Top things to see in first ALFS

Hui Zhou zhouhui at wam.umd.edu
Wed Feb 2 07:04:21 PST 2005


On Wed, Feb 02, 2005 at 02:42:28PM +0000, jamie bennett wrote:
>Quoting Matthias Berndt <Berndt.Matthias at gmx.de>:
>> The client has to parse the profile to give the user a chance of
>> intervention. After that you want so send the profile to the server
>> with some additional information about what and how to be executed and
>> the server has to parse the profile again. IMO sensless duplicated work.
>
>I think your missing the point a bit. The current idea for alfs is that the
>client is only going to read and display the profiles and allow you to
>add/edit/remove bits and pieces. It need not know how the server is actually
>going to implement the function on the target host. For example, a <download>
>tag may be implemented on the server as a wget, a curl, or some other network
>transport method. 
>
>The client should be _dumb_ when it comes to the actual XML translatation. The
>server on the other hand will take the profile and actually do the intelligent
>bit. If each client has to first parse the XML, translate it into some kind of
>language the server will understand and then send it, the server will still have
>to _parse_ the stream at the other end. Why not just keep it all XML?

Jamie, you missed the point of his. The server don't need parse the 
stream since no profile stream is being sent. Try re read his post.

He is voicing another Client/Server model.

-- 
Hui Zhou



More information about the alfs-discuss mailing list