Re: project BIG AppleCrate

["Michael J. Mahon" <mjmahon@aol.com>]

aiiadict@gmail.com wrote:
> On Dec 30, 1:24 pm, "Michael J. Mahon" <mjma...@aol.com> wrote:
>
> > Likewise--and that's the option that currently requires the
> > $400 bytes of ROM space taken up by the self-test code in
> > unenhanced //e's.  (The enhanced ROM replaces the rather poor
> > code in the unenhanced self-test with an improved $100-byte
> > self-test--not enough to hold the NadaNet equivalent of "RWTS".)
>
>
>
> and I have a ROM burner..

Good--you'll need it.  The modified $C000-$DFFF 2764 ROM image is
included in the disk images on my site.  ;-)

> > > contiki running webserver, you could request page "RUN PROGRAM".
> > > webserver sees that it doesn't exist on disk.  or prefix any nadanet
> > > commands with NADA or something similar
> >
> > Does this permit remote running of programs on a ProDOS control
> > machine?  And how about user-supplied programs?
>
>
> I don't see why not?  "command line" commands can be executed like
> this:
>
> request page "123.123.123.123/NADAOPEN FILE
> request page "123.123.123.123/NADAREAD FILE
> request page "123.123.123.123/NADAWRITE FILE
> request page "123.123.123.123/NADARUN FILE
> request page "123.123.123.123/NADABRUN FILE
>
> webserver gets command, recognizes it as being for NadaNet.
> It sends the command over a serial connection to the Master
> machine...
>
> the webpage used to control this internet enabled NadaNet could
> look like a 40 column apple II text screen, with a prompt...  In fact
> the webpage could display the Master machine's text output!

That would be ideal.

In fact, a good prototype project would be simply to hook up an
Apple to the Internet exactly like this, so that it could be used
remotely.  Once you can access a machine (and move files to it),
using NadaNet or a 'Crate is no different.

> > In general, to run a program on an AppleCrate requires setting up
> > the control machine with one or more programs and then being able
> > to run those programs on the control (or "master" machine) and
> > being able to get the output from that machine--basically a remote
> > command line interface.
>
>
>
> Everything is controlled by the master machine, correct?  If a serial
> connection (data, in and out aka I/O) is made to the master machine,
> and the other end is connected to the net, I don't see why you can't
> easily send and receive data..  Just like a keyboard connected to your
> master.

I agree.  Of course, the programs are not usually loaded through the
keyboard (except in some ultimate sense ;-), so some file transfer
protocol would be needed.

I can imagine that Contiki allows files to be transferred to its disk,
so the only issue then is how they get to the master machine's disk.
(I suspect that NadaNet would not be politely compatible with Contiki,
since it disables interrupts for many milliseconds at a time.)

> > > Send the command
> > > to the nadanet master through a serial card.  Program output can
> > > be sent to the webpage through the serial port.
> >
> > Cool.  Does this mean that programs would have to be EXECed and
> > then saved?  General I/O sounds like a problem...
>
>
> That would be one way to do it....
>
> request page "123.123.123.123/NADAOPEN FILE FROM NET
>
> So we have some options on how to get the "remote" code to LOAD.
>
> EXEC like you say...  still need a way to send from net to the
> NadaNet.
>
> 3 ideas:
>
> 1)enter code into a text entry box on the host webpage, have a button
> to "LOAD" AppleCrate with the entered code
>
> 2)specify a file that is on the net with the page request:
>
> request page "123.123.123.123/NADAOPEN FILE FROM NET 222.222.222.222/
> FILE.EXT
>
> 3)choose a file from your hard drive to upload and then use a web
> *browser* to
> recieve the file.  just like the DSK downloader works.  could be a
> BASIC program
> or a Binary file...  whatever
>
>
>
> I don't understand why I/O is a problem...  That's what the Apple II
> is for, right?
> a computer?  Input, process, output.
>
> Code would include a section for I/O.
>
> ---I--- would require connection to net, to get input from remote user
> ---o---would require connection to net, to show output to remote user.
>
> What kind of O is required?  data...  we have the ability to send and
> receive data with Uthernet and that other ethernet card.
>
> I think doing IN#2 on Master with a serial card in slot 2,
> PR#2
>
> with a webpage showing this text would be neat, and we can
> all experiment with it and view eachothers efforts...

My concern is that although some programs can be run on a 'Crate with
just a couple of short Applesoft programs, some will require data files
that may be quite large.

Currently, the NadaNet File Server gives all networked machines access
to the File Server machine's (hard) disk(s).  So my implicit assumption
was that all working data was read from and written to the File Server.

The question, then, is how to make transferring files between a remote
user and the File Server a relatively painless and natural operation
(assuming that the remote command line interface already exists).

I'd envision something like:

1. Move <input directory> to file server
2. Run program(s), with screen responses
3. Move <output directory> to remote user

When I'm using my AppleCrate, I find that I frequently look at the
network activity light and the individual 'Crate machine "send" lights.
The patterns are sometimes quite useful for debugging, as is plugging
a video monitor into various 'Crate machines to see what they're doing.

They are especially useful for observing that a setup is working as
intended.  ;-)

I very frequently keep a monitor plugged into 'Crate machine #2--the
Message Server--since it displays a dynamic histogram of the lengths
of each of the queues it is handling.  This would not be easily
"exported" to a remote user.

I also use the speaker LED (since speakers are not attached) as a kind
of visual indicator of some condition, by (silently) BEEPing a machine
under that condition.

It's also possible to use ATTACH to remotely interact with a 'Crate
machine, but that activity is often guided by what's going on in the
lights.

I suppose it would be possible to interact with a 'Crate without
those cues to its activity, but I haven't really tried using it
as a "black box".

-michael

NadaPong: Network game demo for Apple II computers!
Home page:  http://members.aol.com/MJMahon/

"The wastebasket is our most important design
tool--and it's seriously underused."