Re: Homebrew STG card - CRAZY-7

["Arne Chr. Jorgensen" <instel-at-online.no>]

Hi,

It is hard to judge what you think,  from the few comments that has been on
this.  Some tends to look upon it as silly.   From the way I presented it,  I
guess I have to expect this.  

But what do we want ?   Is this EMC software meant to dictate us,  or educate
and help us ?   Has it been done in the hope that "we" should be able to
participate ?

It has been stated several times that some of the intended functions would be
to run parts over a network,  and coordinate several tasks.  In such a context
I would say that we have a test vehicle for such attempts.   This is not much
different than to have a GUI on one machine,  networked to another running EMC.

The other thing is that there has been proposed to use PC104 and other
candidates for EMC.  So how would you do this ?  For any such developments,  I
do think this could be a "test vehicle",  as one of the interesting parts
would be to maybe download parts of the software.  

This has been some of my thought behind this,  but focusing on something we all
can get our hand on cheap,  and have an opportunity to learn the bits and
pieces needed.  Or should we just "cry" out for support by NIST on such tasks ?

In this case,  I do.  There is too much I just don't have been able to
understand,  and  I hoped for some sort of project we could work with and learn
from,  without making a havoc out of everything.  

At least for me,  it is hard to understand a lot of what NIST have been doing. 
They have created several libraries with some thought,  - else they would not
have done it.  But there is a problem to read out of the code in EMC, how they
intended it to be used and how it can be used.  

If we want to be able to use this, I think we need to work on some examples
from the start.   So I didn't intend this idea to be something very smart, but
more a practical example and test project. 

Sometimes that is not so easy.  Why I "dragged"  out some older computers was
not only because most of us can get our hand on it,  but these are at least not
as complicated as the last ones on the marked,  and it is easier to get hold of
specifications.  

We may be able to make simple ISA hardware,  but for anything else, we need to
understand both hardware and software to interface such things.  

My hope was to be able to find some "starting" point,  even if it would never
be an up to date and high performance "thing".  But there is not much that
would hinder it to be working with a high grade of performance - if we worked
along such lines.  This last part would of course make it so much more fun to
do.

Sure,  I could get something working - typing 250.000 lines of code,
reinventing the wheel - but that is not the point.   I hoped to get some help
along the way - as how to use the libraries and code that NIST have made. And
not alone,  but together with other that would find such a thing interesting.

I am not sure if that is being "silly" .    Well,  whatever,  this was the
idea behind it.

//ARNE