Matt,
The important distinction is between syndication of content and
unauthorized distribution.

In general, copyright law (at least as interpreted by courts in the US)
prevents any unauthorized copying of protected content that is not required
in the normal act of reading or consuming that content. In other words,
facilitative copying is permitted as it is seen to be part of the mechanics
of reading. However, even while such copying is permitted, it is only
permitted for the express purpose of facilitating the reading of the
content and other things that can be done with copied content do not become
permitted simply because one of the reasons for copying was to facilitate
reading. Thus, just because you can copy content into temporary caches,
screen buffers, etc. doesn't mean that you are allowed to then republish
that content in a modified form (unless necessary to permit reading), as
part of a collection, or whatever.

Syndication via feeds, PSHB, etc. should be seen as providing nothing
different from what is provided by TCP/IP, web protocols, etc. These are
simply delivery mechanisms that move content from a publisher to a reader.
The fact that copying is involved in these transfers is no more interesting
than the copying that occurs for any TCP/IP packet or the copying that
occurs in a web browser. Thus, just as we say that a publisher who puts
content on the web in the form of an HTML page implicitly licenses
facilitative copying by browsers, we can say that publishers who "publish"
content within a syndication network (either by creating feeds or by
pushing fat-pings to hubs) is also implicitly licensing that content to
flow through that network. The key thing to understand, of course, is that
the mere fact that some copying was done to facilitate syndication and
subsequent reading does not in any way weaken copyright protection for any
other purpose. The copies are just as protected as the originals.

One might argue that the implicit license to syndicate doesn't exist in the
case where a scraper is used to convert content such as a web page into a
syndication format without the permission of the publisher. However, while
this case might generate some debate, we shouldn't question that publishers
who knowingly convert, or permit the conversion of, their content to
syndication formats have provided a limited implicit license to syndicate.

If the rule I suggest above is *not* accepted, then what you'll find is
that none of the syndication systems in use today can be safely operated
without fear of legal problems. The problem is that if format conversion or
injection isn't the signal for the implicit license to syndicate, then
*any* publisher could create feeds and inject content and then object when
the syndication network works as designed and intended. This is known as
"poisoning the stream..." (There have been cases of people who created
feeds that contained "licenses" in them and then insisted that they were
being damaged by all feed syndicators that didn't pay attention to those
licenses. This sort of entrapment by poisoning the stream clearly should
not be permitted.)

Many have suggested that greater control over distribution can and should
be given to publishers via some form of machine readable license to
distribute content. However, whether or not such machine readable licenses
would be useful (I think they would create a mess), it is important to
realize that no non-legislative authority, such as a standards group, can
usefully define the format of such a machine readable license. The reason
is that only legislative bodies (i.e. governments) can define the means by
which one party is encumbered with legal responsibility to a second party.
A non-legislative body might define a format, however, that wouldn't given
any a legal requirement to pay attention to stuff encoded in that license
-- unless there were some explicit and private contractual relationship
between the parties.

You may argue that Creative Commons is an example of a machine readable
license and if CC licenses are possible, then others should be as well.
However, it is important to note that Creative Commons licenses do not, in
any circumstance, impose greater restrictions on use than what is provided
by copyright. Creative Commons licenses do not restrict usage, rather they
only grant rights that would otherwise be prohibited by copyright. (Note:
The CC "non-commercial" license doesn't actually "prohibit" commercial use,
rather it simply says that commercial use rights are not granted and are
thus are restricted to whatever copyright law would provide or not provide.)

Sorry for going on so long. I've just heard this concern before and am
trying to anticipate a number of directions that this conversation normally
goes in... Basically, I don't think we have a problem as long as we're only
talking about syndication and as long as the content syndicated comes
either in a format which is known to be a "syndication format" (i.e.
RSS/Atom, etc.) or if the publisher injects the content into the network.
In both cases, there is an implicit, limited license to syndicate.

bob wyman



On Sun, Nov 20, 2011 at 5:02 PM, Julien <[email protected]> wrote:

> This is a message sent my Matt Terenzio posted in another topic. As I
> believe this is a topic worth discussing, I'm reposting it here :
>
>    I brought an issue up in the early days and got a few decent
> responses and
>    a number of irrelevant attacks which I guess was because I was
> considered
>    the RSSCloud guy on the PuSH list. Just thought I'd toss that in
> here.  ; )
>    But it had to do with the architecture of PubSubHubbub and
> respecting
>    copyright.
>    At some point in a a widely grey area there is a line between
> syndication
>    and unauthorized redistribution of content. I don't know where it
> is and it
>    might even begin with the publishers intention or implicit license
> they
>    give by making a feed available.
>    While I tend to lean toward more open licenses for content, not
> everyone
>    does. And because hubs can daisy chain content down lines, whether
> or not
>    your hub is respectful might not mean you aren't part of a
> questionable
>    distribution chain.
>    That last part is certainly not the strong part of what I'm
> saying. Just
>    saying we should think about what it means to redistribute parts
> of the web
>    that owners might not have intended for syndication.
>    Aside from that concern which I'm sure you have already thunk
> about, I
>    think it has incredible potential with the explosion of semantic
> web data
>    arriving on the web.
>    So much so that I could see feeds being unnecessary for many sites
> since
>    all the pages are marked up well enough that the description of
> the content
>    is just as easily digestible from the web page as it was from the
> feeds.
>    Almost, at least, though there would still be the overhead of the
> crawl, I
>    guess. But for many blog style sites, a sematically marked up home
> page is
>    practically as good as a feed.
>

Reply via email to