Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!swrinde!kent
From: kent@swrinde.nde.swri.edu (Kent D. Polk)
Newsgroups: comp.sys.amiga.programmer
Subject: Re: An Amiga Conversion Facility (2 of 3) LONG!
Message-ID: <2407@swrinde.nde.swri.edu>
Date: 17 Apr 91 17:51:20 GMT
References: <790@tnc.UUCP> <791@tnc.UUCP> <1813@madnix.UUCP>
Sender: news@swrinde.nde.swri.edu
Organization: Southwest Research Institute, San Antonio, Texas
Lines: 83

In article <1813@madnix.UUCP> perry@madnix.UUCP (Perry Kivolowitz) writes:
>In article <791@tnc.UUCP> m0154@tnc.UUCP (GUY GARNETT) writes:
>>           A Conversion Utility for the Amiga (Part 2 of ?)
>>            (A modest proposal from "Wildstar"; 200 lines)
>
>This is all very nice. But it has already been done. Why do it
>again?

Where does it exist? What exists that allows me to programmatically
pass a pointer, etc. to a bitmap in a message & get an encapsulated
Postscript file as a result?  Or a set of DR2D commands & get it
rendered, or get the Postscript output of an IFF bitmap file?  And have
the server automatically loaded and purged when no longer needed?

I need an extensible client/server mechanism like this & I sure haven't
seen it anywhere. Remember the recent "Renderer" thread? I see it and
this "A Conversion Utility for the Amiga" to be the same thing, as they
address the same problems.

If you talking about "The Art Department", I agree that TAD is an
excellent program, but I sure don't see it performing the types of
rendering/conversion functions that is being discussed here.

I brought up this concept about 2 years ago specifically with regard to
rendering 'devices'. My mailbox was flooded with messages explaining to
me that there is absolutely no need for this type of thing on the Amiga
so I shut up and listened. I still disagree, and I still need these
types of operations since they still don't exist.

In the meantime, I developed a pseudo-generic event oriented data
acquisition system which, for once in my 14-year professional career,
allowed me to write an extensible set of programs to perform
acquisition and filter operations ONCE (Ok, maybe twice :^).  No more
writing massively duplicated programs for every blasted test I had to
run!

It's like going shopping now. "Let's see, I need a Tek 2430 sampler
here, two of those plot windows here and there, a gating function
there, a real fft over here, a Hamming window here, a envelope function
there, a time shift based on the gate threshold over yon. Also want to
save the results after that fft and when it's all done. There, off we
go. Hmmm, that's interesting, I wonder what channel 1 would sound like
with a different set of gates than the fft was done on? Ok, another
gate function at the end here and that Play routine. Very
Interesting!!"

Each of the filter operations is (usually) a tiny little program which
only has to handle the specific function for which it was designed. It
was also designed around PPIPC (Fish disks) which made the messaging
and server loading mechanism almost trivial and extremely fast.

------------------------------------

Now, I want to extend this mechanism to allow for a whole list of
things which I need to accomplish. Renders, converters, etc. accessed
from within my small, self-contained processes.  I need "Real" Modular
Programming, not just modular at link time...

What would it look like?  A configuration where you would list your
available servers. This could be a server to do about anything you
wanted on the Amiga. A program needs a service performed on its data? A
simple message... a server gets loaded if not already...  Results.
(BTW, PIPC provides the mechanism for this stuff already - just have to
write the servers and the final config routine).

I guess what I am saying is that I want a different way of computing
and am trying to do something about it. I also think it would be nice
and quite convenient if others decided to do the same, but I have found
that I think a bit unconventionally, so it is usually only after I have
finished needing something that others discover its' significance :^(

BTW Perry, This wasn't personally directed to you. I just don't want to
see this concept thrown away. I also don't think it will necessarily
compete with TAD. I don't work with pictures & don't need a picture
conversion utility, but I do need data rendering & output capabilities.

Why don't we all go take another look at PPIPC and open minds a little?

Thanks,
=====================================================================
Kent Polk - Southwest Research Institute - kent@swrinde.nde.swri.edu
 "What will happen to our memory now that we can keep it on paper?"
=====================================================================