Hi guys, this week I'm learning about the importance of posting the
correct header responses. On some pages of our site I use named
anchors, the problem is a click to named anchors trigger a rule to
send a 404 header, since
[jump-to-paragraph-is-right-here] is not a page, but it works fine in
a browser.
If I parse the #named anchor part it defeats the objective of
improving navigation.

I''m hoping the spider has no need to follow a named link, but I
thought I'd ask to gauge the priority of a fixing.

www.example.com/somepage#jump-to-paragraph-is-right-here

Wes

A yes or no reply would be Ok.

Need a real example to test headers.

In general a named anchor even if absent on a page will never cause a
404 reponse. Named anchors are normally only handled by the browser
(so client-side). As long as the url without the anchor exists and
returns a 200, it doesn't matter if the named anchor exists or not. In
fact named anchors are not even included in server variables like
request_uri.

Perhaps you have an Ajax applicaiton which actually  makes use of
named anchors and it interacts with the server  too.

On 2 juil, 11:03, WWW wrote:

WWW-

webado's right.  #anchors are not included in the GET strings sent by
browsers to a server.  Try using a browser or browser extension that
lets you see your HTTP requests (e.g. the "Live HTTP Requests"
extension for Mozilla FireFox), then visit any site with an anchor
link, like this one:

http://www.google.com/webmasters/tools/#AN_ANCHOR

The first two lines of the request will say something like :
GET /webmasters/tools/ HTTP/1.1
Host: www.google.com
...

Notice anything missing from that URL?  That's right, the anchor isn't
in the request on the first line.

If you're using AJAX or some non-browser way to interact with your
server, it is likely not configured to deal with URL string #anchors
by default.  You may want to contact your host or, more likely, switch
to using ?arguments when constructing URLs for client-server data
transfer, but that's not a Crawling, Indexing, or Ranking issue per
se.

The issue may arise that Googlebot, like all visitors using browsers
to your site, will not send #anchors in its request strings.  You
might be changing this behavior with AJAX or Flash, but we typically
count on Googlebot to run complex Javascript like AJAX or--despite the
recent (exciting!) announcement about better flash indexing--
dynamically load Flash content.  Current best practice is to use
#anchors and ?arguments with their historical meaning--#anchors to
pinpoint information client-side, ?arguments to pass information to a
server.

-Bergy

On Jul 2, 8:24 am, webado wrote:

Thanks for your reply webado.
I don't have an ajax system, as you may remember i a few weeks ago you
answered a query with regard to mod rewrites. Bascially i have a cms
system that i wrote to remove url file extensions. The other problem i
encountered was the index file is the default for any page request.
What that meant was the index file has to parse every url, check if
the name exists in the database, redirect to the template file and
print the page. (I would have bought an off the shelf package if i
knew the website would become a full blown application) So the index
takes all via htaccess rewrite, that means i have to force 404s as the
final else statement gives a 200 code and loads the default index
page. Incidentally during development every request gave a 200 (i was
unaware at the time. Now is payback time with 404s for all types of
unruly urls which have been spidered)

HI Bergy

The seo consultants web site i found has a great service for a basic :
http://tinyurl.com/3a9p47
which led me to believe otherwise as it tries the full #link which
caused my concern. (I tried the firefox live headers plugin, but
that's a bit to advanced for beginners.)

I can see from your example a 302 header is sent. That’s a tricky one
for me to solve.

Here's the application / site working
Named anchors here:
http://tinyurl.com/3u3ejf

I can't decide if this a good or bad idea to use here, on the last
site where this content is redirected from every reply was a page in
its own right, but that leads to duplicate threads?

http://tinyurl.com/3pvp77

Thanks Bergy for you time and knowledge !