[edit - tim beat me with a concise comment... that will teach me not to be verbose! :-)] Forest, I don't follow that. It seems to me that if gdb has been opened it would be easy for Manifold and anybody else to use it, true? Why would anybody hesitate? If it were really opened that would make it easy for Manifold or any other GIS using the "open" thing to be procured and installed side-by-side with Arc, which is enough for Arc to lose more share (no need to force a switch). If on the other hand it has not been opened, that argues against it for people who wish to avoid being trapped in a "silo." There is also the fundamental concern that gdb is not a particularly efficient or reliable format. Setting aside the fragility of embedding SDE-like structures or shapefiles (both of which have significant problems from a modern perspective) into a DBMS format, if one was to pick a DBMS format MDB is not a good choice because it is really old and fragile. Microsoft's mdb (as Microsoft is the first to say) is a notoriously fragile format, especially for multi-user access or multi-process access as with a GIS-enabled web site. Microsoft has plenty of blood-curdling knowledge base articles about how to attempt to revive a corrupted .mdb, what best practices to use to try to keep a .mdb from being corrupted and so on. That's why many years ago they introduced MSDE as a more robust option and now have introduced SQL Server Express as an alternative. Last but not least, isn't a .gdb limited to only 2 GB because of the 2 GB .mdb limitation? It's difficult to believe individual consultants doing significant work will settle for that, let alone government departments or enterprises. In the year 2010 they usually want more, and expect to need yet more again in 2011 and onward. I don't know. I get the impression ESRI has been trying to push GDB adoption for years and not getting enough headway. With a few customers, yes, but to the degree necessary to replace even shapefiles, no. It could be they've finally gotten some traction with some users, but that would be unfortunate for those users if that traction finally comes at a time when expectations have moved on. 2GB max is way too small, fragility in multiprocess or multiuser ops is no good for workgroups, government departments and enterprises, and the low performance of mdb compared to more modern technologies (heck, even early 2000's MSDE is far faster) is altogether looking back before the year 2000 planning for the past instead of standing in the year 2011 looking ahead planning for the present and the future. The more time goes on the more dated and less attractive gdb looks, so I'm not surprised if they are talking about opening it to try to get some traction before the only people left who know what it is start thinking more about Medicare than GIS. It could be a moot point anyway, since if it ever got real traction Manifold would just open it anyway and if it doesn't get traction it is a don't care. After a zillion formats it's just one more that either will go somewhere or it won't.
|