Go to try it... Thanks for the info
From the task manager, if the app is launched with -ui the CPU remain at 25% busy! while without -ui is near 0%....
Salut,
Josep M
How pass parameters to exe
Moderators: FourthWorld, heatherlaine, Klaus, kevinmiller, robinmiller
-
- VIP Livecode Opensource Backer
- Posts: 9842
- Joined: Sat Apr 08, 2006 7:05 am
- Location: Los Angeles
- Contact:
Re: How pass parameters to exe
If that's idle time you might want to double-check for any timers or other periodic tasks. While the IDE and GUI support in general takes much more RAM than running facelessly, I don't believe the IDE itself uses any timers that should be eating CPU time like that. Here (though on Ubuntu rather than Windows) LC uses 0% CPU time on idle, as it did on Mac last time I checked.JosepM wrote:From the task manager, if the app is launched with -ui the CPU remain at 25% busy! while without -ui is near 0%....
Richard Gaskin
LiveCode development, training, and consulting services: Fourth World Systems
LiveCode Group on Facebook
LiveCode Group on LinkedIn
LiveCode development, training, and consulting services: Fourth World Systems
LiveCode Group on Facebook
LiveCode Group on LinkedIn
Re: How pass parameters to exe
Hi,
But I think to remember that I had something similar with my socketserver too. Runs as a faceless service on a Win2K3 server, and when I started it with "-ui" it would constantly hog the CPU. I may try this again at the weekend when I can play with it ...
RAM usage/ time until gone: *1)
I'd rather have it run minimized ;-)
Have fun!
*1): Update:
This SA can be started with the shiftkey down, then it populates its fields (for setting prefs) and stays open. Else it doesn't touch any fields, only does its work and quits. Now I changed it so that it always populates its fields. (= forced GUI operation)
The RAM/ time values didn't change, only the time needed to quit with "-ui" is even longer, 30 sec.
I know for sure that I had StandAlones on Windows that wouldn't quit properly when called using "-ui", and I'm sure their job was done (they write files), and nothing more to do but to quit. Tried some of these, but couldn't reproduce, for sure.FourthWorld wrote:If that's idle time you might want to double-check for any timers or other periodic tasks.JosepM wrote:From the task manager, if the app is launched with -ui the CPU remain at 25% busy! while without -ui is near 0%....
But I think to remember that I had something similar with my socketserver too. Runs as a faceless service on a Win2K3 server, and when I started it with "-ui" it would constantly hog the CPU. I may try this again at the weekend when I can play with it ...
Interesting, tried with a little SA (5MB on disk with all files) that reads & decrypts its prefs, fetches a file from FTP, then renames some files, finally writes the FTP-file to disk and quits.FourthWorld wrote:While the IDE and GUI support in general takes much more RAM than running facelessly, [...]
RAM usage/ time until gone: *1)
- With "-ui": ~6MB RAM/ ~20 secs,
- w/o "-ui": ~7MB RAM/ ~6 secs.
I'd rather have it run minimized ;-)
Maybe a problem of the Windows implementation?FourthWorld wrote:Here (though on Ubuntu rather than Windows) LC uses 0% CPU time on idle, as it did on Mac last time I checked
Have fun!
*1): Update:
This SA can be started with the shiftkey down, then it populates its fields (for setting prefs) and stays open. Else it doesn't touch any fields, only does its work and quits. Now I changed it so that it always populates its fields. (= forced GUI operation)
The RAM/ time values didn't change, only the time needed to quit with "-ui" is even longer, 30 sec.
All code published by me here was created with Community Editions of LC (thus is GPLv3).
If you use it in closed source projects, or for the Apple AppStore, or with XCode
you'll violate some license terms - read your relevant EULAs & Licenses!
If you use it in closed source projects, or for the Apple AppStore, or with XCode
you'll violate some license terms - read your relevant EULAs & Licenses!
Re: How pass parameters to exe
Hi,
In my case, the app execute two update over a mysql, nothing more. But still on memory...
In my case, the app execute two update over a mysql, nothing more. But still on memory...
-
- VIP Livecode Opensource Backer
- Posts: 9842
- Joined: Sat Apr 08, 2006 7:05 am
- Location: Los Angeles
- Contact:
Re: How pass parameters to exe
This sounds like a bug. Whether it's in Windows, LiveCode, or your code is hard to say, but I'd start in inverse order of the number of people testing the code.AxWald wrote:I know for sure that I had StandAlones on Windows that wouldn't quit properly when called using "-ui", and I'm sure their job was done (they write files), and nothing more to do but to quit. Tried some of these, but couldn't reproduce, for sure.
But I think to remember that I had something similar with my socketserver too. Runs as a faceless service on a Win2K3 server, and when I started it with "-ui" it would constantly hog the CPU. I may try this again at the weekend when I can play with it ...
...
RAM usage/ time until gone: *1)On a Win 7-64, LC community 6.7.1
- With "-ui": ~6MB RAM/ ~20 secs,
- w/o "-ui": ~7MB RAM/ ~6 secs.
I'd rather have it run minimized
...
*1): Update:
This SA can be started with the shiftkey down, then it populates its fields (for setting prefs) and stays open. Else it doesn't touch any fields, only does its work and quits. Now I changed it so that it always populates its fields. (= forced GUI operation)
The RAM/ time values didn't change, only the time needed to quit with "-ui" is even longer, 30 sec.
My experience with faceless apps on Windows is limited, but I have used them on that OS now and then and frequently use -ui on Mac and Linux. My optimistic view of the init time for faceless execution is not only borne of that experience, but reinforced logically: graphics and UI messaging eat up a lot of time and space, so when neither is present we should expect things to run much leaner.
Can you post the code your app is using? If it's taking too long and not quitting normally those two symptoms may well be related.
If your code checks out at least we'll be one step closer to arriving at a recipe for what might be a recently-introduced bug in the Windows engine for LiveCode. And if that checks out if may be a bug in Windows. But let's start in inverse order of the number of eyeballs that have worked on the code and move up from there.
Richard Gaskin
LiveCode development, training, and consulting services: Fourth World Systems
LiveCode Group on Facebook
LiveCode Group on LinkedIn
LiveCode development, training, and consulting services: Fourth World Systems
LiveCode Group on Facebook
LiveCode Group on LinkedIn
Re: How pass parameters to exe
Hi Richard,
did quite some testing now, but as to be expected, no real reproducible results yet. Cought a suspect yet, but it does use "ask" and "answer" - in "-ui" mode it fails to quit (even if ask/ answer aren't called), in minimized mode it runs quite well.
I'll keep on testing, and as soon as I find code that proves my claim, I'll post it.
Sorry for no better results yet, this bug (if it is one) seems to be a real bitch (not related to any sex or gender, used as plain simple derogative).
Testing my socket service with "-ui" is impossible atm, I'd get stoneth as soon as I tried. It's used all around the clock, and even with 20 concurrent sessions it rarely uses more then 10% of a 2x Xeon 3GHz ancient server. Trying to overload it with 200+ sessions I managed to get the CPU usage to 50% - after this the IN connection was the limit.
I'll report as soon as I find out more.
Have fun!
did quite some testing now, but as to be expected, no real reproducible results yet. Cought a suspect yet, but it does use "ask" and "answer" - in "-ui" mode it fails to quit (even if ask/ answer aren't called), in minimized mode it runs quite well.
I'll keep on testing, and as soon as I find code that proves my claim, I'll post it.
Sorry for no better results yet, this bug (if it is one) seems to be a real bitch (not related to any sex or gender, used as plain simple derogative).
Testing my socket service with "-ui" is impossible atm, I'd get stoneth as soon as I tried. It's used all around the clock, and even with 20 concurrent sessions it rarely uses more then 10% of a 2x Xeon 3GHz ancient server. Trying to overload it with 200+ sessions I managed to get the CPU usage to 50% - after this the IN connection was the limit.
I'll report as soon as I find out more.
Have fun!
All code published by me here was created with Community Editions of LC (thus is GPLv3).
If you use it in closed source projects, or for the Apple AppStore, or with XCode
you'll violate some license terms - read your relevant EULAs & Licenses!
If you use it in closed source projects, or for the Apple AppStore, or with XCode
you'll violate some license terms - read your relevant EULAs & Licenses!