[wp-hackers] Canonical integration into core

[Lynne Pope]

2009/2/18 Alan J Castonguay <alan at verselogic.net>

> Keep in mind that the question mark (?) in the path is not meaningful for
> HTTP. It could just as easily be a capital Z. The semantics are provided by
> the browser's handling of HTML forms and the HTTPD/CGI script parsing them.
> Each different url, regardless of ? or Z, is a different resource, and can
> be referenced and/or cached as a standalone entity. Using query parameters
> (?foo=bar or Zfoo=bar) for things like sorting or paging duplicates the same
> data on multiple pages, for two goals: To permit bookmarking/referencing a
> particular view/sort/subset, or to ask the server to generate a rendering of
> that particular view/sort/subset for a dumb client.
>
> The former (eg, comment paging) can be addressed using Fragment identifiers
> (#comment-page-2) and javascript (to fetch/display/sort that result set).
>
> If there is a need to preserve the latter use case, shouldn't the server
> only provide server-side paging for clients known to require/request it, and
> instead yield a single large page (eg, of comments) for search engines and
> printers, along a javascript/css manipulation for the smart clients?
> -


That makes sense and would also improve accessibility. It would also make
sense to provide the content on the first (post) page and not on the
subsequent pages of commentshould a JavaScript solution not be used.

I still feel strongly that the paged comments issue should be treated
separately to the canonical link. Preventing duplicate content at the core,
without relying on search engines to do it seems a more robust solution to
me.

Matt Cutt's support for this in the core also contains a warning. But, as
for me, I will wait to see how you implement this and reserve my opinion. I
do feel that WP is straying too far into SEO - the nofollow is an example of
this (it is, after all, simply a tag for optimising page rank by reducing
leakage to other sites). And I fail to see how software can determine which
content is the content that is definitive on any given site.

So, I will shut up now :)

Lynne