Re: HTML as GNAT source

[dewar@merv.cs.nyu.edu (Robert Dewar)]

Robert Leif said

<<<<I agree with everything you wrote except for, "The tools that produce the
  input can be either proprietary, public domain (as would probably be case
  if I do a Y2K tool), or copylefted." One condition I wrote was that Ada
  tools should be available under conditions that permit their use by ALL
  compiler vendors. I will not develop software that is potentially is
  covered by a copyleft. Firstly, I am a professional inventor. I will NEVER
  under any circumstances agree to give up my constitutional right to obtain
  a patent on my intellectual property. However, I will agree to join one or
  more patent and copyright pools including giving up my right to be a sole
  source in exchange for reasonable royalties. >>

>>

One further point here is that it is a misconception that because something
is public domain or copylefted (i.e. copyrighted with the GPL in effect),
that it means that other vendors could not use it.

Certainly if something is public domain, of course it can be used, for
anything, by anyone, in any manner. You can even take a public domain
product, modify it slightly, and copyright the result.

As for GPL'ed software, it can often be quite freely used. Any stand
alone tool is certainly usable by anyone even if used in conjunction
with a proprietary compiler. For example, if some vendor of a proprietary
compiler decided that the new GNAT stub generation tool (gnatstub) would
be useful in conjunction with their compiler, then they would be free
to distribute gnatstub (and the necessary components of GNAT) with
their proprietary compiler. There would be no legal problem in such a
distribution, and we would have no objections at all.

Similarly library units from GNAT, given that they are distributed with
the modified GPL that allows free use in a very wide range of circumstances,
could perfectly well be distributed with some other proprietary compiler,
and indeed at least in some previous versions (don'
t know if this is still true), the Aonix Object Ada compiler included
some of the GNAT library routines in its runtime. Again, that is
perfectly legal, and perfectly fine with us, providing that the 
appropriate distribution rules are met (e.g. if Aonix wanted to modify
one of these units, they would have to distribute the source of the
modified unit, and maintain the result under the same license).

So I understand the concern here, but there are two points.

First, anyone building Ada tools or components is most certainly free
to make the decision to distribute their work under the GPL, and if
Robert Leif decides not to use it on this basis, that is his problem!

Second, in actual fact, Robert is probably over-reacting, most likely
he will be able to use the resulting tools without disturbing his 
"right" to copyright and protect his own work in whatever restrictive
manner he wants to.

Robert Dewar
Ada Core Technologies