That's not necessarily a good thing. Do you want to write Java for a low-resource platform and incur the ridiculous overhead not to mention not being able to know what actually goes on? Both languages have their own strengths, and neither one is going anywhere soon.

I agree that learning a language is a lot easier in a text editor, but when programming, on the other hand, is much helped by an IDE, because it takes care of all the little nit-picky type of things for you.

I've never once used an IDE for C, C++, D or Ada. I'm quite fine without. When I do go with an IDE, I usually use Netbeans 6.9.2 along with Java, Ruby, and Groovy. Ruby I can almost absolutely say that an IDE is the way to go, as there are many functions built in that would be torturous to use without either autocomplete and/or in-IDE documentation. Netbean's documentation of Ruby commands is quite superb, and while it is almost directly ripped from ruby-doc.org, it's much easier to access and I love it so.

An IDE is much more than just a souped-up text editor. Debugging the code is at least as important as writing the code in the first place, so having quick access to debugging tools (which typically integrate with the text editor) is extremely useful.

If you're having to resort to your IDE to autogenerate most of your code for you then I'd say the language has a pretty big problem.

I don't have it autogenerate anything non-trival, I just might forget the small difference in something like string.partition and string.rpartition, so I can press ALT+ENTER to bring up suggestions and I can compare the two and see which one would allow for me to produce less hassled code.

I'm not saying that the IDE autogenerates all of my code, it just makes coding a lot less work. For example: when I type in the first part of a method , it will give me a menu of all the available methods for the class that fit. This saves time, and also helps when using Java's extensive code libraries.

And like dinosaurs, C powers everything from cars to planes to gundams, and even to the very computer you're posting from.

(Can we please stop talking about Java? It is not even tangentially relevant.)

Knowing nearly nothing about CP/M, I think it is most certainly feasible. I think porting CP/M to a z80 calculator would be an excellent endeavor.

As far as programs go, it's more than possible to rebase a program to get it to run somewhere else, especially programs a simple as what CP/M would be using.

You'd likely even be able to retain compatibility with existing programs, to a degree.

Your main issue is going to be handling I/O, both user-side and file-side. As previously mentioned, CP/M wants an 80-char wide screen. That'd be one pixel column per char on an 84. If you want to retain compatibility, you'd have to add some sort of scrolling for the screen.


Also, once this OS is on your calc, how are you going to get files to it? Linking is dependent on Ti-OS.

As far as your poll goes, yes, it's possible. Feasible, no.

(Also, java sucks.)

Unfortunately, plants are what coal and oil are made from, not Dinosaurs.



1. Contact Dan Englender, creator of USB8x, and ask him about how the calc-calc protocol works.
2. Write an app for TI/M for "calc-to-calc" transfers.
3. Write a Java app for the computer using the jUSB libraries. (Will trick the calc into thinking that another calc is just sending it data)

Nobody said anything relating to plants, coal, or oil. And about C, Java relies on C because many JVMs are written in C/C++. Back on topic about TI/M (which doesn't use java)...

1. Ask BrandonW also (as that's where I got my USB driver help from).
   
2. Writing a TI/M <=> AOS will be harder to do, but TI/m <=> TI/M will be easier. *wink* Silent Link transfers *wink*
   
3. Firstly, you will not use USB8x code, but instead periph8x (which may not work as dev. stopped on it) because your OS will have to run as a USB peripheral.

If you want to do USB, then look at Linky for host/peripheral modes and usage

Thanks AHelper, can you explain:

If you want to send/receive data to/from AOS (TIOS), then you will have to have both periph8x and usb8x working, handle silent link transfers and virtual packets, TI style, and you will have to handle the data storage types (appvars, apps, lists, w/e it throws at you).

With TI/M to/from TI/M, you can make your own protocol doing only what is needed for simple file transfers (still periph8x and usb8x).

I looked up where to find perph8x, and I found a site with this directory listing:

calc.asm
equates.inc
header.asm
hook.asm
interfaces.asm
log.asm
logview.asm
low.asm
msd.asm
periph8x.asm
periph8x.zws
recvio.asm
routines.asm
settings.inc
ti83plus.inc
util.asm

What do I need from this? Where can I get docs?

You'll have to use an assembler like Brass (inside the Doors CS SDK) or ZDS (which is BrandonW's assembler of choice, and would probably work best for this).

Which ones do I use? There's no documentation.

I would read up on how CP/M works and get exact specifications, such as RAM locations used/free, OS calls, output specs, and more. And, I would read through those things you found to see how they work and what the code does If you can't understand it, pastebin it and we'll look at it.

I once read a book on this... That's how I know about it. It wasn't too specific though...

I once heard of a site called "The Google"... you should look into it.

I found a link to that obscure site, "The Google", and I found some pretty decent documentation: http://www.cpm.z80.de/drilib.html

EDIT: I personally would look through every scrap of information in that site, it has everything from assembly documentation to how the OS works and even how it uses resources and more. The things on this site are nice and specific, and tons of resources at that. However, I still don't think it's a practical project