Building OSs for Flash ROM Calculators
|
Posted on 6 September 1998
The following text was written by Matthew
Stits: When one looks at the evaluation of the TI series of calculators, one
sees more and more people trying to push the envelope of what one can do with them. At
first, just a few basic games, then assemblers, all the way up to memory expansion kits.
With this in mind, TI did make it a bit harder to make an assembler on the TI series with
the 92. I remember many people discussing the problems (of which I do not recall the exact
reasons) which gave way to making fargo a very stable shell built on an Operating Sytem
never intended for it. With time and the presence of fargo, TI has seen that someone will
always find a way to get around what ever obstacles presented and has now put an assembler
on their TI-89 and TI-92 Plus models. At first this would seem great for the TI's.
In one single step they have erased the need of so many people who enjoyed their work. Now
after explaining some of the history to this saga, I feel TI has given themselves a bit of
an Achilles heel. With the ever growing cost of the college student's calculator, TI said,
"Hey! Let's put Flash ROM in so they will only have to buy one calculator for a little
more." With this in mind a hole was opened that none had previously thought about. Why
doesn't someone now make a complete OS for the calculator? It could be anything from a
small unix box, to a full fledged GUI OS. Here I'd like to present some examples of it why
it should be done. All the registers are out and I am sure that a 10 MHz chip is more than
enough for a GUI interface or at least a basic lunix shell to start from. I think
that the biggest problem would be in making a joint inter face for both the TI-89 and TI-92
Plus. It would most likely have to be recompiled for each version with different specs for
the first few builds until a set amount of memory is dedicated to output for the LCD screen.
There are at least 3 OS's made from this chip and its children already! Mac OS, Norton "that
pseudo Palm Pilot" and Sega's very basic ROM reading OS for its genesis and probably a few
more. This is by no means to say the that Fargo has no purpose, but what if they made it
into a full fledged OS and not a shell on top of an OS never intended to work in the back
ground? When looking at this from the a different angle, one sees a few possible
problems. Some (actually most) of us don't have a TI-92 Plus, so Fargo is all that many can
use. Fargo is probably a lot better planned than whatever TI had made. Fargo can use
libraries, make TSR's, and many other things that I for one doubt TI put that much work.
There are already many good programs for Fargo "that could be ported at a later date". As
for making your own OS for the calculator, all you could do is turn it into what most (at
least at first) would consider a novelty or GameBoy, not to be taken seriously. Why
reinvent a calculator that TI paid lots of people to make? I believe a person or small
group of people not getting paid would make anything as good or better.
|
|
Reply to this item
|
Re: Article: "Building OSs for Flash ROM Calculators"
|
Andrew Hockman
|
There is a big problem here that is being overlooked. In the past, TI calculators were made with true ROM... read-only. They were confident that when the technology became outdated, consumers would buy more calcs. However, the new "flash rom" gives the possibility for an almost infinately upgradable calc. This would never be good for business, so TI added a little "feature" to the new flash calcs... a limited number of upgrades. Theoretically, at some point in the distant future the calc will completely shut down and refuse to work (at least this is how I understand it). Consumers are then forced to buy more calcs. Perfect... almost. While this may be fine for normal users, in my opinion TI has totally neglected all the programmers and developers that drive the popularity of their products. How many programmers out there want to write, flash, crash, restore, modify, flash, crash restore, and everything else necessary to write quality software if they knew that each calc crash was "eating up" a little bit of their calc's limited life? I'm wary.
|
Reply to this comment
|
7 September 1998, 05:35 GMT
|
|
Re: Article: "Building OSs for Flash ROM Calculators"
|
Aj
|
Well, I have a lot to say but i am going to make it short cause i have to try to get to sleep. All of us here are talking about making an os and how cool it would be, nnow, not to imply that i know anything about 68k asm, but i am will to start a group to begin searching for ways to do this os project, I do have a few ideas on how to approach this project. my e-mail is: ajp259@hotmail.com
Somethign that i think that we NEED to talk about it how in the last post the guy said that there is a limmited number of upgrades, this needs to be investigated this is huge, i really need to know wwhat is up with this. I will nwo say anymore about how serious a limit on upgrades would be if there is infact a limit untill i find out in wroiting where that is from ti, and what the actuall number is, okay, I am sorry for all ther mispellings there must be in this thing, i was typing very fast cause i really do have to get to sleep,
later
-Aj
|
Reply to this comment
|
7 September 1998, 06:34 GMT
|
|
Re: Article: "Building OSs for Flash ROM Calculators"
|
Jeff
|
Okay, I have a thought about emulating a Mac on a TI-89 as a second OS:
The orriginal Mac was less powerful than the TI-89, at least memory and speed wise. For those of you not familiar with it, it used a 8MHz 68000 with 64k ROM and 128k RAM. With the memory available in the Ti-89, I believe it would be entirely possible to emulate one of these orriginal Macs, even though it would be rather pointless. (The only programs I have that can run on one are System 1.1n and a database program)
|
Reply to this comment
|
7 September 1998, 07:26 GMT
|
|
Re: Article: "Building OSs for Flash ROM Calculators"
|
Jonas Minnberg (Sasq)
(Web Page)
|
First of all: MacOs on a TI? What's the point?
Without a mouse and with such a small screen it
would be pointless. What we need is a unix-like
OS, and that shouldn't be impossible. Also, a
few 68000-based OS:es was mentioned, but not
the (IMO) most interesting; Amiga OS.
There's a lot to learn from this OS that was
(still IMO) superior to all other OS:es in the
late 80s and early 90s - mainly because of the
fast multitasking and the fact that the entire
OS was composed out a few, well-documented "libraries" (known as DLL:s in the windows world).
I have been thinking about how to make a good OS
for the TI92 for some time... If I ever get
around to it, it will be unix-like with a micro-
kernel (something along the lines of QNX perhaps)
with very simple functionality (priorityless,
round-robin preemptive multitasking) but with
the ability to override and extend everything
with new, more advanced funtionality. This would
allow for a very small OS if all you want is the
ability to run some games (say a kernel around
1-2k) but with the ability for a full-fledged
unix.
I also think that anyone considering developing
an OS should think twice about it; it is not
a very good "first" project. I have been
programming for many years, most of the time in
68000-assembler - and I've a good deal of
experience with unix-like OS:es and how they're
built but I still feel it's a very tough
project, especially if you want to get it "right"
and not make all mistakes again (even with lots
of money, time and experience this may happen,
just look at windows ;).
About the flash-limit; If I start this project I
will of course use my emulator as much as
possible :)
//Jonas Minnberg (Sasq)
|
Reply to this comment
|
7 September 1998, 13:40 GMT
|
|
TI-89 GNU OPEN TI OS
|
Nathan Cassano
|
First off I do think new rom's are very possible. I person dumped their 89 rom mistakenly and it still booted, but with no extra algebra and calculus stuff. There is a TI OS ROM (size=?), then added software ( 250K-500K ), user archiving ram ( 384K ) and system ram ( 250K, 180K of that available to the user ). That's almost a meg of ram ( the Palm Pilot has 2 megs ). You would have to boot to the TI OS then exec a boot loader from there. Here is my proposal for a new TI-89 OS, I wanted to voice my opinion on what the OS would be like. The entire idea and purpose of an OS would be to make the TI-89 as open and as accessible as possible. The current OS, that TI provides doesn't give us the power we desire out of our calcs. We need more. We need to push our TI's to the max! Here we go.
TI-89 GNU OPEN TI OS
Proposal
PC to TI connectivity station
External GNU C/C++/ASM Compilers ( why aren't we using these with good libraries? )
Stable System Kernel ( does the system chores, interupts, blah, blah )
A standard programming library
File System (with compression options)
GUI and text Shell interfaces
Multi-tasking
Networking Protocal
Standard system DLL's program interfaces
Security System
Scripting language (similar to TI-BASIC but more power and robust, like
Java Script)
Word Processing and other associated applications ( extras )
|
Reply to this comment
|
8 September 1998, 07:23 GMT
|
|
1 2 3 4 5
You can change the number of comments per page in Account Preferences.
|