> I've been working on my second book on XML which will include a lot more
> practical examples than the first. Right now I'm writing a
> chapter in which
> I describe a dozen or so different XML applications. Eventually, I'll
> devote a chapter each to several of these. But I may not be able to do as
> much as I want or as much as the authors of these specs might like. The
> reason is excessive restrictions on reuse and republication of DTDs and
> theirm associated markup languages.
You might also like to do something similar to my book (The XML & SGML
Cookbook, ISBN 0-13-614223-0 see fourth leaf): I put the notice "The author
does not assert copyright over any SGML markup declaration in this book,
unless noted." It is not enough to expect others to provide copyright-free
declarations, authors have to do it themselves.
The copyright problems were big enough that I kept away from directly
including large chunks of non-ISO or non-W3C DTDs, but that served my
purpose of finding the underlying patterns in DTDs.
There are many tricky cases. For example, ISO claims it has copyright on
country codes: so people can "use" the codes, but only ISO (or the delegated
registartion body) can publish them (this is strange, since many of the
codes came from UN in the first place, so it is dubious). (I got
permission, but there is a widepsread feeling among the authors of standards
that no such copyright exists on codenames. However, presumably this crazy
new US Bill on the phone directory issue may alter that.) But, even if the
codenames themselves may be used by impunity, it is possible that the
particular file which contains codes may be copyright: this is why I got
permission from whoever rekeyed or transcribed some the codes I used (e.g.
the language codes).
One issue that came up was that the ISO entity sets have a copyright notice
which certainly allows the sets to be used in software, but does not
explicitly mention the status for books. Charles Goldfarb (my series editor,
an erstwhile lawyer, inventor of SGML) who I think drafted that notice,
informed me that a book was certainly something "for use with SGML systems"
and so there was no copyright problem. The same approach was taken in in AW
book (by Mikes) on X-Windows, which gives a right to "publish" providing a
copyright notice is given.
One thing you should do to minimize any copyright if you are worried is to
reformat the source code, and replace all comments with your own. Comments
and layout both have "originality", which certainly may be copyrighted. If
you do this, you should also alter the formal public identifier for the DTD,
if it exists:
move the owner field across into the right-hand description, and put your
book's ISBN into the vacated owner field:
-//old owner//DTD foo//EN
becomes
-//new owner//DTD old owner::foo//EN
Rick Jelliffe