Re: Preprints/Reprints of Academic Papers in Packages
On Mon, Mar 18, 2002 at 12:15:41AM -0700, John Galt wrote:
> Okay, provide a definition of source that includes interpretive languages
> such as Perl. I submit that any definition of source so broad as to
> include a perlscript must necessarily include a postscript document.
The form of a {program,document} that is intended for modification. This
includes perl scripts (unless they've been run through an obfuscator),
human-editable HTML, and human-editable PDF. It clearly doesn't include
most generated PDF.
I recall Roxen coming with a Tetris module, called GPL (I believe), with
obfuscated source. Gah.
There's also the case where there's no human-editable forms; ie, a document
created in Word, saved as DOC and exported to HTML. Now there's no source at
all.
> >I think that a PDF is source if it's human-editable, and not if it's
> >practically uneditable PDF code generated from something else. The
> >GFDL tries to make this distinction for HTML.
>
> ...and fails miserably IMHO. One thing that must necessarily fall into
> the "not source" category is ASCII-armored encrypted text, yet the GFDL
> allows it as a transparent copy, for an example. GPG is available to the
> general public, it is editable with cat or sed with the proper key if you
> so desire, and the output from gpg is pipeable.
How does it fail miserably? It's obviously extremely poor in the
general case, but it seems pretty close (if imperfect) for HTML:
"Examples of suitable formats for Transparent copies include ...
standard-conforming simple HTML designed for human modification.
Opaque formats include ... the machine-generated HTML produced by some
word processors for output purposes only."
Other than "simple" (complicated, hand-written HTML isn't transparent?)
and "standards-conforming" (if I release HTML that uses the <BLINK> tag,
it's not human-editable?), this seems reasonable.
--
Glenn Maynard
Reply to: