On 2000-06-02 at 17:46 +0100, Colin Watson wrote:
> Mike Bilow <[EMAIL PROTECTED]> wrote:
* * *
> >Further, it is doubtful that invisible (that is,
> >unrendered) parts of the HTML source, such as META tags, could be
> >protected by copyright at all.
>
> Clearly, however, such <META> tags are more important than, say, HTML
> comments would be. For instance, a search engine that encounters the API
> website may classify it using the keywords given, and may even choose to
> display the description ("Debian GNU/Linux is a free distribution of the
> GNU/Linux operating system ...") as a short description of that site. In
> this context, the normally unrendered parts of the HTML source *are*
> significant, though of course they are not the sole problem.
Significant, yes. Protectable by copyright, not necessarily.
An index basis is not itself any kind of intellectual property. Libraries
traditionally sort books by title, author, subject, call number, or some
similar set of indices. Book authors would have no copyright claim if
libraries decided to index their works idiotically, say by cover color.
Libraries and bookstores could actually do fairly extreme things by way of
indexing, such as placing all books about religion under "fiction," and
still not infringe the copyright of their authors.
In any case, I think this is off the point. Assuming that the META tags
are a copyright infringement, the remedy would be to remove them. If the
API web site removed the "swirl" graphic and the META tags, would those
changes alone satisfy Debian? I think not.
The critical question, then, is: what remedy should Debian seek here?
-- Mike
