Re: Commentary: RSS Roundtable

[Aaron Swartz <aswartz@swartzfam.com>]

Aaron Swartz <aswartz@swartzfam.com> wrote:

> Rael hasn't mentioned it yet, but this week's Open Source Roundtable at the
> O'Reilly Network is all about RSS:
>
> http://www.oreillynet.com/linux/rt/08252000/

Dave Winer comments at:

http://scriptingnews.userland.com/stories/storyReader$916

Here are the comments in full, with my reply. I don't want to start another
fight, but I think this would be a good opportunity to respond to some of
the misconceptions about RSS.

> These are my comments entered after listening to O'Reilly's radio show on "RSS
> 1.0".
> 
> Had I been on their program this is what I would have said, at a mininum. The
> format and process they describe are highly complex.

I disagree, they are certainly more complex but I don't think they are
"highly complex" and even if they are, complexity is different from
complication. And I don't understand what you mean by process.

> They are over-estimating
> content people's technical sophistication and interest in working on new
> formats.

Once again, content people can use the tools that we're creating to convert
from simpler formats and write files through a Web interface.

> People can extend RSS without namespaces, they already have and this
> has not caused problems afaik. Namespaces will create new problems.

Actually, I believe it has caused problems. The extensions to RSS have not
been well documented, are not widely used and are not agreed upon. Dave,
please explain the problems you see with namespaces, so we can try and fix
them. The only problem I see is that they cause problems for people whose
parsers do not support them, but parsers are being rapidly written and
updated so this should be remedied quickly.

> The basic item structure of RSS remains the same. This is the core problem
> with RSS, it does not model Web content in Y2K, it's not clear that it even
> modeled Web content in 1998. Clearly they want it to do a lot more than it was
> designed to do, and they aren't dealing with the tough issues.

RDF is not meant to model "Web content", it's meant to model site structure.
Please tell us what you think the problems RDF's model are, we would like to
fix them. Yes, we want it to do more than it was designed to do, but it was
very minimally designed (a list of links and titles attached to a URL)
allowing many possibilities for expansion.

> Imho, the new format should not be called RSS. There's been a fork, and the
> peaceful solution is to each go our own way. Calling their spec RSS is unfair.
> We never considered moving RSS forward without getting O'Reilly on board
> first. RSS 1.0 was a surprise, we found out when the spec went public. I've
> said this over and over to the O'Reilly people, I would wish them godspeed if
> they hadn't called it RSS. Should we call our spec RSS 1.0 too?

Hmm, perhaps we should consider changing the name. The problem is that many
of us have so much invested in the current name, making it painful to change
it. Having two RSS 1.0's would be even more confusing. I think the name is
also deserved, considering the large amount of work spent on making the new
spec backwards-compatible with RSS 0.9. It would be different if their we
were creating a radically new spec, but we're not -- instead we're simply
adding namespaces and more RDF support to an already existing spec.

> BTW, it was Netscape's decision to take the RDF out of RSS, one we heartily
> supported. We considered calling it Really Simple Syndication. That's the core
> thing about RSS, simplicity, it's almost an end-user format, easily explained
> in a four-screen spec designed for people who understand HTML and not much
> more. Once Guha left, Netscape totally dropped the RDF pretense. Now it's
> back. 

We disagree on the importance of simplicity. Yes, I like simplicity, but it
needs to be balanced. I don't think that's the core thing about RSS, I think
the core thing about RSS is what it stands for: RDF, sites and summaries.

Actually, I don't believe it was all Guha. Dan Libby was the original author
of the format, and has recently commented that further RDF development was
the intention from the beginning:

http://www.egroups.com/message/rss-dev/239

> Imho it's worthwhile to come to agreement on new tags to keep it simple for
> developers, designers and writers. But we clearly have a different technical
> philosophy from O'Reilly, and a different ethic for working with people. We
> have good ideas. None of them are in their spec. That was their choice. Should
> we stop moving? No way.

No new ideas are in the spec -- we want people to share them and add them
through namespaces. You are free to do this if you want, but you are also
free to create your own format.

Dave, I understand you are unhappy with the spec, and I apologize for that.
However, please don't let your unhappiness interfere with the work that is
and can be done. You are free to create your own spec if you wish, no one is
stopping you -- this is the Web.

Let's try and be civil here. We have our differences, and we'll each go our
different ways. I respect your difference of opinion, and let's keep it at
that. 

-- 
        Aaron Swartz         |"This information is top security.
<http://swartzfam.com/aaron/>|     When you have read it, destroy yourself."
  <http://www.theinfo.org/>  |             - Marshall McLuhan