Script runs in SciTE but not when compiled

[Lord Maim]

r u sure beta is toggled when using the right click compile option?

To be sure, I used the Tools > Beta Compile option in SciTE and the same issue occurs.

As for the msgbox idea, to save myself some time I put a msgbox command in line 1, recompiled, and ran it. No box.

This is about where I start beating my head against a wall.

EDIT: I just checked the right click compile options, and it is the beta aut2exe that is being used. Version listed is 3.2.1.13

Edited December 20, 2006 by Lord Maim

[someone]

I could be very off and you do have a lot of people that will forever know more(a lot more) about autoit then I do, but....

You said you compiled your script with a Msgbox in line 1 and it didnt run? Can you compile any script and have it run? Someone correct me if I'm wrong but doesn't the fact that a msgbox in the first line won't run when compiled mean something?

Maybe it doesn't and like I said theres a lot smarter people in here but it still seems weird. Did you put the msgbox above any #include lines?

I say try to compile anything and see what happens. Can you break up your code as a whole starting at the begining so maybe you can just compile several hundred lines at once or something similar?

Good luck

[Lord Maim]

I say try to compile anything and see what happens. Can you break up your code as a whole starting at the begining so maybe you can just compile several hundred lines at once or something similar?

Good luck

Yeah someone, in my last(?) post, I mentioned that I tried compiling some older versions of my script, and yes they did run. Also, the line 1 msgbox still runs in SciTE, so no help there. Thanks, though.

Lar: ........ Yes I did add some regions in a recent version to manage the 3300 lines of code. I don't know if my last compile had them or not, but its worth investigating. I'll yank them out and report back.

EDIT: No. Yanked them all out, and no effect. Dammit.

Edited December 20, 2006 by Lord Maim

[klsgfx]

Lord Maim,

first one hint for next time: Use a versioning system (like CVS or SVN or whatever). This way you can trace all the changes you did easilyb back to different versions of your script.

Ok, but now for my tries for helping:

1) Backup the actual script and then put a Msgbox on top of the script and remove (or comment) all other lines. Compile, what does happen?

2) You still have an older runnnig version of the script in source? Get Winmerge (google for this) and compare the old script to the new one and perhaps check the different lines for possible errors. Or perhaps post the differences here.

That are my ideas, now back to the professionals.

Take care

Klaus

[Lord Maim]

Did you happen to save the script as unicode or something? Or copy/paste something with crazy characters in it?

Lar.

No, and I'm pretty sure the only special characters are expressed with a CHR command, but I'll give WinMerge a try and see if I can come up with some answers.

Thanks much for that tidbit, Klaus. I think that's really going to be a big help!

[1905russell]

This is only valid when you use Encodit...

I woke up this morning hoping to find that you had resolved your problem but alas not yet.

I know my similiar problem was with encodit and its not the same thing but I was thinking -

Why don't you try encodit and see if that either throws it out or lets you run it in scite.

If it does run in scite maybe then the compiled encodit version will run?

Encodit seems to pick up all kinds of otherwise undetected problems so who knows?

[Lord Maim]

I woke up this morning hoping to find that you had resolved your problem but alas not yet.

I know my similiar problem was with encodit and its not the same thing but I was thinking -

Why don't you try encodit and see if that either throws it out or lets you run it in scite.

If it does run in scite maybe then the compiled encodit version will run?

Encodit seems to pick up all kinds of otherwise undetected problems so who knows?

Actually, using WinMerge last night, I went through the code change by change. I copied sections from the 20th version (works in SciTE, but not working compiled) into the 18th version I still had the base code for, piece by piece and compiled them (right click compile) to see if they'd work. Fifteen sections later, I had copied over everything relevant, but it still worked. I am now not only frustrated, but completely confused. The only difference between the 20th version and the updated 18th, is the absence of regions and several sections of commented out lines.

Jdeb, it looks like I won't need a decompile of my 1.19 anymore, but if there's someone on staff that you know of that might be able to use the non-working code I have to help track down the problem for AutoIT debugging, I'd be willing to provide it. (Eventually. I'd still have to pull out all the work related stuff, but if it helps out I'll take the time.)

Thanks anyhow for the assistance that everyone gave in trying to diagnose the issue, given the unusual circumstances. I do appreciate the help. Though I'm still interested in finding out what the heck the problem was in the first place.

Unless I've missed something small and random, I don't know what else it could be. I pulled out all the regions yesterday before working with WinMerge, and that compile still didn't work, so I don't think that was the issue. Is it possible that something in the comments could affect a compile? That seems more than a little counterintuitive. Overall program length limit? Would blank lines affect it in any way? Maybe its the lack of sleep talking, but I'm paranoid now that any little change I make might result in the same issue. If nothing else, it reiterates the age old adage "Save early, save often"