Need Testers for WinAGI

[AGKorson]

I finally have a version ready. But I don't want to do a general release until I get a chance to have it tested by a few other people to work out any last minute fixes.

I would like to get four people who are willing to do an indepth review/test of the program. As a thank you, I am offering $10 (via paypal) to the first four people who agree to help. (I'd love to offer more, but I'm not that rich!  ) As a bonus, I'll offer an extra $10 to whichever tester provides the best and most helpful feedback.

If you are interested, send a PM, and I'll provide a link to download the installer.

My goal is to release a public version before end of the year, so if you do want to help, I'd appreciate if you could get comments back to me as soon as possible.

Thanks in advance!

[Collector]

I would love to without the monetary incentive, but don't have enough time to give it right now.

[klownstein]

I am actively (weekly, if not daily) using your previous version of WinAGI (ver 1.2.1) and would be happy to use your newest version, although I don't know that I would be able to put in the effort to actively look for bugs to fix.  Mostly, I have been using the picture and view editors lately.  If that would be helpful, let me know.  I also would not need or expect to be compensated.

P.S.  The most common error I run into currently with ver 1.2.1 happens when editing views.  If I copy and paste a portion of the view, move the highlighted section so part of it is off the right hand side of the screen, and use the flip horizontal feature, I get an error that says:

Run time error: '9':
Subscript out of range

The program then crashes and I lose any unsaved data (usually the view I've been working on up until it was last saved).  If I do the same thing but flip vertical, I get no error and it works like I would expect.

[AGKorson]

I am actively (weekly, if not daily) using your previous version of WinAGI (ver 1.2.1) and would be happy to use your newest version, although I don't know that I would be able to put in the effort to actively look for bugs to fix.  Mostly, I have been using the picture and view editors lately.  If that would be helpful, let me know.  I also would not need or expect to be compensated.

P.S.  The most common error I run into currently with ver 1.2.1 happens when editing views.  If I copy and paste a portion of the view, move the highlighted section so part of it is off the right hand side of the screen, and use the flip horizontal feature, I get an error that says:

Run time error: '9':
Subscript out of range

The program then crashes and I lose any unsaved data (usually the view I've been working on up until it was last saved).  If I do the same thing but flip vertical, I get no error and it works like I would expect.

Good catch! I hadn't considered someone flipping a selection that overlaps an edge. I also forgot to add an error handler to the horizontal flip function; the vertical function will ignore the error. But both of them won't work correctly unless the selection is entirely within the cell drawing area.

I will add this to the bug list and get it fixed.

And if you would like to try the new version, even if you don't have a lot of time, I'd be happy to share. Shoot me a PM and I'll send you the link.

thx!

[Kawa]

Off to a good start in 1.2.3! Turns out because the Preview window is too damn small by default, clicking a Picture resource causes RTE 380, "Invalid Property Value". This also leaves behind windowless processes. But now that I found out about this I can safely test the rest.

Looking at KQ3 logic 0, the first line is already considered an error:

Code: [Select]

if (error_code > 0) {
  call(98);
}"Argument syntax error in 'if' statement". I didn't write that statement, WinAGI did.

[AGKorson]

I will make sure the preview window gets a reasonable default size, and check for that error.

Can you check your Settings, on Logic tab- is the "Use Reserved Variables/Flags Names" option checked or unchecked? It should be checked by default, but if it's turned off, and you haven't defined 'error_code' to be a variable, this error gets displayed. If it is checked, and you are still getting this error, let me know. (I think I will add something to the help file for this error to point to that solution; it's the most likely scenario when looking at a new game for the first time in WinAGI)

[Kawa]

That was it exactly. Now I just get a bunch of warnings about various variables and flags being "normally read-only", but it does compile.

[AGKorson]

You can get rid of those warnings by changing error detection to Low. The compiler won't complain unless it is unable to generate valid code; only a few 'critical' warnings ever get displayed.  Error detection level is set on the Logics Tab in Settings.

But I prefer using the Ignore Warning feature; you can individually ignore a particular warning that you don't want to see again. In the Warnings window, you can right-click a warning, and choose 'Ignore Warning', and the compiler will never show you that warning again. Use the 'Reset Message Box Warnings' option if you want to re-enable previously ignored warnings.  (Oh, and if you double-click an error or warning, WinAGI will automatically bring up the affected logic and highlight the line where the warning/error occurred.)

[pmkelly]

I'd be very happy to take a look and send you some detailed feedback!

I've never used WinAGI before and have been away for the scene for 20 years so would be keen to check it out.

No payment necessary

[AGKorson]

Hi Peter,

I'd love to have you test WinAGI. If it hadn't been for AGI Studio, I doubt I ever would have written WinAGI. so THANK YOU for that!

I'll send you a link to the install file by PM.

[AGKorson]

I've fixed the first couple things that have been pointed out so far. And I'm continuing to do my own testing (and found a few more minor issues). Things are progressing well. I hope to get more feedback from testers soon.

In the meantime, after thinking more about the snippet of original source code from Leisure Suit Larry (in another post here), I am thinking about tweaking the syntax used in WinAGI.

I am strongly considering dropping support for block comments (i.e. '/* … */'); I get the impression that block comments were not an option for original AGI source - it appears the square bracket '[' was the only comment token in use. AGI Studio included the bracket token, but also added the double slash token ('//') as well as block comments. If these were not in original AGI source, I'd like to take them out of WinAGI.

But I also have another reason to eliminate block comments - it will significantly improve performance of the rich text editor. There are no existing rich text editor controls available for VB (either as native OCX files, or other object that I can use API calls on) that have all the features found in modern code editors (most importantly syntax highlighting); I had to write a custom control. It works pretty well, but with large logics, the potential existence of block comments really kills performance (because the control has to re-highlight everything on each edit in case a block command was added or deleted - perhaps a smarter programmer could figure out how to make it work more efficiently, but I've run into a wall with this).

If I were to drop support for block comments performance would not suffer at all (because I only have to reformat lines/text as it is entered, and only the lines that are affected). How big of a deal would that be? In the few fan game source files I've seen, I have not seen much use of block comments.

As a compromise, I am looking to add a 'block comment' feature that will add comment tags at start of all lines in a selection, as well as an 'block uncomment' feature to remove them (this seems to be a common feature in the few source editors I've worked with).

Are there any major objections to dropping support for block comments?

[Kawa]

I agree with this entirely, and I'm reminded of SCI Companion's block comment feature.

[AGKorson]

I've gotten a few comments back, but I was hoping to get more by now. If anyone who's testing has any feedback, I'd love to hear it.

In the meantime, I have a question for you all - it's been suggested that a synth instrument such as 'Lead (square)' (instrument #80) would be a better default MIDI instrument for sounds than 'Grand Acoustic Piano' (instrument #0). Does anyone have any thoughts on that?

[Kawa]

It'd better match the PC speaker, so...

[Collector]

Agreed.