Re: TI-H: TI Network - TIN

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

Re: TI-H: TI Network - TIN


>At 09:35 PM 10/19/98 -0800, you wrote:
>>
>>>1. As a programmer, you should know that a compiled app for one calc will
>>>not run on another without an "emulator" of sorts.  I don't think that an 86
>>>prog would run on an 82, do you?  And what about the 89/92?  They even have
>>>a different processor!  How the heck could you run an 82/83/85/86 prog
>>>natively on an 89/92/92+?  The ZTetris linking protocal is the only thing
>>>cross-calc about it.  If you don't believe me, find someone at school with a
>>>different calc than yours without ZTetris, and try to send it to them.
>>
>>Yes.  But porting an easy program like the I2C driver is easy.  Most of it
>>is simple timing and I/O.  Comeon.  As a programmer, you should know that.
>>:)
>>
>>>2. I think that the best way to handle multi-calc storage is for the calc to
>>>attach an ID to the stored file, so another calc could see that "hey, this
>>>just isn't my thing", and wouldn't run it.  I do not know if things such as
>>>lists and matrices are cross-calc or not, but I imagine that an 82/83 list
>>>would need to convert to an 85/86 list, which would need to convert to an
>>>89/92 list.  (I hope I am wrong here.)
>>
>>The lists can be stored anyway you want.  It depends how you store the
>>bytes to the EEPROM.  If you knew the formatting you could just store it in
>>a generic format and then format it specifically for each calc when the
>>driver writes it to ram.
>>
>come on grant!  we better be using your sram idea.....

DRAM idea?

>>>3. And about addresses, this just means that more calcs can attach, right?
>>>So, if everyone in school that had a calc, they all would be able to use
>>>this network?  How is this going to be implemented?  I don't think that
>>>someone is acctually going to build a 200+ hub/connector.  Or, is it
>>>possible to connect multiple hub/connectors together?
>>
>>I'm working on an ethernet to parallel/I2C converter so you could just
>>subnet I2C networks using the UDP protocall.  :)  That would be cool...  If
>That wouldnt be cool........ that would be ___KICK ASSS_______!!!
>
>
>>I spent the time to figure out the new ARM+NET chip we could have dial-in
>>access for calcs and they could use the net.  :)

I've decided on 10BASE-T and I might do 100BASE-TX if National will send me
those...


>>I think a special 'basic' (or such) should be written for all calcs, but
>>programs can be run off the I2C net.  THe programs could be tokenized and
>>calcs like the 85/86/89/92 could just stretch the screen output to size an
>>82/83 calcs.  Just an idea...  Then every netprogram would work on any
>>calc...
>and idea indeed.  it would be an alright idea, but it would be pretty slow
>and stuff.

Not too slow if it was used for databases or presentations...
References: