Re: A82: OS-82 bug/drawback


[Prev][Next][Index][Thread]

Re: A82: OS-82 bug/drawback



The misreported size of any variable do not have any influence on the
calcs speed ! The misreported size is due to an incorrect calue in the
size bytes of the var, and the calc normally does not use this at all.

The size byte is used to calculate the size of one singel variable. This
is done when the clac shows the size f one var (in the mem menu), and when
the size of that variable is changed. So unless this variable is used in
calculations it does not matter.

When the system allocates mem it does it in several ways, but the rutines
which have allready been found covers it all. Of cause there might be more
functions, but it is not likely. All the functions found in the ti82 rom
does not use the size bytes, but some pointers placed in the system part
of the memory. These pointers are not affected by the misreported size.

I am not do not know exactly how OS-82 is coded, but i seems very unlikely
that it affects the math operations. When shells like Ash and OS-82 are
not used, they act as normal variables. It is only when they are executed
they act different. So having OS-82 on your calc should in no way affect
the normal operation of the calc. The same problems has several times been
reported on the ti85. All these reportes has been due to programs with
bugs. If you run asm programs which messes up the VAT or the system part
of the ram, the clc might crash. These is however not a problem with the
shell, but with the programs you run.

Of cause the shells might include bugs, but it is not very likely taht
they cause the problems discreibed here.

Dines

_______________________________________

Dines Justesen
Email: dines@post1.com or
       c958362@student.dtu.dk
WWW  : http://www.gbar.dtu.dk/~c958362/
_______________________________________

On Thu, 21 Aug 1997, Thomas J. Hruska wrote:

> At 10:57 PM 8/21/97 +0000, you wrote:
> >
> >I of course am aware of this, your version is just outdated.  Get OS-82
> >version 1.1 from ticalc.org.  That fixed that little mishap, and as far as
> >I know, there are no other "bugs" in OS-82, which has been out for about a
> >month or so.  The "slowing down" of calculations is simply because of the
> >memory being used by programs.  If you have a lot of normal basic programs
> >on your calc and do a calculation, it's a LOT slower than if you did the
> >same calculation with reset memory.
> 
> I have to disagree.  The slowing down of the calc is probably caused by the
> mis-reported size of one of OS-82's variables in memory.  BTW, OS-82 has
> only been around 17 days...not "a month or so."
> 
> >At 10:31 PM 8/21/97 -0400, you wrote:
> >>At 08:59 PM 8/21/97 -0400, you wrote:
> >>>I have been using OS-82 and have found an interesting bug/drawback to using
> >>>it.  Whenver you use a table type function with a menu bar (list editor,
> >>>matrice editor, y/x table), the menu does not show up on the screen.  If
> you
> >>>press graph to look at the graph screen, the menu is on there where it
> >should
> >>>have been before.  Can anyone else explain/verify this?  Does this happen
> >>>with OSHELL-82 or ASH?
> >>
> >>This problem is due to the fact that (IY+$14) bit 7 is SET.  The only way
> >>to correct it is by changing the exit code on the shell so that it is
> >>reset.  Talk to Jason Todd about this.
> >>
> >>(IY+14)
> >>Bit 7 set - Write text to graph mem
> >>Bit 7 res - Write text directly to LCD
> 
> 
>                  Thomas J. Hruska -- thruska@tir.com
> Shining Light Productions -- "Meeting the needs of fellow programmers"
>          http://www.geocities.com/SiliconValley/Heights/8504
>                     http://shinelight.home.ml.org
> 


References: