On Tue, Aug 20, 2013 at 5:06 PM, Rob Weir <robw...@apache.org> wrote:
> The trick is to work with the Apache CMS. We already have a system > where source files (like MDText and HTML) are converted into HTML > which we then review and publish. That's how our templating works. > The glue is Perl. And I bet Python is used behind the scenes for the > MDText conversion. > > This is not interactive on a per-request basis. But if we can think > of this as a data-driven problem where we have an XML (or CSV or some > other structured data file) that we want to convert into some web > content like JSON or HTML, and have the web content automatically > regenerated whenever the underlying data file changes, then we might > consider the CMS as part of the solution. > > This could work well for our update notification scripts as well. > These should also probably be data drive. And I bet they would use > some of the same data that is needed for the download pages. > > Wouldn't it be cool to update an XML file describing our release files > and then have it automatically generate the download page logic as > well as the update notification XML's? > It'd probably be good to have the info in generic XML anyway. That could be AJAX-ed by JS client-side or used server-side to dynamically generate the page. Don
