Re: Improving gweather

On Mon, Mar 10, 2003 at 07:51:24AM +1300, Mark McLoughlin wrote: 
> 	Of course, its not just the hard coding of a URL thats a problem.
> Previously, AFAIR, gweather parsed an arbitrarily formatted HTML page
> and now it parses XML. A change in this format will break the applet
> just as much as the URL changing ... I think there's a few options for
> fixing that:
> 	* Mirror the service on in our own "standard" format.
> 	or
> 	* Allow the applet to pull from both the location of the
> service and enough information to parse the output. Maybe a list of
> XPath paths into the output for each piece of data we use might work,
> maybe we could supply an XSL transform to apply to the ouput and parse
> the result, maybe ... I'll leave that discussion up to people who know
> what they're talking about.

I'd say simply guaranteeing the format (via mirroring ourselves if
required) seems the simplest by far.

One thing we might consider is gzipping the data, don't know if that
would help with bandwidth or if http has built-in compression ability
these days.


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