[Coco] [Color Computer] OS-9 Editor and NitrOS-9 Download

[Robert Gault]

David wrote:

> On Mon, Dec 08, 2003 at 07:36:33PM -0500, Robert Gault wrote:
> 
>>David wrote:
>>
>>>On Sun, Dec 07, 2003 at 08:22:46AM -0500, Gene Heskett wrote:
>>>
>>>
>>>>On Saturday 06 December 2003 22:52, KnudsenMJ at aol.com wrote:
>>>>
>>>>
>>>>>Scred was an OSK editor that came with the MM/1 and was an official
>>>>>Microware product.  Didn't know it had been back-ported to 6809. 
>>>>>Those who used it on 68K systems said it was very crash-prone. 
> 
> 
>>>>Its just as crash-prone on the coco.  Not of much use really.
>>>
>>>
>>>I found the cause of one crash - actually, I think it was the same thing
>>>in both the coco and OSK editors.  The pointer to one of the lookup
>>>tables for something was defined as a signed integer rather than an
>>>unsigned integer.  ALT-<key> input was, of course, defined as a negative
>>>number, causing a jump to a wrong location.  I posted a patch for this
>>>on Compuserve and/or maybe Delphi, plus some other places.  If anyone
>>>still uses coco scred, I can post the modpatch file to fix it.  I
>>>attempted to do the same thing for OSK, but never was able to track this
>>>one down. I went to ved anyway..
>>>
>>>
>>
>>Yes, please post the patch and better the source code for the patch if 
>>you still have it.
> 
> 
> In response to this and at least two others, here's the patch file.
> 
> -------------- cut --------------------
> * Fixes(?) bug in "scred" where alt-key crashes ???
> l scred
> c 0012 3c 3d
> c 05db 2c 24
> c 05f1 2e 22
> c 05f7 2f 23
> c 108d 30 31
> v
> -------------- cut ----------------------
> 
> You apply the patch with modpatch (correct file name?)  IIRC, you can't
> have blank lines in this file.  Let's see ... do you apply this with the
> module in memory?  Of course, you all know to keep a good copy of the
> original for in case things go bad.
> 
> To test it, try running scred before the patch and see if you don't get
> crashes with ALT-<key> combos.  IIRC, this will eliminate these.  There
> may be other bugs that this doesn't address.
> 
> Robert --  All I have -somewhwere- is some rudimentary disassemblies of
> the scred program.  I didn't do a very good job of disassembly, and
> really narrowed the "better" disassembly down to this bug.  If you are
> really interested, I can probably look up what I have.  Would you be
> wanting to see the complete source for scred?  FWIW, from my messing
> with coco scred and OSK scred, it appeared that they were probably
> compiled from the same C source.
> 
> 
> 
I was just curious about the nature of the bug and the method of fixing 
it. However, it could be of interest if we had the true source (not a 
disassembly of C produced ml code) to create an improved 6309 version 
for Boisy's NitrOS-9 project.

Thanks very much for the patch!