paperclips-discuss
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Paperclips-discuss] HEAD, was: API_23_ALPHA_2_1


From: Nic Ferrier
Subject: Re: [Paperclips-discuss] HEAD, was: API_23_ALPHA_2_1
Date: Tue, 26 Jun 2001 11:42:32 +0100

>>> "Christopher K. St. John" <address@hidden> 26-Jun-01
2:50:49 AM >>>

    Hmm, I've never messed with the DD mime type 
    mappings, but it looks like WWW servlet should 
    be doing:

     ServletContext.getMimeType()

    in order to access the mime types mappings in the 
    DD? I'm assuming that DD mappings wouldn't affect 
    URL.getContentType()?

Yes... but it might be just as easy to set them on the PURLH and have
the ServletContext call that. If the PURLH was in charge of mime types
it might make more sense....


   - WWW should use ServletContext.getMimeType()

So where do we get mimetypes from when we're serving non-webapps?


   - WWW has a built-in DD (distributed as a war?)
  - The built-in DD has the appropriate mime-type
    entries, instead of using a separate mime.types
    files.

Ugly. I have thought about overloadable webapps.... but it seems
overkill to me for things like default setup.


   - Don't use URL.getContentType(), since URL is
     a random nasty unpredictable class.

URLs are a lot nicer in Paperclips. I plan to cut out the URL stuff
and make a generalized URL implementation for Classpath.


   - WWW must handle the case where the mime type
     is really unknown, default it to something
     reasonable. Probably text/plain.

Ok.


    Random question:

    - Any reason WWW doesn't return clickable links
      in directory indexes?

It used to but I took it out to try and force welcome-files to be
used.

I've no objection to it going back in as long as it's nice and
simple.


Nic



reply via email to

[Prev in Thread] Current Thread [Next in Thread]