[thelist] Search engine and index.cgi

[Richard Bennett]

Hi,
First thing is, proceed with care, because the site already has a google
pagerank of 5/10, and many sites linking to it.
If you change the navigation abruptly you'll loose more than you gain.
So whatever you do, you should make sure the current URLs will still be
supported.

It is generaly accepted that Google will spider pages with short
query-strings, with no more than 2 parameters or so.

I think the easiest would be to edit the cgi to accept other forms of
strings too. That way existing links will still work ok, and you could
change the links on the site like this:
Current:
www.afairesoimeme.com/index.cgi?action=srubrique&id=109920&rubrique=2

new:
www.afairesoimeme.com/index.cgi/srubrique/109920/2

Optimised with server-side friendly-name mapping:
www.afairesoimeme.com/index.cgi/srubrique/bougies-traditionel-decoratief-kit
s

Having your key-words in URLs people link to you with makes a big difference
for search-engines.
It might be nice to drop the '.cgi' in the new script, but leave the current
script too for legacy links.

And indeed, all those forms can go too.


Richard.


----- Original Message -----
From: "Damien COLA" <damiencola at wanadoo.fr>
> I have to work on a site that wasn't developped by me. Everything goes
> through index.cgi which then takes a html file and dislays it.
> So the links, when they're not as a <form> are like that :
> index.cgi?page=mypage.html
>
> My issue is about referencing, is it good for robots ?
> I'd guess NO, the robots don't index pages with ?variable=value
> but on google you can find one particular message in a forum thread
> where the url has arguments in it.
> As for the <form> links I am quite sure they are BAD and won't be
> indexed.
>
> So in the end I don't know if having a central point of entry index.cgi
> is bad for search engines.
> And since I am free to change whatever as long as it's not too much
> work, how may I improve that situation ?
>
> Thank you for your insights, the site as it is before I work on it is:
> http://www.afairesoimeme.com
>