🚀 go-pugleaf

/ gnu.misc.discuss / butcher@g.gp.cs.cmu.edu (Lawrence Butcher) / Aug 1, 1990 /
> Look-and-Feel Copyrights, and Copyrights on anything besides Source Code:
> 
> Utter disaster.  Catastrophe.

Yeah, but...

> Anyone who invents a new computer language and writes a compiler for that
> language, can prevent anyone else from writing a compiler for that language
> or any directly derived language [like this:]
> The compiler contains within it a list of keywords.

Yeah (unless it's for APL:-), but...

> The
> keywords are the single most important piece of the copyrighted compiler.

Oh, really?  Great, time to go shorten all those compiler writing courses!
The ones I know of tend to stress insignificant little things, like
tokenizers, parsers, parse trees, symbol tables, optimizers, code
generators... 

> Any compiler which compiles the same language must contain the same list
> of keywords.  Clearly this would be an intentional copyright violation.

On the list of keywords, if that can be copyrighted.  Not on the compiler.
At least, that's the common sense interpretation.  A brain-damaged
interpretation that allows people to copyright user interfaces would make
your friend right.  That's one of the things that the Leage for Programming
Freedom is fighting against (have you joined?).

> There are many ways to express the same ideas that the reserved word list
> expresses for a new language.  For instance the author of a clone compiler
> might use BEGIN and END instead of { }.

If a language specifies "BEGIN, END", then a compiler that uses "{, }"
instead compiles a different language.

> The use of identical keywords can
> only be the result of an intentional effort to copy the keyword list in
> the original compiler.

Contrary to whidespread opinion, any idiot can make up a language.  It
takes a lot of skill to design a GOOD language.  There is hardly a shortage
of languages out there (thousands? tens of thousands?).  There are many
questionable reasons to invent yet another language, and "so that it uses a
set of keywords disjoint from those of all the other languages in the
world" is the most ridiculous reason yet.

> There is no written law which says that a person is allowed to write a
> cross-compiler for Lotus Macros.

In this country, everything is legal unless it's expressly forbidden (at
least that's how it's supposed to be).  Is there a law that says that a
person is NOT allowed to write a cross-compiler for Lotus Macros?

> There is no written law which says that a
> programmer can take the keywords in the C language and add a couple, and
> distribute the new compiler as a separate work.

Oh yes there is.  The C language is in the public domain.  I don't think
languages are copyrightable at all, though certain *&%&@*$'s will attack
small companies with law suits which claim that they are.

> It is derived.

The language is.  The compiler may or may not be.

> Given present laws, I see no reason to be optimistic
> about my freedom to write software which is plug-compatible with existing
> software.

I think the present laws are being abused.  I'm all in favor of clearer
laws that would prevent such abuse.  But I do not accept the premise that
the present laws are designed the way their abusers claim they are.

> [The copyright office] accept the first and
> last several pages of the source code as complete specification of the
> expressions embodied in the code.  No matter that the pages be blank, or
> contain comments.  When the copyright is up, the public gets nothing.

This is really orthogonal to the interface issue.  I don't have anything
against copyrights on SOFTWARE (as opposed to copyrighting the way it
behaves).  Yes, it would be more fair if the source code to the whole
program was held in trust until it goes into the public domain.  But
frankly, 50 years after the author's death it is not likely to be that
desirable.

Jacob
--
Jacob Gore		Jacob@Gore.Com			boulder!gore!jacob