In-Reply-To: <9505141612.AA12299@fitt-85.fir.fbc.com>
Message-ID: 
References: <9505141612.AA12299@fitt-85.fir.fbc.com>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Status: O
X-Status: 


On Sun, 14 May 1995, Mike Fischbein wrote:

> In article  you wrote:
> 
> 
> : On 14 Apr 1995, DrA wrote:
> 
> : There is one BIG difference between PDF and HTML/SGML... the availability 
> : of specifications in public archives, and the availability of reference 
> : model source code under the terms of General Public license.

Before I take yet another drubbing for the specific terms of GPL vs. the 
15 various flavors of "Freely Redistributable Source Code, with Required 
Give-backs of enhancments" Licenses, I will quickly yield that there are 
specific distinctions that are unique to the FSF-GPL.  When I spoke of 
GPL I was referring to it in a more generic sense.

> : The internet exists in it's current form, not because of the technical 
> : superiority, but because of it's evolutionary development under GPL.
> 
> Nonsense.  Most of the evolutionary development of the
> Internet is despite the GPL, not because of it.

> Stallman's crusade is his own business, those of us
> who make code freely available without the stringent
> restrictions of the GPL resent your insult.

I apologize for what was perceived as an insult.  Again,
I will apologize for lumping all "Freely Redistributable Source
Code - with restrictions on proprietary enhancements" in
the same boat with Stallman's Specific wording of GPL.

I know there were several who were very upset when Stallman put GPL
copyright notices into the source code generated by BISON because
bison generated macros were part of his code.

>  I can
> only speak for myself directly, of course; but I'll
> bet that code like the X Window System, the Berkeley
> Software Distribution, all the Internet specifications
> (RFCs, IENs, etc), including IP itself have done more
> than all the GPLed code around.

Absolutely.  The key is it was the contributions to net.sources and the 
related archives and newsgroups that has been the driving force behind
these standards.

> I'd much rather deal with code "Copyright the Regents
> of the University of California", or code released into
> the public domain through COSMIC, than with GNU's stuff.

I am still a Linux fan, but I agree that the Licensing for UC and UI
are better.  The concern with public domain is that you end up with 
proprietary "killer upgrades" which render the generic clients useless as 
soon as they hit upgraded servers and content.  A perfect example is
the chaos Netscape causes Mosiac and fire-wall proxies, or the delightful
experience of having a PDF file "dumped all over your console" because
the publisher never tested it with anything but Netscape.

> : For those who are not familiar with it, GPL permits you to make/sell 
> : binary copies, but you must provide access to original source code and
> : may not make proprietary enhancements without returning them to the 
> : original authors.
> It also has the effect of prohibiting reasonably
> priced technical support.

It does depend on who is doing it.  Linux seems to get excellent 
technical support because Linus is very consistant about incorporating 
updates on a regular basis.  On the other hand, I've seen some apps die
of neglect.

> : The most effective strategy, given these economics, seems to be to 
> : publish infrastructure under GPL, and publish "Glitz" under private 
> : labels.  Use X11 and widgets as the infrastructure and use Motif as
> : the "Glitz", reference models can be published using athena widgets.

> Why do you contradict yourself?  You recommend GPL, then
> X - yet X is freely distributed without the GPL restrictions.
> Then you even recommend a completely proprietary, for-sale
> chunk of code, Motif.

I objected to Motif when it came out as well.  On the other hand, Motif
is relatively inexpensive, available in source for reasonable fees, and
doesn't turn an X-server into confetti or wipe-out half the controls.
(Yes, it does kill push-pins :-)  At least ICCCM provided a structure for 
interoperability.

> : PDF may be a better solution for some purposes, but without available 
> : reference model source code, there is a high risk that some "loser" like 
> : JPEG, compressed ghostsscript, or list3820 could end up as the ultimate 
> : "winner".
> So, where's this reference model source for Motif?
Many applications are developed in reference model using Athena or HP 
Widgets.  On the other hand, OpenLook and the Andrew libraries are
"Pretty" and are widely available under "Freely Redistributable" licenses.

In principle, we are saying the same thing.  The availability of source 
code under a "freely redistributable/publish upgrades" license has lead 
to a powerful infrastructure.

The question is:
	Do we now wish to see proprietary upgrades and "killer fixes" making 
the entire internet as reliable and compatible as  
MS-DOS/Windows?  This, I feel, would be the consequences of allowing PDF
to become a "DeFacto Standard" through pure Marketing, without the
security and stability of the "freely distributable source".

Or-- Maybe we should all switch to WindowsNT and let Billy do
Everything :-).

> 	mike


> Mike Fischbein   mfischbe@fir.fbc.com   CS First Boston

> Any opinions expressed are mine only, and not necessarily
> those of any other entity.  They may not even be mine.
> 

	Rex Ballard
	Standard & Poor's/McGraw-Hill
	Opinions expressed do not necessarily reflect
	the Management of the McGraw-Hill Companies.
	And subject to change without notice.



From rballard@cnj.digex.net Tue May 16 02:53:05 1995