can someone PLS find a way to loop this program>!

[mccoyn]

The new version has a GUI. You just need to double click the file and it should launch. It'll have a place to enter the location for super_pi_mod.exe.

I'm not sure what the 0xc0000135 error is exactly. If you don't have the DotNet framework installed you will have to install that first (I probally should have mentioned that.) http://www.microsoft.com/downloads/...E3-F589-4842-8157-034D1E7CF3A3&displaylang=en

 

[DaWiper]

Installing .NET framework did the trick.
It seems to work just fine. Good job so far.
Heres a couple of things:
-Make pi_loop remeber the path to super_pi_mod
-If the user presses stop within superpi or closes application pi_loop will hang.
-Maybe make a timer wihtin pi_loop that records how long each run lasted. And when all runs are done put this in a txt file. This will be a faster way for the user to evaluate the results.

I'm still testing, will come back with more if I find anything. I'm having problems with getting my testrig back to its super unstable state.

 

[mccoyn]

Incidently, those are all things I was planning on fixing.

Currently there is a command line syntax. "pi-loop 1 11 c:\super_pi_mod.exe" will run 1 loop of option 11 (32M) using c:\super_pi_mod.exe. Using this command line will automatically fill in the options and all you have to do is click begin.

The hanging problem is something I will correctly. Its just a remnant from the earlier command line version. I just have to restructure the code a bit.

I didn't put in a timer because it would be as much as a second off. When I fix the hanging problem I will be able to improve this precision problem and I will add a timer.

I hope to add options to save a log, not save the screen shots, auto-run when loading (command line) and auto-exit when done. I also want to have an export options button that will save the selected options to a shortcut file. I might make a seperate file to save the super_pi_mod path.

I used speedfan to switch off a fan and make my computer unstable. This saved the trouble of rebooting. As you can imagine I didn't want to compile code on an unstable computer, but I couldn't test it on a stable computer.

Thanks for the tips.

 

[deathstar13]

i have it working is decent for a beta type proggy.
as dewiper said the 3 items would be nice if fixed.

i did notice one thing in a few simple runs without testing far is if i use the machine and dont have the program and pi on top it wont loop.
not a really bad thing as i dont expect to use the machine and would be annatended anyhow.

ill try some error testing now.
so far i really like it

"edit"
it will error out exactly like the original program also
this is exactly what i was talking about lol excellent job!

you say you will have it log results from each run? now will that record the errors also? as from what i seen it will restart pi if it finds an error wich is good also but an error count would be nice and mabey where it erred or how far in the run.

beside getting the stability issue with closeing and stopping fixed this is super sweet even as it is.

 

[mccoyn]

The log will be exactly what appears in the textbox at the bottom of the pi_loop window. I will add a timer output to this log as well.

The program as its implemented will never allow users to do other stuff at the same time. I can improve things slightly, but I still need to take control of the mouse from time to time.

 

[deathstar13]

The log will be exactly what appears in the textbox at the bottom of the pi_loop window. I will add a timer output to this log as well.

The program as its implemented will never allow users to do other stuff at the same time. I can improve things slightly, but I still need to take control of the mouse from time to time.

i dont think its such a big issue as if they were at thier pc they could manually restart the 32m run.
so i really dont find it much an issue myself.

i dont think this will need to be perfect either as its basically a tool very specific people only will be useing.
but i think you have an excellent grasp on where to take this and how to impliment any improvemnts.
might wanna add some credits for you or name the app after you or something also.

when you get it exactly how you want it id suggest releasing it in the amd section of cpu's as thats mainly where the idea was spawned and much feedback could be offered there also.even tho its softare its main purpose is getting the last mhz out of an oc while maintaining a decent stability.

 

[mccoyn]

Yeah, when I get all the features I want in I was thinking of posting it in the general CPU discussion forum.

I could add an about box with a credit for me and a link to the sourceforge page.

 

[mccoyn]

I uploaded a new version of pi-loop today. http://sourceforge.net/project/showfiles.php?group_id=144006

I think I got most of the major concerns covered. Heres my release notes.

Several additions here.
- Pi Loop no longer hangs while waiting for super_pi_mod to finish. Instead, it switches to a new panel that gives an estimate of when Pi Loop will be finished.
- An about button and dialog was added.
- The log is now saved to a file.
- Elapsed time for each iteration is now part of the log (this is an estimate.)
- The path for super_pi_mod.exe is now saved to the registry.
- The super_pi_mod window is automatically moved so that it does not overlap the Pi Loop window.
- Whenever Pi Loop moves the mouse it will move it back to its last position when it is done. This makes it a little nicer when you accidently try to run infinite 16K loops.

I'm aware of a couple bugs right now. It seems that if you run a set of calculations and then after its done you try to start again it will hang. The other bug about putting other windows on top of super_pi_mod still exists.

One other concern I have is that pi-loop polls to find out if super_pi_mod is done. This uses processor time and can slow super_pi_mod down. I would like to instead use a hook procedure, which should avoid unneeded thread switching. It will also improve the accuracy of pi-loop timers.

 

[macklin01]

Wow, this is great work! Keep it up!! -- Paul

 

[deathstar13]

taking it around the block now

 

[maxxoverclocker]

wow man i have no idea how it works... but it works great lol

 

[deathstar13]

i dont know why but 1m wont loop.
16k and 32k did tho.
i didnt try 32m yet due to lenght of the runs.

 

[mccoyn]

I get the same thing on 1m. That seems really strange. I just assumed that each option should work since the code is so similar so I never tested that. I'll look into it next week some time.

 

[mccoyn]

I've made another minor release. http://sourceforge.net/project/showfiles.php?group_id=144006

Fixed a couple bugs.
- Pi Loop no longer stops when it runs long loops.
- Pi Loop no longer stops when it is rurun without restarting Pi Loop.

 

[mccoyn]

I've made yet another release. The changes in this one are mostly internal. You might notice it is ever so slightly faster. That took a lot of work, so I expect it introduced some bugs.

From now on, I'm not going to use this thread and forum to anounce releases. Instead I've created a public mailing list. If the traffic on this list gets too high I'll create a new list that only I can send to for just releases. You can subscribe to the list here:
http://lists.sourceforge.net/lists/listinfo/pi-loop-users

 

[El<(')>Maxi]

Excellent work mccoyn, now to dev one from scratch with a snazzy new interface

 

[mccoyn]

I'm hoping to finish pi-loop this week and start the next thing next week.