[RFC] Search engine optimization: canonical link elements

[Noxwizard]

To help search engines with duplicate detection, I think it would be a good idea to add a canonical link element to the page header. It is currently possible to reach the same pages under many different URLs, so we are not providing singular addresses to search engines:

• index.php
  • SID
  viewforum.php?f=#
  • SID
  • start offset
  • sort key
  • sort direction
  • sort days
  viewtopic.php
  • ?f=#&t=#
  • ?p=#
  • SID
  • start offset
  • sort key
  • sort direction
  • sort days

The memberlist also has sort parameters, but I think at the least we should have those three pages set up to provide a canonical address. This would also aid sites that use "pretty URLs." The canonical forms should be:

• index.php
  viewforum.php?f=#(&start=#)
  viewtopic.php?f=#&t=#(&start=#)

Canonical links are supported by:
Ask.com - Article link is dead
Bing
Google
Yahoo

[MichaelC]

+1. Sounds like a good idea to me.

[Oleg]

Search engines should not be getting urls with sids, but users may use such urls and as a result search engines may index them, therefore I agree with the sid part.

Start offset is needed to index pages > 1, it has to stay.

Sort key, direction and days should only be selectable via a form which search engines do not use. I don't recall ever seeing an indexed viewforum view not in default order. Theoretically we can handle the sort bits but I doubt it's worth the effort.

Viewtopic is interesting as the usual view is the topic view, not a single post view. From my experience with search engine indexed forum posts, sometimes reading just the one post is enough and sometimes I want to see the entire topic. I doubt we can select one or the other for all posts on all boards.

[naderman]

+1 on adding canonnical urls in the header where applicable. I agree with Oleg that we should be careful what to consider duplicate content URLs.

[bantu]

+1 on adding canonnical urls in the header where applicable. I agree with Oleg that we should be careful what to consider duplicate content URLs.

+1

[Louis7777]

Was this developed at any point or has it been forgotten?

[bantu]

https://tracker.phpbb.com/browse/PHPBB3-12555
https://github.com/phpbb/phpbb/pull/2468 has been merged.

[Pete77s]

Perfect! 3.1 has begun to change my thinking on the need for seo prettyfications when I upgrade.

Great job, guys

[rfdy]

Why wouldn't the view=print be included in the canonical? It seems like obvious duplicated content. If it isn't a canonical it should at least be include a robots "nofollow" meta.

Related: is there not a default robots.txt file?

[Vinny]

Why wouldn't the view=print be included in the canonical? It seems like obvious duplicated content. If it isn't a canonical it should at least be include a robots "nofollow" meta.

Related: is there not a default robots.txt file?

https://github.com/phpbb/phpbb/blob/dev ... nt.html#L5