[Tfug] RDBMS reprise

[Bexley Hall]

Hi, Judd,

--- Judd Pickell <pickell at gmail.com> wrote:

> Well I guess there is no system is not frought with
> peril.. :)
> 
> Okay, how about non db solution? Why not have the
> dbms save an xml
> file to disk with the query result, and use a script
> to parse return

There's still the same amount of entropy involved.
It boils down to whether you store the results in
the client, on the server, or take it in bite-sized
pieces.  And, how vulnerable you are to changes
in thedataset *while* you are doing the query.

I.e. there is no free lunch.  It's just deciding
who to pay!  :-(

> results to the client. Either via AJAX, or via
> simple call the page
> load the results type of thing. Of course there is
> no tried and true
> process for making sure created files are removed,
> but I guess if you
> timestamp it with the sessionid, you could easily
> discover which ones
> should be removed. This would keep both server
> resources and client resources low.

--don


      ____________________________________________________________________________________
Looking for last minute shopping deals?  
Find them fast with Yahoo! Search.  http://tools.search.yahoo.com/newsearch/category.php?category=shopping