released: MISFIT for MFAKTx version 2.11.0 (31-Jan-2017) - Page 2

Xyzzy · 2017-02-21, 15:52

Are you worried the downloads might be compromised?

swl551 · 2017-02-22, 16:03

I have an account on the ftp server, but I no longer want to trust using it as my code may get compromised and make me look bad.

Xyzzy · 2017-02-22, 16:42

Wouldn't posting checksums work towards preventing that?

kracker · 2020-10-11, 23:43

download link(s) to MS drive are dead.

swl551 · 2020-10-21, 00:09

Quote:

Originally Posted by kracker

download link(s) to MS drive are dead.

Fixed.

LaurV · 2020-10-23, 06:59

Wow! Scott! Welcome back man!

Now that you put your nose back into Misfit, can you beat its butt until it will accept "PFactor" lines? That way, we could use it with gpuOwl as well as MfaktX, because everything it does, is to take assignment lines from a common pool and distribute them to the monitored folders' worktodo files, as well as taking back the results from there and send them to PrimeNet.

That's all we need.

So, there is no difference in behavior, the only issue is that it only grabs "Factor" lines from the pool, and ignores the others, spitting errors at us

As well as it will accept P-1 lines too (starting with "PFactor") we are good to go. We can manually add assignments to the pool, if you can't make Chris come to reason

I may go to GPU72 once per week and grab 500 lines by hand, I am ok with that. But managing the local splits and specially balancing the "worktodo"s is bloody (and of course, one card is always faster than the other).

As you gave me the sources some time ago, I already did my own "patch" to it (as I remember, there were some changes in global.cs, and another file), but it is better if you do the changes and release by yourself. Having the "accepted lines" customizable (like a text or config file where I can write "Factor=", "Factor=N/A" "PFactor=" blah blah) would be even better...

This request may seem trivial, "who the hack is doing so much P-1??" but if you look to the evolution of the things since you left, one guy with two or three Radeon VII cards can run 4 to 7 instances of gpuOwl and each of them can cycle 10 to 30 P-1 assignments per day. Not long ago I finished about 800 P-1 exponents from GPU72, and splitting/balancing those was not easy. Maintaining those manually is slowly becoming a pain in the butt.. (well, there are scripts, pythons, and all kind of other scary animals there outside, but we like Misfit

)

swl551 · 2020-10-27, 23:02

Hi.. I'll have to fire up my dormant windows 10 pc and install visual studio.... it has been so long I probably won't even recognize my own code.

Give me a few days...

swl551 · 2020-11-03, 00:11

Quote:

Originally Posted by swl551

Hi.. I'll have to fire up my dormant windows 10 pc and install visual studio.... it has been so long I probably won't even recognize my own code.

Give me a few days...

Ok. I have Visual Studio up and running...

Now I need a sample of these factor lines that MISFIT is OK with and the P1 lines it is unhappy with. You see I can't remember the distinction so I cannot code for it from the top of my head....

LaurV · 2020-11-18, 14:29

TF line properly handled:

Factor=2CCClahBlahBlah164D5,211836929,73,74

P-1 line ignored:

PFactor=249DlahBlahBlah8E5D52F,1,2,101190059,-1,76,2

(they won't be mixed in the same files, as the tools used to work them are different, but this doesn't really matter, you just handle the second format same as the first, and we are good!)

Thanks in advance!

kotenok2000 · 2022-04-21, 11:51

Misfit says "unsupported feature" if V5UserID and ComputerID are set in mfactc.ini
and uid is present in results.txt

2020-10-23, 06:59	#17
LaurV Romulan Interpreter "name field" Jun 2011 Thailand 10281₁₀ Posts	Wow! Scott! Welcome back man! Now that you put your nose back into Misfit, can you beat its butt until it will accept "PFactor" lines? That way, we could use it with gpuOwl as well as MfaktX, because everything it does, is to take assignment lines from a common pool and distribute them to the monitored folders' worktodo files, as well as taking back the results from there and send them to PrimeNet. That's all we need. So, there is no difference in behavior, the only issue is that it only grabs "Factor" lines from the pool, and ignores the others, spitting errors at us As well as it will accept P-1 lines too (starting with "PFactor") we are good to go. We can manually add assignments to the pool, if you can't make Chris come to reason I may go to GPU72 once per week and grab 500 lines by hand, I am ok with that. But managing the local splits and specially balancing the "worktodo"s is bloody (and of course, one card is always faster than the other). As you gave me the sources some time ago, I already did my own "patch" to it (as I remember, there were some changes in global.cs, and another file), but it is better if you do the changes and release by yourself. Having the "accepted lines" customizable (like a text or config file where I can write "Factor=", "Factor=N/A" "PFactor=" blah blah) would be even better... This request may seem trivial, "who the hack is doing so much P-1??" but if you look to the evolution of the things since you left, one guy with two or three Radeon VII cards can run 4 to 7 instances of gpuOwl and each of them can cycle 10 to 30 P-1 assignments per day. Not long ago I finished about 800 P-1 exponents from GPU72, and splitting/balancing those was not easy. Maintaining those manually is slowly becoming a pain in the butt.. (well, there are scripts, pythons, and all kind of other scary animals there outside, but we like Misfit ) Last fiddled with by LaurV on 2020-10-23 at 07:14

2020-11-18, 14:29	#20
LaurV Romulan Interpreter "name field" Jun 2011 Thailand 3·23·149 Posts	TF line properly handled: Factor=2CCClahBlahBlah164D5,211836929,73,74 P-1 line ignored: PFactor=249DlahBlahBlah8E5D52F,1,2,101190059,-1,76,2 (they won't be mixed in the same files, as the tools used to work them are different, but this doesn't really matter, you just handle the second format same as the first, and we are good!) Thanks in advance! Last fiddled with by LaurV on 2020-11-18 at 14:31

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Mlucas version 17.1	ewmayer	Mlucas	96	2019-10-16 12:55
LLR Version 3.8.19 released	Jean Penné	Software	11	2017-02-23 08:52
LLR Version 3.8.16 released	Jean Penné	Software	38	2015-12-10 07:31
LLR Version 3.8.15 released	Jean Penné	Software	28	2015-08-04 04:51
MISFIT for MFAKTx	swl551	MISFIT	186	2014-02-05 23:27

2017-02-21, 15:52	#12
Xyzzy Aug 2002 7×1,223 Posts	Are you worried the downloads might be compromised?

2017-02-22, 16:03	#13
swl551 Aug 2012 New Hampshire 2³·101 Posts	I have an account on the ftp server, but I no longer want to trust using it as my code may get compromised and make me look bad.

2017-02-22, 16:42	#14
Xyzzy Aug 2002 20561₈ Posts	Wouldn't posting checksums work towards preventing that?

2020-10-11, 23:43	#15
kracker "Mr. Meeseeks" Jan 2012 California, USA 2⁷·17 Posts	download link(s) to MS drive are dead.

2020-10-27, 23:02	#18
swl551 Aug 2012 New Hampshire 2³×101 Posts	Hi.. I'll have to fire up my dormant windows 10 pc and install visual studio.... it has been so long I probably won't even recognize my own code. Give me a few days...

2022-04-21, 11:51	#21
kotenok2000 Mar 2018 2·5² Posts	Misfit says "unsupported feature" if V5UserID and ComputerID are set in mfactc.ini and uid is present in results.txt