To: perot@pallas.amp.uni-hannover.de
Subject: Re: [5] New StiK (also Megafile 20)                                             


On Sun, 29 Dec 1996 14:29:11 Peter Rottengatter <perot wrote: 
>
>On Sat, 28 Dec 1996, Ronald Andersson wrote:

----- snip ----- re: LocalTalk on MIDI port

>Concerning the Midi port, you might have noticed that I got another 
>proprietary protocol in mind which I called MidiNet. You'll find a dummy 
>driver for this in the package I sent. It is a many years old dream of 
>mine to connect Ataris via Midi in a ring shaped network, just like the 
>MidiMaze game requires. This will have a few features from the IBM token 
>ring, but different in other respects. I have a fairly clear idea how it 
>will work, but it is such a low priority again ...

Yep, I noticed, and it sounds interesting.  This is one of the many things
we can start working on once the basic internet functions are debugged.


----- snip ----- re: LocalTalk speed dependencies
>>
>> That would be a problem of course, but it sounds rather messy to begin with.
>> This means that the protocol will be extremely sensitive to any variations
>> in execution speed too, since that determines when the software will actually
>> 'see' the characters as received.  Are you sure that this char by char timing
>> problem is not actually the result of a too short hardware buffer.
>> (Same problem as we would have with the one-byte buffer that is.)
>
>I dunno, I'm not perfectly familiar with LocalTalk either. I added the 
>dummy driver because Dan said he'd like to have such a thing, and I know 
>it's quite a fast and efficient network, and most of the hardware is 
>there on many machines.

Ok.  This is not the time for any final decisions on such matters anyway.
Once the new STiK has been debugged sufficiently to replace the old for all
current clients, with all current ports and protocols, that will be the time
for starting the experiments with entirely new features. Until then we can
only speculate, and try to keep the interfaces extensible for future needs.


----- snip ----- re: Midi port SLIP/PPP for 2-computer network
>> 
>> Right.  I'd probably prefer the SLIP/PPP approach anyway, once we get some
>> appropriate client/server software.  Until then I'll have to use DUET and
>> similar utilities instead.
>
>Saves you a few kilobytes too, because the driver is included in the 
>SERIAL.STX. But I guess you're not as short of kilobytes as Martin ? ;-)

Hardly...  I have a 16 MByte SIMM in my Falcon and 4 MBytes in my Mega ST.
Actually the drivers planned should be affordable by practically everyone,
especially in the final assembly versions.


----- snip ----- re: FSERIAL.CPX by Roman Hodek (was erroneously Bolek)
>>
>> Yes, but if he did give you the authors name correctly we can perhaps also
>> find him through an Alta-Vista search for (Roman NEAR Bolek).  If he has a
>> home page it probably contains his name somewhere, and with such an unusual
>> name there should not be too many irrelevant 'hits'.
>
>Dunno, maybe it's worth a try. Do you wanna do it ? I just looked up 
>again Harun's mail, the name was Roman Hodek, not Bolek. I'll do an archie 
>search for fserial too.

I planned to do both at once before but something else intervened, and I still
haven't gotten around to it.  I guess you've already tried it by now, and I
hope you've managed to reach him.

Hey!  I just thought of checking for him again (with the correct surname) in
the AIP hypertext, and I _found_ him...!!!

Here's all the data (some is hopefully still valid):

<Roman Hodek
<
<  eMail:  Roman Hodek @ N
<          roman_hodek@n.maus.de
<          Roman.Hodek@cip.informatik.uni-erlangen.de
<
<Programme:
<
<- Suprima Mailbox-System

This does not mention any "FSERIAL.CPX" either in the works or finished,
but the program mentioned is at least a comms-type program, so I'm fairly
certain that this is the guy we're looking for.

BTW: Is it really legal to have a "." character in the user name of an email
     address, such as is used in his Internet address above...?
     Perhaps it should be a "_" instead (but AIP uses the ".").

-------------------------------------------------------------------------
Regards:  Ronald Andersson                     mailto:dlanor@oden.se
                                               http://www.oden.se/~dlanor
-------------------------------------------------------------------------
