AxKit - Site List
These are huge (24 and 28 print volumes) resources, and we used Axkit to transform the xml using mostly xpathscript stylesheets. The resources themselves are the most authoritative reference works of their kind, and are flagship products for their publisher, Macmillan press.
Although our site (http://www.howlingfrog.com) is fairly simple, AxKit helped make it a WHOLE lot easier to come up with and keep consistent. We're primarily using simple XPathScript to do the transformations from XML to HTML, and aren't using any of the super-cool features that AxKit gives us (e.g. different output styles from a single content-base), but have been extremely happy with AxKit.
We knew we wanted something that'd let us keep our site content in XML as we've worked with too many customers in the past that ended up with nothing but nightmares because their HTML designer did everything by hand and didn't use any sort of management tool/style.
The one "nice" thing that we've had AxKit do for us, is cascaded XPathScript stylesheets. Generally, we build up most of the content for our site in a custom "page" XML schema that we use internally. For other types of pages though (e.g. "software"), we've got a stylesheet set up that'll take the "software" page and then convert that to the standard "page" XML first, then crank that through the end-result "page->HTML" stylesheet, and we're done. This alone saved us tons of grief that'd otherwise be spent trying to maintain several separate styles that all "look sorta the same". By the time we were done setting up all of the styles, we had it to the point that if we wanted to update the "look+feel" of the site we were down to a single stylesheet that needed to be updated; all of the others just rippled-through and picked up that style on their way out.
The main web site of Oxford University Computing Services is done in AxKit.
All about satellite TV and Radio with up-to-date schedule for more than 200 satellite TV channels.
The main pipeline of the site is the following: Postgres database >(data)> Perl scripts >(XML)> AxKit >(HTML)> Browser
Under AxKit transformation pipeline looks like this: Apache::RegistryFilter >>> Apache::AxKit::Provider::Filter >(XML)> AxKit first stage XSLT >(XML)> AxKit second stage XSLT >(HTML)>
The key is two stage processing by AxKit. First stage XSLT is chosen on per-page basis. The second stage XSLT is the same for all pages. It separates browser/internationalization support logic from interface sprecifics. The browser type and language requested is chosen by Perl script based on HTTP_USER_AGENT setting for browser and URI for language. So at the second stage XSLT adds (or removes) attributes from HTML tags based on browser selected. And it removes all text, which is in languages others than selected.
I can't say (yet!) our site is successful. But I'am sure it is very easy to change its look-n-feel or to add languages. Thanks Matt and all the people who contributed to AxKit.
We (upmystreet.com) use AxKit to drive our Sky Digital WapTV service. We use our own XSP taglibs built using TagLibHelper to interface into our pre existing perl libraries. The XML is transformed using libxslt.
We use a multiple stage pipeline to simplify the WML creation at various stages. We don't use the internal AxKit cache system, due to us providing customised pages based on postcodes (that's 1.7 million variations of each page). We do however have our own SQL caching layer.
Performance has been excellent, and I've always had a high level of support from the various AxKit support channels.
These pages are copyright © 2002 Apache Software Foundation