Re: Universal Link Library


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

Re: Universal Link Library



Mmmm... that appears to be some false info.

>Well, yes and no. one wire(red I think) sends data, while the other
>sends a strobe. If a one-wire protocol was developed(using low level
>hardware calls) data could be sent two ways, or could be sent twice as
>fast. Also,  some sort of error checking and fixing could be
>implemented.

Nope. SCSI and Parallel port have a strobe ability. that is, you can
tell the pport to turn on the strobe. can be easy for devices without
their own clock. However, the TI link port is very simple and to the
point. more importantly, every single TI calculator, inluding the 92,
has the exact same linkport. you can communicate between them just as
if it was another 85/86/whatever. see TI-Tetris. There isn't any
'special low level hardware call' used to send data two ways. also,
everything in asm is low level :-).. It sounds like you are trying to
sound complicated while the linkport isn't.

Take your calc<->calc wire. start from the outside. That is the Red.
then there is a blue band, which doesn't conduct. the next one is the
white. another blue band. the last one is ground.

Hook up a voltmeter. when the wires read 11, you can measure +5 Volts
between the red and ground, and you can measure +5 Volts between the
white and ground.

Error checking has NOTHING to do with how the linkport has been
designed. It's completely unrelated.

The protocol is simple. a white and red wire, 1 means there's +5V
between the wire and ground. 0 means there's no volt between em. All
calcs have this protocol inherent.

>Right. Plus if other features were needed, the protocol would stop being
>standard.

Exactly. in other words, protocols are quite useless for the TI-calc
series. That is, universal protocols. it's already there. A real
'protocol' can only exist between two exactly the same games. ie:
ti-tetris/85 and ti-tetris/86 and ti-tetris/92. these calcs don't
detect what calc is on the other line and send appropriately- they all
send the same protocol, and that protocol hasn't been particularly
modified to work with all 3 models.

>Of course it would. And not just for game data: Imagine being able to
>transfer equations, variables, matrices, formulae, etc. between ti8xs
>and ti92s.

Please read the thread. the 'protocol' suggested would somehow
magically make all games work together. Such a particular feature is
quite simple to write. take any variable browser shell, and write up a
little protocol. However, there is not much use for sharing data
between different types of calc. If there would be a need, people
simply transfer it to their computer, rename it, and send it to the
other calc. Unless you need to constantly share data between, say, the
85 and the 92, without a computer ready, and in amounts too large to
just copy, it would be of use. quite limited, imho.


References: