[TYPO3-english] CoolURI 1.0.20 inconsistent?

[Axel Joensson]

Jan Bednarik <info at bednarik.org> wrote:

> Hi,
> 
> > thx for your reaction. That behaviour (the starting page of a
> > de/en/ja/zh site being cached as "id=2" without L-param) occurs, when
> > somebody calls the site only by domain name (www.example.de). There is a
> > general redirect active sending them immediately to
> > www.example.com/index.html. I don't think, this may cause severe
> > problems, so that link may stay in the cache, I suppose.
> 
> what do you mean by general redirect? And why would you do that? Why do
> you redirect to index.html?

The site was static before, all files ending with *.htm. That site was
replaced by the T3 site. In the htaccess, there is first the part added
with TYPO3 4.4.1 to gzip css and js-files. Additionally, lots of 301
redirects from old files are defined in advance of the part above, like

Redirect permanent /index.htm http://www.[domain].de/index.html
Redirect permanent /index.htm] http://www.[domain].de/index.html

I added that second one after I found a "crawling error" in the Google
webmaster central. I don't know, who linked to the site that way. All
other 301s redirect old files to new files, nothing else. There is no
redirect for / to index.html in this part. Another .htaccess exists on
another domain, which is no longer used, it redirects a bunch of
formerly used URIs to the present ones in their respective language
folders (en, ja, zh). Finally, there is the "standard" part on bottom:

RewriteEngine On 
RewriteRule
^/(typo3|typo3temp|typo3conf|t3lib|tslib|fileadmin|uploads|showpic\.php)
$ - [L] 
RewriteRule
^/(typo3|typo3temp|typo3conf|t3lib|tslib|fileadmin|uploads|showpic\.php)
/.*$ - [L] 
RewriteCond %{REQUEST_FILENAME} !-f 
RewriteCond %{REQUEST_FILENAME} !-d 
RewriteCond %{REQUEST_FILENAME} !-l 
RewriteRule .* /index.php 
AddType text/x-vcard .vcf

Well, the last line is not standard, doesn't matter. 

I am not even sure why www.[domain].de redirects to
www.[domain].de/index.html, because it is not defined anywhere in the
htaccess, as far as I can see, but I am no expert in htaccess. If you
call the domain without "index.html" appended, you are redirected there.
Anyway: I guess it an advantage to have such an URI, so the same page
will always show up under one address only.

> To be clear
> 
> users come to example.com, they are redirected to 
> example.com/index.html. A new row in link cache saying:
> 
> /     id=2
> 
> appears.
> 
> Or is it somehow different?

Searching with the % operator in the CoolUri Mgr. under section "Cached
links", the first link showing up always comes with an empty field in
the column "Cached URI", no slash there. Under "Parameters" it is shown
as id=2. No sense to delete it, it returns with first time calling
www.[domain].de/. *All* other listed links - without exception - show
"L=[0,1,2,3]&id=[xy]", partly with "&no_cache=1" added manually and
labeled sticky, where necessary. 

Alphabetically way down below comes the Cached URI "index" with
Parameter "L=0&id=2", as expected, and sticky as well as "en/index" and
the indexes of the other two languages are sticky: I regularly had
problems with calling these files through sr_language_menu, they came up
as "../en.html" instead of "../en/index.html", which is stopped by
manually adding the slash and "index" and then marking them sticky.

In the TS template, baseURL and canonical addresses for the head are
defined:

  config.baseURL = http://www.[domain].de/
  ...
  lib.canonical = TEXT
  lib.canonical {
    typolink.parameter.data = TSFE:id
    typolink.returnLast = url
    wrap = <link rel="canonical" href="http://www.[domain].de/|" />
  }
  page.headerData.1000 < lib.canonical

> I think the problem is with the redirect to index.html, how exactly is
> that performed?

I'd like to know ...