> The problem that I see here is collecting all pages into a .application file. > Personally I do not think this belongs in Tapestry (the runtime costs would be huge), but in a pre-processing tool such as Ant or Spindle. Coding an Ant task that finds all .page files in the project and adds them to the .application should not be hard to do at all, and I think it would resolve your issue. > In a sense, this is similar to what 'make depend' or the Visual Studio IDE do for C/C++. > What do you think? > -mb
I'm not so hot on the idea of a post-process modifying my application file, and I'm also not so keen on having to integrate that process into our build process (which has so far been trivial). But perhaps with the use of search-paths, or something similar specified in an application/library the runtime expense of searching for a page would not be so bad? I don't know enough about servlet containers or even resource loading to know if this is even possible. Howard mentioned putting HTML in the root of the web application context - is that somehow searchable where a deployed jar (containing the page, its template and its java code) is not? R ------------------------------------------------------- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf _______________________________________________ Tapestry-developer mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/tapestry-developer
