Re: Bug found in C64?

From: Spiro Trikaliotis (ml-cbmhackers_at_trikaliotis.net)
Date: 2005-12-06 07:17:27

Hello Ruud,

congratulation, your Pegasus Mail did not break threading! ;)

* On Mon, Dec 05, 2005 at 09:09:41PM +0100 Ruud@baltissen.org wrote:
 
> But the point is that this is a bug indeed. But I cannot remember
> anyone mentioning it. And I never experienced a wrong behaviour when
> outputting strings. But OTOH I haven't found out yet under what exact
> circumstances this behaviour will occur.

It is simple: If you open a file with a logical file number < 128 (that
is, bit 7 = 0), then all text outputs should be as you are outputting
them. This is especially true for CR, which just remains a CR.

This is not true for logical file numbers >= 128 (bit 7 = 1). In this
case, this part of the BASIC routine wants to make sure that a LF is
appended after a CR.

I do not know anyone who ever used that behaviour. Some printers come to
mind, but I do not know if they were heavily used. Anyway, this wrong
behaviour could only be experienced if this feature was used.

BTW: I did not check if the KERNAL does that CR -> CR/LF conversion,
too. IFF it does, then this would not be an error anymore.

BTW2: The "wrong" JSR does nothing bad, as only an EOR #$FF; RTS is
executed, and the Accu is not used afterwards.

Regards,
   Spiro.

-- 
Spiro R. Trikaliotis
http://www.trikaliotis.net/
http://cbm4win.sf.net/

       Message was sent through the cbm-hackers mailing list

Archive generated by hypermail pre-2.1.8.