18088[/URL][/QUOTE]

Thanks but not needed. MFAKTC always runs full bore; I don't do any gaming.

Additional question: If I did a MISFIT "Start Batch" with my four instances already running, would it start an additional four? I guess I could try but why flirt with disaster...

[QUOTE=Chuck;339284]Thanks but not needed. MFAKTC always runs full bore; I don't do any gaming.

Additional question: If I did a MISFIT "Start Batch" with my four instances already running, would it start an additional four? I guess I could try but why flirt with disaster...[/QUOTE]

If MISFIT starts more instances then you have a misconfiguration regarding instance names. I cannot answer it for you... You must flirt with disaster to prove out our setup.

Just try not to hurt anyone....:smile:

In my setup, selecting "Start Batch" in MISFIT, or running the underlying batch files manually more than once will result in multiple instances in different windows.*

Conversely, selecting "Send Ctl-C" when there are no mfaktX instances running produced an error message along the lines of "I'm sorry, Dave, I'm afraid I could not do that. (See actual message below!)

This is Win 7 64 bit running on AMD hardware.

* Obviously, not an optimal situation. Just be sure to kill the newer processes and leave only the oldest running. Or kill 'em all and start over.

[QUOTE=swl551;339290]If MISFIT starts more instances then you have a misconfiguration regarding instance names. I cannot answer it for you... You must flirt with disaster to prove out our setup.

Just try not to hurt anyone....:smile:[/QUOTE]

I tried "Start batch" with the jobs already running, and a window opened that said "Process Launch Failed". Good behavior. No harm done.

If you have at least one instance running, and you try to start more from the START PROCESS menu, MISFIT will give you dialog error. If MISFIT does start more instances the INSTANCE name configuration is incorrect.

Stalled process detection is in charge of telling which ones are not running.

[QUOTE=kladner;339300]In my setup, selecting "Start Batch" in MISFIT, or running the underlying batch files manually more than once will result in multiple instances in different windows.*

Conversely, selecting "Send Ctl-C" when there are no mfaktX instances running produced an error message along the lines of "I'm sorry, Dave, I'm afraid I could not do that. (See actual message below!)

This is Win 7 64 bit running on AMD hardware.

* Obviously, not an optimal situation. Just be sure to kill the newer processes and leave only the oldest running. Or kill 'em all and start over.[/QUOTE]
Kladner,
Does MISFIT STOP [B]all[/B] your instances when using the PROCESS STOP menu. Regardless of how you started them it should stop [B]all [/B]of them.

[QUOTE=swl551;339509]Kladner,
Does MISFIT STOP [B]all[/B] your instances when using the PROCESS STOP menu. Regardless of how you started them it should stop [B]all [/B]of them.[/QUOTE]

Yes. Process Stop has always worked without fail and taken down all instances.

[QUOTE=kladner;339513]Yes. Process Stop has always worked without fail and taken down all instances.[/QUOTE]

Excellent. I just wasn't clear on how to read your previous post.

I was not exactly clear in my statements. :no:

Starting from this morning, I was getting the following error when MISFIT tried to submit results to GPU72. (it was repeated each hour)
[QUOTE]Error during ProcessAutoFetch:Index was out of range. Must be non-negative and less than the size of the collection.
Parameter name: startIndex[/QUOTE]
From looking at the logs, I found this warning page (a snippet of the HTML) as GPU72FETCH_[bunch of characters].HTML, logged each hour:
[QUOTE]<h1>Work assigned past expiry date!</h1>
<p>
Sorry Mini-Geek, but you already have too many assignments, <i>or</i> one or more assignments are past the due date. Please see your <a href="/account/assignments/bydate/">Assignments Report</a> for details.
<p>
Since you joined the project you have done on average 177.67 GHz Days of Trial Factoring work per day.
<p>
You currently have 47 assignments totalling 616.26 GHz Days of work assigned, or 3 days worth based on your history. The oldest is 31 days old.
<p>
In order to have access to additional assignments, please complete some more of the work already assigned <font color="red"><i>(particularily the oldest)</i></font>, or <a href="/contact/">Contact Us</a> and explain why you should be exempted from this automated policy.
[/QUOTE]
I've since noticed the warning, extended the assignment, and regenerated my worktodo to ensure that this old assignment will be completed soon, so MISFIT should start working properly again.

I'm pretty sure that while it was doing this, it was not fetching new work for me (the "47 assignments" figure stayed the same, and MISFIT apparently kept trying to fetch work). I see that this was really mainly GPU72's doing, but MISFIT could've given me a relevant error message.

I can provide the full HTML pages, or other info/logs, if needed.

In short: can haz better error message plz? :cat:

Thanks!

p.s. on investigating why this assignment went so long without completing/reporting, I found that mfaktc processed it, but the last bit level of it wasn't reported, so GPU72 never noted it complete. I'm not sure what caused that, but in any case, that wasn't apparently MISFIT or GPU72's fault (I was using submit_spider.pl at the time, so the fault was probably there or in PrimeNet).

[QUOTE=Mini-Geek;339887]Starting from this morning, I was getting the following error when MISFIT tried to submit results to GPU72. (it was repeated each hour)[/QUOTE]

Actually, technically, this was when your MISFIT instance was fetching work from GPU72 -- results aren't submitted to GPU72 (except through the proxy).

However, thank you! Your bug report has reviled a problem when robots talk to robots. My fault, really. :smile:

I have removed the assignment limitation when MISFIT is in the loop.

I now need to figure out how to tell people that they have outstanding (read: old) assignments. Scott -- in-channel return codes? Or out-of-channel communications (probably via e-mail)?

[QUOTE=chalsall;339891]Actually, technically, this was when your MISFIT instance was fetching work from GPU72 -- results aren't submitted to GPU72 (except through the proxy).

However, thank you! Your bug report has reviled a problem when robots talk to robots. My fault, really. :smile:

I have removed the assignment limitation when MISFIT is in the loop.

I now need to figure out how to tell people that they have outstanding (read: old) assignments. Scott -- in-channel return codes? Or out-of-channel communications (probably via e-mail)?[/QUOTE]

if GPU72 pages could have a clearly defined "error" string I could detect it and pass the error back up through the UI.

something like:

PAGE_PROCESS_ERROR:User has too many assignments
PAGE_PROCESS_ERROR:User not logged in
PAGE_PROCESS_ERROR:Request had invalid parameter, cannot service

Let me know.

Scott

VERSION 2.5.3 released
 

VERSION 2.5.3
1. Improved detection and error dialog if GPU72 returns HTML without usable factor rows in it.

Just curious, will MISFIT work with dropbox?

[QUOTE=kracker;340858]Just curious, will MISFIT work with dropbox?[/QUOTE]
I have no idea.

[QUOTE=swl551;340861]I have no idea.[/QUOTE]

Ok, Thanks :razz:

[QUOTE=kracker;340881]Ok, Thanks :razz:[/QUOTE]
did it work?

[QUOTE=swl551;340967]did it work?[/QUOTE]

No, actually I haven't tried it yet, just was curious if you knew it worked or no.

Uh...I'm a bit puzzled. What do you expect MISFIT to do regarding Dropbox? :huh:

[QUOTE=kladner;340976]Uh...I'm a bit puzzled. What do you expect MISFIT to do regarding Dropbox? :huh:[/QUOTE]

Well, just a experiment using dropbox to "control" a gpu in another computer not connected by LAN... but just a crazy idea haven't tried it yet, life is busy :smile:

@chalsall
 

[QUOTE=swl551;339895]if GPU72 pages could have a clearly defined "error" string I could detect it and pass the error back up through the UI.

something like:

PAGE_PROCESS_ERROR:User has too many assignments
PAGE_PROCESS_ERROR:User not logged in
PAGE_PROCESS_ERROR:Request had invalid parameter, cannot service

Let me know.

Scott[/QUOTE]

I'm assuming no changes to the pages will be made.

MISFIT 2.5.4 beta 1 is available
 

At FLASHJH's request I made all email subject lines configurable.

Get from site. [url]http://mersenneforum.org/misfit/[/url]

Looks nice, it will be useful when I get another gpu that I can't access physically.. Also, is it possible for a "Start Minimized" option or command line? :smile: (The new gpu will probably go in one of the office computers, I think you see what I mean)

[QUOTE=kracker;342490]Looks nice, it will be useful when I get another gpu that I can't access physically.. Also, is it possible for a "Start Minimized" option or command line? :smile: (The new gpu will probably go in one of the office computers, I think you see what I mean)[/QUOTE]

create a windows shortcut and set it to "start minimized"

[QUOTE=swl551;342492]create a windows shortcut and set it to "start minimized"[/QUOTE]

Well Duh. Why didn't I think of that...

Thanks :smile:

MISFIT 2.5.4 RELEASED
 

VERSION 2.5.4
1.All email subject lines are now configurable


Get from [url]http://mersenneforum.org/misfit/downloads/MISFIT/[/url]

This morning I was surprised by error messages from MISFIT. I first replaced whatever I was using with a newly extracted copy of V 2.5.4. This put me into reconfiguring the settings, and I have not been able to get it working again, since. I did not make note of the earliest error messages. However, the one I'm getting now is-
[CODE]7/6/2013 11:15:40 AM UpdateGridStats Error! Analyzing Results.txt rows Error parsing no factor for M66422269 from 2^73 to 2^74 [mfaktc 0.20 barrett76_mul32_gs] for calculation of GHZdays
StartIndex cannot be less than zero.
Parameter name: startIndex
[/CODE]

I'm sure I'm overlooking something. I'm not sure what screwed the pooch, but screwed it is.

[QUOTE=kladner;345436]This morning I was surprised by error messages from MISFIT. I first replaced whatever I was using with a newly extracted copy of V 2.5.4. This put me into reconfiguring the settings, and I have not been able to get it working again, since. I did not make note of the earliest error messages. However, the one I'm getting now is-
[CODE]7/6/2013 11:15:40 AM UpdateGridStats Error! Analyzing Results.txt rows Error parsing no factor for M66422269 from 2^73 to 2^74 [mfaktc 0.20 barrett76_mul32_gs] for calculation of GHZdays
StartIndex cannot be less than zero.
Parameter name: startIndex
[/CODE]I'm sure I'm overlooking something. I'm not sure what screwed the pooch, but screwed it is.[/QUOTE]

There is corruption in one of your results.txt files. Open manually to inspect. Look for corrupted lines, missing CRLF, invalid numbers or anything else that is weird. Send to me if you don't see the issue.

thx

[QUOTE=swl551;345461]There is corruption in one of your results.txt files. Open manually to inspect. Look for corrupted lines, missing CRLF, invalid numbers or anything else that is weird. Send to me if you don't see the issue.

thx[/QUOTE]

OK. Thanks. I'll clear or recreate the files manually.

.....some interval elapses as I mess around with things trying to get it working. I seem to keep loosing configuration info. I will keep after it. Here's another screenshot - I have a similar one which ends with "parameter name: count".

I'll try setting up in an entirely new directory, with new-extracted or created files for mfaktc and misfit. I'll take it on when I'm not quite so distracted by a guest and music, etc.

Well, I've now gone through the independent directory attempt. I still can't get it to work, though the messages have changed (see attached). This error message is followed by the loss of config data. Fields I just filled in, (location, bat files, name to stop, etc.) are blank again.

This makes me think that some OS structure (.net?) has gotten borked somehow. Any ideas?

mmh, maybe a deinstall of the whole .net-framework and install again is the solution ? give it a try.

other drives then E: tested ?

firewall/antivirus maybe block the access ?

a older version of misfit works ?

It was working until a few days back, but that [U]was[/U] an older version, I think. I should try rolling back MISFIT before I take on all of .NET. The latter would be a real bear to deal with as there are so many versions and updates listed.

[QUOTE=kladner;345573]Well, I've now gone through the independent directory attempt. I still can't get it to work, though the messages have changed (see attached). This error message is followed by the loss of config data. Fields I just filled in, (location, bat files, name to stop, etc.) are blank again.

This makes me think that some OS structure (.net?) has gotten borked somehow. Any ideas?[/QUOTE]

1. What version were you on before upgrading?

2. You say losing configuration? Please elaborate...

3. Did inspection of results.txt uncover anything?

4. Manually inspect MISFITconfig.txt does it anything like the attached file?

5. Don't uninstall .net. MISFIT is probably 1 out of 300 apps on your PC that uses .NET. If .NET was corrupted you would already be rebuilding your PC.

6. Do your MFAKTx instances run fine with using MISFIT. (they should)

7. delete or rename MISFITCONFIG.TXT and restart misfit. The wizard should take you through configuration again. Do one simple config and see if it saves correctly. If it cannot save correctly try a different drive. You could have a disk problem.

[QUOTE=swl551;345602]1. What version were you on before upgrading?

2. You say losing configuration? Please elaborate...

3. Did inspection of results.txt uncover anything?

4. Manually inspect MISFITconfig.txt does it anything like the attached file?

5. Don't uninstall .net. MISFIT is probably 1 out of 300 apps on your PC that uses .NET. If .NET was corrupted you would already be rebuilding your PC.

6. Do your MFAKTx instances run fine with using MISFIT. (they should)

7. delete or rename MISFITCONFIG.TXT and restart misfit. The wizard should take you through configuration again. Do one simple config and see if it saves correctly. If it cannot save correctly try a different drive. You could have a disk problem.[/QUOTE]

1) Not sure about previous version. I am currently working with 2.5.0.

2) I have gotten some or all of the configuration file entered through the config editor on quite a few occasions now. I've taken to doing it in stages so I don't commit the entire effort of setting up email, etc. At this point, I make it through setting up the Process Setup tab, and save. The Locations and Instances fields get filled in on the main MISFIT window. At this point I have, at times, gone back into config Editor and entered and saved more information. On this trial, I have exited MISFIT after the first Config save. When I restart MISFIT, I get the error message box "Error applying configurations! Number of copies to create must be strictly positive. Parameter name: count." I have attached screenshots of this to previous posts. When I dismiss this "Info" dialog, a second one appears, "No work directories specified! Use Configuration Editor to correct!" This is the data I have now entered repeatedly, which has disappeared in this way more times than I have counted.
When I dismiss the second dialog box, the fields on the main MISFIT window are again blank. This has occurred with or without deleting MISFITconfig.txt. It just occurred on a different physical drive than previous attempts. In none of these attempts did MISFIT ever display the contents of the worktodo.txt files. The results.txt files all looked perfectly ordinary. I have tried the configuration with empty, or no results.txt files. MISFIT has been able to call the batch files and run mfaktc instances. Mfaktc runs fine independently of MISFIT.

@Kladner
 

I found the cause of "object no set....blah" that would occur in a totally clean directory on the first config save. It is minor.


I cannot debug your issue without a copy of your directory as there are hundreds of lines of code involved in ApplyConfig(). The error you see is the outer most exception and is not really expected to ever be hit.

If you zip up and send me your directory structure exactly as it is I will point my debugger at it and find the answer.

I'll send you a link of where to put your zip file over private message

Scott

Thanks, Scott! I'll get it off to you. :smile:

First Kladner debug test
 

Everything worked. (not using a debug version. just the latest compile)
(sans decrypting your passwords. That I cannot do)



I'll do the second installation later.

[QUOTE=swl551;345674]Everything worked. (not using a debug version. just the latest compile)
(sans decrypting your passwords. That I cannot do)



I'll do the second installation later.[/QUOTE]

Thanks. This sort of seems like there's a kink on my end. When I get a chance this afternoon I'll try again with the latest version.

[QUOTE=kladner;345676]Thanks. This sort of seems like there's a kink on my end. When I get a chance this afternoon I'll try again with the latest version.[/QUOTE]

My mistake. I am not using the latest version. I am using the version in your zip file.

Here is your 2.4.9 install
 

No errors beyond MISFIT not being able to decrypt your passwords.

Here is your install running with 2.5.4
 

See yellow as it even calculated you statistics OK.

[QUOTE=swl551;345690]My mistake. I am not using the latest version. I am using the version in your zip file.[/QUOTE]

That makes sense under the circumstances.

It had been working well for so long that I did not even remember which version I was using. I'll get back to you in a bit on further experiments.

EDIT (just saw your last post): That sure is weird! I never got anything meaningful down in that section of the screen. It still seems like something is strange on this machine. Thanks again for taking time over this.

Version 2.5.5
 

VERSION 2.5.5
1. Added more error messages if errors occur during initial loading of files
2. Fixed a bug where productivity data is trying to be loaded before the directory exists. This would only occur during initial setup.

Get from [url]http://mersenneforum.org/misfit/[/url]

This is a long-overdue follow-up to my earlier reports. I still don't know what specifically was involved in the problems. It is quite clear that it was something wrong with my system, probably in the Registry.

After the last post I went through a search and ran a variety of .NET repairs with Microsoft tools for that purpose. These did not resolve the issue. I was, about the same time, involved in a rebuild of Windows. After one setup and settings transfer, MISFIT was again working. Then I lost it again. I finally drew a connection between the failures and some portion of Norton Utilities 16. It was most likely either the Registry Cleaner or the Speed Disk defragger. Unfortunately, running Speed Disk wipes all System Restore Points, so a rollback was impossible. After another setup and settings restore, MISFIT was again functional. I got rid of NU 16 and have not had the problem again (so far.)

@kladner
 

MISFIT does store a few values in hkeyCurrentUser including the password salt. If deleted the worst condition would be messages relating to inability to decrypt passwords. So Norton definitely was doing more than touching registry.

I am sure all MISFIT users happy to see the root cause being a third party and not bugs in MISFIT... I know I am!

Sorry it took me so long to say so. I really have no clue what happened, but it seemed pretty clear that NU 16 had something to do with it. I've had that suite for some time and used it without problems, but the difficulty has not returned since I ditched it.

VERSION 2.5.5.1
 

VERSION 2.5.5.1
1. Fixed limitation working with WorkToDo.txt files with greater than 32,767 rows. Functionality confirmed with 102,696 rows in a single file.
[B]Thanks to NormanRKN for highlighting this limitation and testing the fix[/B].

Get from [URL]http://mersenneforum.org/misfit/[/URL]

[QUOTE=swl551;347542]VERSION 2.5.5.1
1. Fixed limitation working with WorkToDo.txt files with greater than 32,767 rows.[/QUOTE]

It's hard to imagine ever reaching such a limit.

NO, it´s easy to reach!

actually they are in gimps in manual testing page available numbers with 65bit.

you could reach this limit fast because there a no many ghzd/d with one assignment. so you need thousands of assignments if you use a "offline-pc" or your are handling/feed a few systems.

so MISFIT was not able to handle that number of assignments (split them) and gave errors and stop working until the "allowed" size was not reached.

[B]thank you Scott for the very fast fix.[/B]

Norman

[QUOTE=kladner;346909]Sorry it took me so long to say so. I really have no clue what happened, but it seemed pretty clear that NU 16 had something to do with it. I've had that suite for some time and used it without problems, but the difficulty has not returned since I ditched it.[/QUOTE]

I recently received notification of a class action settlement against Symantec involving Norton Utilities and similar software with other names. The $9 and 3 months of free Norton Anti-virus don't mean jack, but it's clear lots of others had problems with the suite.

[QUOTE=kladner;347575]I recently received notification of a class action settlement against Symantec involving Norton Utilities and similar software with other names. The $9 and 3 months of free Norton Anti-virus don't mean jack, but it's clear lots of others had problems with the suite.[/QUOTE]

To put on the table...

My advise is not to respond to any such "offers". They're simply trying to collect data on you.

IMO, such "class action" contact requests provides no upside to anyone except the sender; almost always never upside to the recipient.

[QUOTE=chalsall;347634]To put on the table...

My advise is not to respond to any such "offers". They're simply trying to collect data on you.

IMO, such "class action" contact requests provides no upside to anyone except the sender; almost always never upside to the recipient.[/QUOTE]

I'm sure you are correct, though this did have all the trappings of a Federal Court notice and the described faults jibed with some of my experiences. However, in this case there is not really anything that Symantec can discover about me that they don't already know. I've been a customer for too long. I suppose that does presume that any part of this was as it claimed to be. But still, any common info on me is undoubtedly easily searchable.

[QUOTE=kladner;347637]I'm sure you are correct, though this did have all the trappings of a Federal Court notice and the described faults jibed with some of my experiences. However, in this case there is not really anything that Symantec can discover about me that they don't already know. I've been a customer for too long. I suppose that does presume that any part of this was as it claimed to be. But still, any common info on me is undoubtedly easily searchable.[/QUOTE]

Class action lawsuits are filed in public court so any case can be verified. I got a check recently for $214 as payout from suite where my local bank was charging customers more than published schedule of fees. I didn't even know about it till a letter from the suing law firm arrived with a check. The letter outlined the suite, the finding and the judgement, the check was issued by the law firm, not the accused bank.
Symantec has been getting sued for decades for one thing or another.

Gross vs Symantec
 

[URL]http://www.computerworld.com/s/article/9223368/Symantec_employs_scareware_sales_tactics_lawsuit_charges[/URL]

The suit alleged that Symantec offered free online tests which ALWAYS FOUND SERIOUS ERRORS!!!!! In other words, they were running the classic Scareware scam.

This seems to be unrelated to my suspicions of bad results from using their Norton Utilities software. Note that I was not lured into buying by the scare tactics. I was lured in by a really cheap offer when I renewed my Internet Security subscription. I have used a number of versions of Norton Utilities over the years so I did not expect problems. I emphasize that these are only suspicions, and that I do not have unequivocal evidence connecting weird behavior by MISFIT to running their registry cleaner and various optimizing functions of the Speed Disk defragmenter.

I do know that it happened twice after a Windows rebuild when I ran the various utilities shortly after. After the second time, I rebuilt yet again after removing NU 16 from the system. Since, MISFIT has worked without problems. I repeat that I had used these functions of the suite for some time previously without these particular malfunctions occurring.

This is circumstantial, at best, but I got really tired of fixing things and did not care to experiment further.

MISFIT source released.
 

The good the bad and the ugly. It is all there on the MISFIT download page.

Please do not interpret the release of the source as a sign I am no longer supporting MISFIT.

[QUOTE=swl551;354460]Please do not interpret the release of the source as a sign I am no longer supporting MISFIT.[/QUOTE]

That's really cool Scott!

Thank you for all your work on this, and the trust involved with releasing the source. :smile:

[QUOTE=chalsall;354473]That's really cool Scott!

Thank you for all your work on this, and the trust involved with releasing the source. :smile:[/QUOTE]

+1 indeed.

Same here! :tu:

+1 from here too! :goodposting: I consider that Scott did a very good move by releasing the sources, in spite of the fact that I was in the "privileged" set of the users from the beginning: Scott gave me the source code to "shut my mouth down", hehe, :razz: when I told him that I won't run exe files from "untrusted" sources, and he continued to send me the sources without my request, for every new version he has done. Yet, this has nothing to do with me "lobbying" the program, the program is very well written, he invested a lot of time into it, and from the user point of view, as I said, since I have MISFIT, I almost forgot to write batch files... no joke! The program is doing everything for you, if is set right. Its utility is better when you have more cards or when you run mfaktX in more computers. Try it and see. What you have to do is set your copies of mfaktX, set MISFIT, go to the beach and relax...

It is true...
 

As Laurv mentioned he was silently providing verification of my builds in-case a concern of my credibility became an issue. I also suspected that getting [B]his[/B] approval was a sure-fire-way to increase MISFIT usage. The more users, the more ideas, the more fun.

Thanks.

MISFIT VERSION 2.5.6
 

VERSION 2.5.6
1. Added option to have Stalled Process Detection suspended on startup. This is useful if you have scheduled start/stop events and MISFIT is starting up but the next scheduled start event's time is greater than the Stalled Process Detection threshold.
2. Added ability to resume Stalled Process Detection regardless of what triggered the suspension. A warning is provided if trying to resume when no MFAKTx instances are running.
3. Added CurrentGHzDaysWorkToDo and StagedGHzDaysWorkToDo to the productivity reports
4. Fixed some minor UI bugs when MISFIT is put into MONITOR MODE

GET FROM [url]http://mersenneforum.org/misfit/[/url]

MISFIT VERSION 2.5.8
 

VERSION 2.5.8
1. Changed the MISFIT support email address to [email]MISFIT.SUPPORT@OUTLOOK.COM[/email]
2. Additional small fixes

GET FROM [URL]http://mersenneforum.org/misfit/[/URL]


Due to the lack of activity in the MISFIT forums I will no longer be checking them.

If something comes up please contact me via email at [email]MISFIT.SUPPORT@OUTLOOK.COM[/email]

[QUOTE=swl551;365292]VERSION 2.5.8
1. Changed the MISFIT support email address to [EMAIL="MISFIT.SUPPORT@OUTLOOK.COM"]MISFIT.SUPPORT@OUTLOOK.COM[/EMAIL]
2. Additional small fixes

GET FROM [URL]http://mersenneforum.org/misfit/[/URL]


Due to the lack of activity in the MISFIT forums I will no longer be checking them.

If something comes up please contact me via email at [EMAIL="MISFIT.SUPPORT@OUTLOOK.COM"]MISFIT.SUPPORT@OUTLOOK.COM[/EMAIL][/QUOTE]

Hey Scott,

I think the lack of activity bespeaks just how steady and reliable MISFIT is now. Again, many thanks for your excellent work!

[QUOTE=kladner;365295]I think the lack of activity bespeaks just how steady and reliable MISFIT is now. Again, many thanks for your excellent work![/QUOTE]

+1!!! :smile:

+1

+2

I wish all programs were this reliable and trouble-free.

Lack of activity because it works well :smile:

Thanks for the affirmations on its reliability.

VERSION 2.5.9 released
 

VERSION 2.5.9
1. Improved GIMPS I/O Manager (GIOM) results uploading : In the event of a multi-file upload the files are now uploaded in chronological order of creation. This would only happen if greater than one file exports were processed while connectivity to GIMPS server was compromised. NOTE this feature is implemented by changing the combined results filename, found in GIOM_STAGED, from a random GUID to FileTimeUTC + 5 random chars. In other words the filename is a string representation of the time it was created with a random component to prevent collisions. When GIOM loads it sorts the file list and they are now uploaded in chronological order. This change may resolve the occasional "RESULTS NOT NEEDED" response from GIMPS if out of order exponent reporting is in-fact a problematic condition. (others can debate if that is fact or fiction, but this is logical and came from user suggestion.)


Get from [url]http://mersenneforum.org/misfit/[/url]

Thanks! :smile:

VERSION 2.5.10 released
 

VERSION 2.5.10
1.Added startup detection of missing MISFIT directories


GET FROM [URL]http://mersenneforum.org/misfit/[/URL]

Scott,

I'm sending a separate email... updates to [url]www.mersenne.org[/url] have broken the results upload.

[URL="http://www.mersenneforum.org/showthread.php?p=366245#post366245"]http://www.mersenneforum.org/showthread.php?p=366245#post366245[/URL]

Help ;-)