[geeklog-devel] Geeklog and 404 Errors

[Tom]

I notice that Google bot was crawling pages of one of my plugins that do not
exist (ie ...&page=2813). I am not sure how Google found them (as the site
didn't create a link to them) but never less Google did crawl a bunch of
these pages.

I checked on a few webmaster forums to figure out what was the right way to
handle this. For example:

1) Return just the last page

2) return a 404

3) do a redirect (301) to the last page

My first thought was option 3 was the best but after a few discussions most
people suggested a 404 error since Google is picky about things like this.

Anyways I thought I would double check on how Geeklog handles an non
existing page number for articles, topic, comments, etc...

A) For non-existing  Articles we redirect back to the homepage.

B) For non-existing Topic page numbers and topic ids we display a No News to
Display Message

C) For non-existing comment pages we just show the item with no comments but
with comment page numbers

Item C needs to be fixed and I would think a 404 should be returned.  I also
wonder if we should do the same for items A and B?

Geeklog also has a 404.php page which as far as I can tell is not used. It
also doesn't seem to return a 404 error code but just contains a nice
message to the user with a few link options.  I am wondering if we should
ditch this page but create a lib-common function that shows something
similar that also includes the correct 404 header so search engines know not
to visit. This way we would have a constant 404 error message across the
site.

Ideas, Thoughts?

Tom

PS This would be implemented in Geeklog 2.0.1 since we are on a feature
freeze for 2.0.0