Re: [syndication] Re: Thoughts, questions, and issues.

[zac <zac@pixelgeek.com>]

Aaron Swartz wrote:

 
> I dropped the file in and it gave me this:

<code snipped>

I think the entire idea of namespaces takes a vastly simple and elegant idea
and complicates it for what appears to be no real benefit.

Aaron's sample file makes reference to RDF, RSS091 and, by assumption,
rss1.0 all in the same document. This strikes me as very inelegant.

Perhaps this is based on my own experience in working with systems that have
very little support for anything other than the absolute basics of XML
parsing. Perhaps if I used C and Java I wouldn't look at the RSS1.0 sample
Aaron posted as an unmanageable. The point being, I think, that simplicity
promotes adoption of technologies. I can currently create an RSS091 or RDF
reader in Shockwave in about 15 minutes. If I had to work around the idea of
namespaces then I don't think it would currently be possible in Director (or
REALbasic the other tool I use from XML work) without adding a sizeable
amount of overhead to the code. Which would mean that I would stick with a
simpler format or create my own.

This decreases the  rate of adoption of the format and the technology

Correct me if I am wrong but this argument basically seems to come down one
group that wants to be able to extend the format without breaking anyone
else's apps and another group  thinks that if enough people adopt a format
addition that it should become part of the standard.

I can see the merit in both cases  but I think it comes down to a case of
least harm. I think extensibility via namespaces (this is the basis of the
point is it not?) creates a format that is going to require more work to
parse. Meaning that fewer sites will use it. Which might just engendered the
type of fragmentation that people seem to be concerned about (at lest I
think this is an underlying assumption in some of the posts I have read).

I think the slower path, simple formats followed by adoption and
experimentation, is ultimately the stronger solution.



-- 

Perfection has one grave defect: it is apt to be dull.


Somerset Maugham



email: zac@pixelgeek.com
web: http://www.pixelgeek.com/