Hugi, By the size of the DB, do you think that the new vehicle registration DB would be huge or poking back at the legacy DB is difficult because it's so huge and doesn't like to be poked?
Still, it might be worth doing. I am always amazed when I talk to large enterprise developers about how awful and full of legacy code big bureaucratic systems are, just last month a couple of Cobol programmers talked to me about some "still in existence" systems. It would be a good conceptual design which could generalize a way of moving these kind of legacy systems forward. Cheers, Baiss On Jul 29, 2010, at 2:52 AM, Hugi Thordarson wrote: >>>> User. Modifiable. PK. WTF?!!? Icelandic banking systems? :-P >>>> >>> >>> I wish - unfortunately it's the Icelandic Vehicle Registry, meaning *I* >>> have to fix this :). >> >> You need to expand your definition of "fix" >> A re-architecting from the DB up is also a fix in my book. > > That would indeed be my first choice, but unfortunately there are tons of > legacy systems using the DB, and it will take 3-5 years to phase them all > out. This means the WO system will have to coexist with them for a while. > > One option I considered was to design an entirely new DB for the WO systems > to talk to and keep it in sync with the old DB using triggers. But due to the > size of the DB, this would be a huge undertaking. > >> The good news is that you can use booleans for the PKs/FKs since that should >> cover both cars. > > ROFL! > > - hugi ---- Baiss Eric Magnusson CascadeWebDesign.com _______________________________________________ Do not post admin requests to the list. They will be ignored. Webobjects-dev mailing list ([email protected]) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com This email sent to [email protected]
