Hmmm this is allgood, but it needs to be able to adapt, and it would also be cool to see a radio reciver for a hookup to a computer. i could plug it into the back of one of the computers at cool and surf the internet without anyone knowing =). Also the range needs to grow. Like a 2meter ham radio using a repeater can go like 25 miles ( dont need that mutch, but if i could get 2 miles out of it i could link up with my home computer from school.

Awesome work! The UHF spectrum seems like an excellent choice for this type of communication because you can use small antennas and the signals will cut through walls and other obstructions much better than VHF would. In the last post, the 2 meter ham band was mentioned. While much more power and flexibility is allowed to hams on their frequencies than on the unlicenced freqs., a few problems arise from using the ham bands for this. #1, most calculator users are not hams, and #2, the protocal used must be well published and it must incorporate a method of identification.

To work out the distance between home and school, someone with a near-by house could possibly provide an internet connection for the students of the school.

that could work i know someone that lives realy close. but i also thought if you could directly plug in one of the radio links to a parallel port or a serial port. Each room in my school has a computer all hooked up to the internet i could plug in one side of this thing into the back of a teachers comp and install some software running in the background and could be in a chat room with the teachers none the wiser, and without the major cost of a cell phone bill. any response to this idea. also the connection to the computer would be nice if it didnt require a graph link make it like those keys for some software that you just plug in.

If you have or have access to a parallel/serial link (not TI's) that uses a jack instead of a plug (the kind that uses a calc link as a pass through), you could just plug the RadioLinc into it just as you would into a calc. Using an emulator would work with this, but to do any kind of non-calculator related stuff (email, newsgroups, etc.) you'd need a program for that, which would just use the standard specs for a link program.

Are these ever going to be offered built for purchase?
It'd be a help of those of us who aren't electronically inclined... which end of the sodering iron do I hold? OUCH!!!... oh well

EverSpiffy

<humor>
Hmm, I did that with a Bunsen burner once :-)
</humor>

--BlueCalx

Not sure if you could get it to transmit further, of course you could always mount a tower on the top of your bookbag and transmit?

We probably will be selling these in a snazzy translucent case in the Y2K, price , unknown as of now.

May I suggest a black or grey case? I'de like to use these to cheat on math tests, and it would be easier if they weren't see through so I could claim that they are "memory expanders". How big will they be anyway?

Please, don't forget to put a connection for an external antenna. It can be very useful, a little wire going to your bookbag or something. Have you seen those super compact antennas that are tighly winded? those could work too.

Try a translucent ash-blue-i have seen some pretty cool stuff in that color

Heh...I actually did that once. A very LONG time ago :)

One word: "Wow". This is completely amazing. Now I'm totally convinced that the possibilities for a TI Calculator are endless. One question though: What happens when 20 people are talking at the same time? What will keep the signals separate so that only the person you intend to receive it does? Sorry about the wording, I hope it makes some kind of sense. LOL

We are working on a way to variate the frequency to select who you want to send to. Not sure how were going to do it yet, but we will. Maybe a type of IP address?

Well with the 89/92 you have a unique serial number you can use

Good point!

I was thinking of making somthing very similar to what you have designed, but I have never even written schematics for it yet. A solution I found to this problem was to set up one calc as a "server" and have everybody joining send a signal on a certain modulation. This "server" woud get the signal and tell all the other calcs to stop transmitting. It would then "log on" the new calc. The entire chat system would be based on a "token ring" type of communication where the "server" calc sends a signal to each other calc in order, requesting any data and giving it data. That way, the "server" calc could be in control of all transmissions at all times, stopping transmissions as necessary to log on newcomers.

Just a thought

Ty

I have horrible experience in such conditions (connected many PC-s over radio at same frequence). Even though that X.25 protocol was used, and lower level links know for colision, there were too much retransmissions that I needed up to 2 hours to transfer 2K file! The reason is that repeater listen all senders, but senders between them not. Senders think that ether is free, repeater receives trash, and resends packet.

I wrote a program for the TI-83 and TI-83 Plus that does this (BASIC). It allows you to chat almost in realtime.

You know... since anyone else in the room with a radio link can intercept the transmissions... wouldn't it be a good idea to include some sort of basic encoding of transmissions?

Yes, that is what we are working on now...

I'm an amature cryptographer and in my spare time wrote a BASIC-89 encryption proggram which is simple enough to be implimented without much proccessor or memory overhead. How ever, it is a symmetrical stream cipher, and the code generating algorithm maybe easy to break. However it is secure enough that only another user listening on a PC with a radio link and specially designed crypto-anylitical software made by a proffesional cryptographer would be able to decode the transmission. So its allright if you talk about "what's the answer to #3" or "do you want to go out saturday" but not "10 kilo's, china white".

The system has the user input a key, then seeds the random generator with it, or some function, possibly a Zeta funciton. Then a random number is generated, xor'ed with the next character in the plaintext and output to the cipherstream.

With ord, char, and mid, this is very easy to implement in basic on the 89, it takes less than 500b, and I imagine it could be easily implemented in ASM on other calculators.

I wrote a n encryption program that generates a random matrix, 10 x 10 of #'s 1-1000. It randomly places the quardinents of the quardinents of a user input number in this matrix. The whole program's about 800 bytes. Too big?