On 7/8/05, Ross Gardler <[EMAIL PROTECTED]> wrote: > Tim Williams wrote: > > I'd prefer project-level locationmaps to be overrides or extensions > > rather than replacements for the app-level locationmap. > > That's the intention, though not well described in the commit message. > > The implementation I put in place is to get things working quickly since > we broke head. I intend to modify the LocationMapModule class to mount > an empty locationmap if it cannot find the project one (i.e. do what the > current workaround does). In the meantime we can start building the > forrest locationmap. > > > Unfortunately, locationmaps don't have > > the same concept of "mounting" so that we could simply mount > > project-level lm's ahead of the app-level one. > > Yes they do - at least in theory, I've not tested it yet. If I > understand it right pretty much any of the map: elements can be used (or > easily be made to work).
It looks to me like it's an implementation that mimics the sitemap syntax but doesn't necessarily make use of it's full power (e.g. TreeProcessor,etc.). Clearly this is at the gray area of my understanding but it looks like the LocationMap implementation implements its own replica of "sitemap" language and not all functionality (e.g. MountNode) is available. I'd really like my reading of the lm stuff to be wrong though. > In other words. I agree with everything you say. I hope others do too ;-) good to hear our goals are the same even if I don't understand the current reality as well as you... --tim
