[pmwiki-users] Google local site search

[Patrick R. Michaud]

On Wed, Dec 28, 2005 at 03:24:27PM -0700, H. Fox wrote:
> On 12/28/05, Patrick R. Michaud <pmichaud at pobox.com> wrote:
> > [...]the approach
> > I'm using isn't modifying <a ...> tags, it's modifying the url that
> > appears in the tag.  Specifically, it's removing any unpermitted
> > "?action=" that appears after "$ScriptUrl".
> 
> So, for example, a page's Edit and History links become self-referring
> links, correct?

Yes.

> If so, I think I'd rather leave the action and include the
> rel='nofollow' attribute.
>
> Bot the Light Skin and Lean Skin include the rel='nofollow'
> attributes, so now there might be unintentional 'nofollow' links to
> normal wiki pages -- possibly resulting in page-rank punishment.  

Google is quite clear that rel='nofollow' doesn't punish a page,
it simply causes the link to not be counted at all for page 
ranking purposes.

And even if the Skins include rel='nofollow' in the templates, 
what about markup...?

    [[OtherPage]]
    [[OtherPage?action=edit]]
    [[OtherPage?action=dc]]

> > > I'd expect most search engines to honor "nofollow" by not following the
> > > link anyway. [...]
> >
> > In reality, for the first six months after rel="nofollow" was
> > introduced, Google followed the links anyway (but didn't weight them).
> > Google didn't stop following the links until sometime in July 2005.
> > So, despite the reasons you give, Google chose to follow the
> > rel="nofollow" links for quite some time after rel="nofollow" was
> > introduced.
> 
> Since future PmWiki versions will not be run prior to July 2005 what's
> important is the fact that they're not following the links now. right?
>  Maybe I'm missing something...

My point was that for at least six months Google thought it was
worth the extra code and expense to follow rel="nofollow" links
but not count the link.  So, if Google chose this path, it's entirely
likely that other search engines will go ahead and follow rel="nofollow"
links.  

In fact, this is *really* the point:  Many spiders/robots won't 
look for rel="nofollow" at all!  The robots exclusion protocol 
doesn't say anything about rel="nofollow" in links, so robots
that are following the protocol are still going to follow such
links.  I've looked at several of the available search engine 
packages (htDig, swish-e, mnoGoSearch, WebGlimpse), and as far
as I can tell *none* of those honor rel="nofollow" in links.

So, the advantage of cloaking "?action=" is that it will work
even for robots that don't understand rel="nofollow".

Pm