Re: A92: Rom dump? ATTN: Shawn Prestridge
[Prev][Next][Index][Thread]
Re: A92: Rom dump? ATTN: Shawn Prestridge
davidell@ix.netcom.com on 01/15/98 11:01:56 PM
Please respond to assembly-92@lists.ticalc.org
To: assembly-92@lists.ticalc.org
cc: (bcc: Shawn Prestridge/TXN)
Subject: Re: A92: Rom dump? ATTN: Shawn Prestridge
>>There are 12 versions of the 92 in the US,
>Is this really true? I've never seen or heard anything about ROMs 1.1,
1.6,
>or 1.9. Do these versions exist? If they do, were they ever released?
Yes. Not all were released for mask ROM production, because just as these
releases were prepared, someone found a RUF (random undocumented feature
8-) )
that they wanted fixed.
>>In a nutshell, what Fargo does
>>is play off of a bug in the 92 GraphLink software. David Ellsworth
>>figured out that if he did a backup of his 92, and then stuck
>>some machine code at the end of it, he could send the whole
>>thing to the 92. When the 92 hits the end of memory, it starts
>>overwriting the first sectors of memory, which allows him
>>virtually unlimited access to all areas of the calculator
>Actually, Shawn, I don't use that method anymore :) I've been using an
>alternate, much cleaner method for quite a while now. It takes advantage
of
>the fact that the memory in a backup file intersects with a global event
>handler callback variable that the TI-92 OS uses. Whenever a GUI event
>occurs, it reads this variable and calls the routine to which it points.
Sorry, I did not realize that you had switched methods!! However, in
contradiction to what NDStein wrote, that IS a defect in the GraphLink
software. GL is supposed to check the size of the image before sending,
and verify that it is of the correct length. As you mention, it obviously
does not!
>I'm curious, how did four object files and a smidget of C code get into
the
>ROM 2.1 image? The object files are aptabled.o, dialog.o, graph3d.o, and
>win.o. The C code is probably from lists.c or something. When I found it,
I
>came up with two theories: 1) that it was a present from some anonymous
>employee(s) at TI, and 2) that it was a mistake caused by uninitialized
>memory that somehow got past everyone and ended up in the released ROM.
When we started working on the 92E, we knew that we were going to make a
TI-92+ (of course, we did not know the nomenclature of the product at that
juncture). Some programmers who worked on the 92 went on to other projects
we are developing, and their modules were subsequently adopted by the
remaining programmers. The remaining programmers have carte blanche to
move the modules in any way seen fit. It is possible in so doing that
somebody screwed up a makefile.
>Also, does TI plan on using the Line-F exception on the 92plus? I'm
>thinking of using it for Fargo II, but I wouldn't like to conflict with
>anything the 92plus does.
Sorry, can't answer that one!! But hopefully in the not so distant
future, you will all have a 92+ module, and you'll find this out for
yourselves!!
>If you're not allowed to answer either of those questions, I understand.
>But if you *can* answer either or both of them, I would appreciate it very
>much. :)
No problem. There's no harm in asking, so feel free to ask!!
>---
>David Ellsworth
>davidell@earthling.net
>IRC: eXocomp
>ICQ: 2300673
Shawn A. Prestridge
Texas Instruments, Inc.
Educational & Productivity Solutions
7800 Banner Drive MS 3908
Dallas, Texas 75251
(972) 917-1698
(972) 917-7103 FAX
"He's back... Kicking bottom or what?!?!" --Holly, Red Dwarf
Follow-Ups: