Mark Wedel wrote: > The mlab maps obviously do not fall into that case - they are clearly > being actively developed, and the person doing so has CVS. > > One problem that arises from this is that there really isn't any way > for anyone to integrate the mlab maps - your trying to integrate a > moving target. > > The other is that this really just seems to be a way to avoid the > mapguide - I know Mike doesn't want to follow our map directory layout > or some other areas of the mapguide. > > [snip] > > But then I'm still stuck figuring out what is going to happen with > all this mlab stuff - either the correct rules should be followed and > this integrated in properly with the maps, or it probably shouldn't be > there. In terms of integrating a moving target, I don't see how it would be any better if they weren't in CVS. Really, the target is still moving if one was trying to work from Mike's periodic mlab release tarballs, so though it being in CVS doesn't solve the moving target problem, it does allow the movement of the target to be tracked better which IMHO is somewhat helpful to the moving target situation, even if only slightly. In terms of mapguide issues, I agree, though if Mike is too unwilling to change, it may be possible to fix many of the mapguide issues by script (i.e. the directory layout could be autogenerated from the map name prefixes that Mike is using), however chances are this would break from time to time, so I don't see this as too appealing an option unless Mike is willing to adopt such fixes for the development of mlab. I know one of Mike's worries with changing his current layout is that scripts could possibly end up creating small errors and breakage. Question to Mike: What exactly in the mapguide are you opposed to? And why exactly? Alex Schultz