Announcement

**Lance Edmonds** · 26 Apr 2000, 01:39 AM

"Non standard" screen column/row combinations are not supported by PowerBASIC/DOS. If you need to use such combinations, you'll need to use the text routines built into the BIOS or poke your data into screen memory directly, instead of the using the built-in PRINT statement and pbvScrnCols, etc.

As far as your hanging problem goes, you'll have to provide a *lot* more information so we can duplicate the problem - for example, what have you done to prove that that this is a PowerBASIC 'bug' rather than something (anything?) else?

Some questions you may wish to consider are:
Have you tried this on more than one PC?
Does it occur with any particular variable class or does it have a pattern to the way it fails?
Does it also happen when you evaluate plain numeric expressions?
Does the effect vary with different source code?
Does it occur if you use a different operating system?
Have you tried alternative memory manager software, or none at all?
Have you found anything that can help in reproducing the problem or identify the cause?

Please consider these issues carefully and let us know what you find.

------------------
Lance
PowerBASIC Support
mailto:[email protected][email protected]</A>

**Tom Hanlin** · 26 Apr 2000, 08:30 AM

You might want to check with Information Management Systems (www.infoms.com). I suspect their PBWiz library can handle nonstandard screen modes. (I wrote the original PBWiz, but it's been a few years. No, I have no financial interest in it, these days.)

------------------
Tom Hanlin
PowerBASIC Staff

**Alan Earnshaw** · 26 Apr 2000, 03:34 PM

Tom, thanks for the plug.

Sebastian,

As Tom mentioned, the PB Wizard's Library can handle non-standard screen sizes but only in graphics mode. (The Library lets you set the screen size to any VESA mode supported by your video card). You can read about the product at http://www.infoms.com/pbwiz.htm .

Alan

------------------
Alan C. Earnshaw
Information Management Systems, Inc.

**Sebastian Groeneveld** · 27 Apr 2000, 01:20 AM

Thanks for the PBWiz link, I hope I can find a PB35 version of the library. However I don't think it's going to help me in this case. The screen already is in 90 column text mode when my program starts, I just have to print coloured text - luckily I have added a STDOUT option

But then about the pbvScrnCols variable... don't you think it's weird that even IF the screen has a non-standard amount of columns, it shows the wrong value, while the BIOS contains the correct value...? I'm still curious where its value is coming from...

Now let me answer all your questions, Lance...

Q. Have you tried this on more than one PC?
A. It's been tested on a 386 DOS system, a 486 DOS system and on a Pentium running in MS-DOS mode (no DOS box) - they all crashed.

Q. Does it occur with any particular variable class or does it have a pattern to the way it fails?
A. No, when I bring up the evaluation box, type anything I like (string var, int var, numeric exp or just random text) and press Enter the system hangs and keyboard does not respond.

Q. Does it also happen when you evaluate plain numeric expressions?
A. Yes, see previous.

Q. Does the effect vary with different source code?
A. No, even after just having started PB with blank editor.

Q. Does it occur if you use a different operating system?
A. No, tested in Windows 3.11, Win95, DOS. In Windows EMS is automatically created, and there's no problem then.

Q. Have you tried alternative memory manager software, or none at all?
A. I only know the problem is solved when using either EMM386.EXE or QEMM

Q. Have you found anything that can help in reproducing the problem or identify the cause?
A. Try removing any EMS driver in your config.sys and boot the computer in MS-DOS mode. Then start PB and try to evaluate 1+1.

Maybe this sounds weird... but I hope your system crashes too,
otherwise I don't know what's wrong with me

Greetings and thanks for help

------------------
Sebastian Groeneveld
mailto:[email protected][email protected]</A>

**Lance Edmonds** · 27 Apr 2000, 05:37 AM

I'll pass your comments about pbvScrnCols along to R&D.

In Windows 2000, there is no "Boot to DOS" mode, so I can't immediately test this where I am here, however, at the office I have a whole bunch of (non-EMS) DOS machines, and today I tried and was not able to duplicate your problem.

Can anyone else duplicate the problem?

------------------
Lance
PowerBASIC Support
mailto:[email protected][email protected]</A>

**Steve Bouffe** · 27 Apr 2000, 02:03 PM

I have discovered another bug in the editor.

On occasion when scroling a line will vanish and leave a blank line in its place, if you cur up / down the line will not show up. If you page up then down, its back !

------------------

**Lance Edmonds** · 27 Apr 2000, 08:58 PM

While we appreciate your feedback, anecdotal (sp?) evidence is unfortunately not going to help us fix the problem - we need to be able to duplicate the problem in order to track the down the cause. We also need more information on your O/S version, etc.

If you can give us exact directions (maybe even with example code as applicable) so that we can duplicate the problem(s), we can get the PowerBASIC R&D department on the job to fix the prpblem.

Personally, I've never come across the "blank line" problem you describe, and I've been using the PB IDE on a daily basis for several years.

As you may be aware, the IDE itself uses XMS memory, so it is conceivable that you could be seeing a problem may be with your XMS memory manager software - there have been several buggy EMM386 and HIMEM releases from Microsoft (not to mention the notorious QEMM), and when they fixed problems they did not always change the file date, time and version number - this makes it very difficult to identify problematic versions of these memory managers.

The easiest way (especially for Win 3.x users) is to use the versions located in the C:\DOS directory instead of the ones installed in the C:\WINDOWS directory (or vice versa) - quite a few people have reported success with this simple switch-over (often the memory managers supplied with MSDOS were more recent than the Win 3.x versions).

Another common cause of problems are TSR's and drivers that may be installed - clean booting will usually help identify these as a source of problems.

------------------
Lance
PowerBASIC Support
mailto:[email protected][email protected]</A>

Announcement

pb3.5 bugs?

pb3.5 bugs?

Comment

Comment

Comment

Comment

Comment

Comment

Comment