forrest-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Helena Edelson <>
Subject Re: access control - user roles by sitemap
Date Tue, 20 Dec 2005 20:15:56 GMT
That should work,
are you saying to handle auth in server.xml and parts in web.xml?

thank you.

Gunther Sablon wrote:

> Hi,
> we are running forrest as a webapp in Tomcat.
> We use the authentication mechanisms of Tomcat:
> - before we had users  in tomcat_users.xml; currently they are in LDAP 
> (can be setup in server.xml)
> - web.xml specifies the parts of the site that need authentication.
> There is no connection to the sitemap, so maybe I have misunderstood 
> your question...
> Best regards,
> Gunther
> Helena Edelson wrote:
>> thanks. will do.
>> i am also working on i18n stuff. when i figure it out,
>> i was thinking it might be help ful to write a how to.
>> everything so far seems
>> fractured in terms of the steps to take to implement.
>> there are many if then's and info is everywhere, from
>> docs to jira to cocoon docs..
>> helena
>> Ross Gardler wrote:
>>> Helena Edelson wrote:
>>>> Is there any way using sitemap  to manage user roles
>>>> via a  login to control roles
>>>> access to app pages for user_role_a
>>>> disallow access for user_role_b type of thing?
>>> Depends on whether you are running in dynamic mode or not.
>>> In static mode you would have to use the mechanisms provided by your 
>>> web server for protecting access to certain resources. On the Apache 
>>> HTTPD server that would be using .htaccess files.
>>> In this case you simply provide the relevant .htaccess files in 
>>> along with your XDocs and they will (theoretically) be copied over 
>>> when you build the site. I say theoretically because I have not, 
>>> personally, tried this. Let us know if you try it and it works 
>>> (preferably in the form of a contribution to our docs).
>>> If you are running in dynamic mode you can, theoretically, use the 
>>> Cocoon Authentication framework [1]. Again, I say theoretically 
>>> because it I'm not aware of anyone having done this yet. If you want 
>>> to go this route then it is really a dev topic, so please move this 
>>> to the dev list where we can help guide you and (hopefully) you can 
>>> write us a How-To on making it work.
>>> Ross
>>> [1]

View raw message