A month ago we started designing and developing a tool which would be able to merge changes made to a derivative copy of an llblgen pro project (lgp file) into the master lgp file. The complexity of this tool is high, as it had to be useful with vanilla v2.0 projects so existing ApplicationCore functionality had to be used.
The target date for a first beta was set to dec 31st 2006. We defined 3 scenario's:
1) several separated projects on the same catalog should be merged into a single master project
2) one master project was edited in parallel by more than one developer and each changed version would be mergable back into the master
3) a developer would build a new feature or set of features on new tables / changed tables and would merge these new / changed features into the main project after the feature had been completed.
For all 3 we have a design and we're confident that it will work OK. We also have a proper command engine design in place which can deal with commands which rely on other commands' results (think of adding a new entity, renaming a field in that new entity, adding a typeconverter to that new entity's field etc. etc.).
We just have a problem: the set of different types of elements to merge is just a lot and for every element a different command and scenario has to be written which takes a lot of time, and also because existing code had to be used, it is also complex to solve problems as often required code is in an internal method in the applicationcore assembly.
So to cut a long story short, we won't make the deadline Dec 31st. As v2.1 dev has to start before or at that date, we can't continue on this toolkit till it's done (which I think will take at least a month if not more). We therefore can't continue with this project merger tool at the moment and have to postpone it till after v2.1's release.
This also makes things easier: we can add code to the applicationcore assembly to serve the project merger, so writing that will be easier.
As said: we're confident the tool will be possible and will work, we just can't finish it on time at the moment and we have to postpone it at this point. We won't cut it, as we do realize it's an important tool for a lot of you out there, and therefore we will finish it, we just have to do that after v2.1's release.
I appologize for this delay in the release of the project merger, though we can't let it delay v2.1.