Re: TI-H: A Virtual Hub: Interlink


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

Re: TI-H: A Virtual Hub: Interlink




Then show him this email and ask him to explain it, since I can't speak
non-technically about technical stuff.  sorry I couldn't be of any help.

CK

Miles Raymond wrote:

>  >there are a few better ways to do this.  First off, the crash
> control is a nice idea, >but very Ethernet-like.  Ehternet is fine,
> but these link portss hit around 5->6K/sec, not 10/100 Mbps.  Token
> Ring based systems work better under low->performance/low-bandwidth
> situations, because only one calc can send at a >time.  Look at the
> MBus routines available for a good software only TI-calc >networking
> protocol.  There's already a chat client out, but unless we get
> "TI->jacks" at every desk in every school, it won't really make much
> sense.  But I like >it.  Very IRC-like, and can be used by over 200
> people at once.
> >For a simple connection system, you want to make the three parts of
> each link >port on every calculator the same point, electrically
> speaking.>>For this, use a "T" or "L" based system.  Your physical
> layout isn't all that bad, >but you may want to put in some sort of
> terminating circuitry (a resistor).  The >pass-through approach is
> great.  But don't count on the "standard" link cable.  >TI changed
> it.  I just saw a TI-86 being sold at Price/Costco with a new link
> >cable design.  No more 90-degree offset, it's a straight cable
> now.>>CK You might want to know that I have just recently joined
> ticalc.org, and have no clue what you are talking about.  I guess I
> should have made it clear in my letter, that I have no electrical
> knowledge whatsoever, and I was only posting my idea so (hopefully)
> someone would see it, and make it.  I have talked about it with David
> Philips in school, and have gotten a few of the previously posted
> ideas from him.  He is a big help since he knows more about hardware
> than I do. I don't think I have gotten a clear answer from all of the
> responses, though.  Is my idea possible via software?  David and I are
> disputing it, but I truly think that it's possible.  Granted, I know
> nothing about how TI makes their link work, but my dad says that it
> isn't possible to create this kind of network via software, and he
> loves the hardware design. About the master Interlink, this would be
> determined by the calc connected to the 'virtual hub' first.  An
> example of setting an Interlink cabled network: 1) connect all of the
> Interlinks to each other2) connect the master calc (calc with ID #1)
> to one cable attached to the 'virtual hub'3) connect all other calcs
> in the network, one by one, each is given an ID according to the order
> in which it was attached to the network OR 1) connect one Interlink
> cable to the master calc2) connect another Interlink cable to the end
> of the Interlink cable attached to the master calc3) connect another
> calc to the end of the Interlink cable4) repeat steps #2&3 for every
> calc remaining. I hope this clarifies any information I have retained
> in the previous message. -Miles Raymond




References: