RE: A89: LinuxTI & Memory Expander


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

RE: A89: LinuxTI & Memory Expander



Title:

See comments given below.

Mark E. Scott Jr.
mscott@dbcity.net


-----Original Message-----
From: KKokiri89@aol.com [mailto:KKokiri89@aol.com]
Sent: Saturday, July 17, 1999 8:41 PM
To: assembly-89@lists.ticalc.org
Subject: Re: A89: LinuxTI & Memory Expander



Great idea! Yes it was mentioned earlier, and i kind of put it aside. That
should probably be a priority, since we would likely have to cut out
functions and features of the programs to reduce size. Also this would act as
a separate "partition" or perhaps drive which you could mount and unmount,
and either store user defined programs or other things. If we end up making
our own calc linker (we will probably have to, if we want to send .exe or
.txt files), we should make it so that we can send files directly to the
extension to make sure that we dont have to load files on the the calc which
are too big, then have to send them to the expansion.
Here are the Top 10 Concerns of this project : (note that i am getting ahead
of myself...but linux took a long time of planning too, so why not?)

- People who use the original TI-OS will get cheesed off because current
games such as SMQ and SF2T etc... will not work on this OS

(Response) Who cares? New programs will come out quicker for the Linux port due to its "neatness".  So many programs used to come out for the ti-85 zshell when it came out because it was so different and cool.  I believe this would generate the same interest.

- TI will be angry if we break there key (and we might have to use parts of
their OS for mathematical purposes.)

(Response) I am of the opinion that we should just make our own boot loader, I could easily make a link program for the pc that is custom.  I haven't understood enough about the bootloader to be sure, but what I think is that you only need the key if you plan to use their bootloader.

- There won't be enough memory and people will have to purchase memory
expansion kits (hopefully you wont charge us _too_ much :)

(Response) I think this is unavoidable, besides most people I know would like an expansion but have no need for it.  Once they see the things we can program and will program, it will become well worth it.  At first it may be deemed not worthwhile, but everything starts out that way.

- We may not even be able to get it up and running with/without ditching the
TI-OS

(Response) This is just my opinion, but TI-OS could go.  We run into problems here if we want to let people run it as a program.  Perhaps we could make it work - somehow convert it to a linux capable program, but I doubt that.  It might be easier to make our own program, but TI has done a good job making all these math functions.  This is the only issue I really see a problem on

- We will have to be able to have a BASIC programming language, otherwise non
ASM people can't contribute

(Response) One word: PERL

- More likely than not, we will have to rewrite any software to include grey
scale, floating point, and black and white (sigh), as well as, the link port,
odd networking hardware (Mark, maybe you could make some Networking adapters
too?) as well as stuff i wont argue about (like the ROM, i enjoy learning
about it from you guys rather than explain it:-)

(Response) This is done in the libraries, and would be essential to the OS.  While this isn't negating your point, its just gonna have to be and has already been talked about I believe.  I believe there are simple network adapters that you can plug into a serial port, but not sure, this is kinda a big if.  Depends on the port speed.  If not you'd have to hook directly into the calc which isn't quite an easy way to do it.

- We need to create drivers and such for our keyboards and link ports and
LCDs (not a big problem i dont think)

(Response) You're right, not a problem compared to porting to OS in the first place.  Don't know about LCD's though

- Other calculators will want this Linux TI for Z80 - could we write ROMs for
that? I'm thinking not.

(Response) I believe TI-89 could be fairly easily ported to due to its being nearly the same calc, but others are out of luck.  This would generate want for the ti-89 and 92 (I prefer 89 though)

- EXTENSIVE bug testing...of course this is the ROM and shouldn't be cleared,
but then again...how did we get it on there (talking from the future :-)

(Response) This is involved with creating any OS.  Big thing is to make sure our bootloader is perfect, I could put my rom back in, but I doubt most can normally, and its a pain in the A$$

- Hmm, okay i didn't make it all the way to 10, but i don't want to diss this
project at all...nor do i want to waste valuable time (i still haven't made
any ASM programs :-(.

(Response) This would be a very good project, and is probably doable, and I bet would get a lot of attention.  I can see big things for it.

Lets go onto the good points...

- We have much respect for getting such a tremendous task completed

- Every could learn about programming who would like to contribute

- Our calculators aren't any longer "a waste of skin"...not that they are if
you need it for math, but TI doesn't exactly push the hardware's limit...its
too easy i guess...(let me know what you think, even though they are probably
all flames :-)

- Make this mother WORK!!! AHAHAHA!!! (excuse me)

- Networking capabilties, with file sharing and permissions

- Much better application development

- The source code is freely free, if we all agree to this lets have it be a
GNU program/OS

- We could do plenty with the different memory types...the ROM is for the
Kernel
  The RAM is just that- the RAM, and the extension (hopefully we get one) is
for programs, like a hard disk

- Okay, this is very very long...(at least it looks like it) so ill sum it up

I dont think it really likely to happen on this calc, but im not in the
position to judge that yet...its real hard to say.
Nothing has been done to work with the ROM except AU (archive utility) and im
not sure if thats even ROM?
Hopefully we DO get passed this issue...if not i have a back up plan!
Perhaps there is some bug waiting to be found in the TI-OS that lets us go
straight to the processor (or am i living in the past...help me out here, the
TI OS handles the ASM and you can't directly program for the 68K...its more
or less programming for the TI-OS?, whatever any way what i was saying) and
we can comletly ingore, disable temperarly, whatever to the TI OS. Then we
can have the program be stored in RAM for the TI part of it, then Load it
self into the expansion, and whatnot...then go from there. (that was most
likely the stupidest sounding thing i have said...or maybe not :-)

Well, I wish you all the best of luck and i hope this opens up to be a great
"hack" that TI wont be happy about, but we will be going around like nerds
with Linux on every thing we own...including our watches (yes im over
exagerating...but you never know..they do run some smart VCRs)

With any luck, we will be discussing this developement for a good long time,
perhaps we should start our own list so the people that dont want to help
dont have to listen to our rambling (I feel like a spammer or something) but
i really hope to see every one in on this...after all, its our calcs, lets
customize!!

(Response) I could easily create a list - I work at an ISP and have access to any of this kind of stuff, and I'm sure ticalc.org would do it too, but we wouldn't have as much control of it.

And if we get a large enough expansion...(im still not real certain on that
idea...its a great idea but i have a strange feeling that i can't
understand...a bad feeling that this isn't going to work and we will be come
the the village idoits, but i wont care as long as we all try at it...you
really can't say at this point though)...like a few megs...5 or 10...then we
really could start something revolutionary here!!

This has got me all revved up...and i hope it goes as well as i think it can

Regards,

Kyle


Follow-Ups: