Recently I have been frequently asked to write up my criticisms about JDO again. The task was already very close to the top of my //TODO: list. Once again a bug simply disappeared with a nice refactoring: JDO was voted out of existence.I was only halfway correct with my forecast more than 3 years ago:"Why Sun's Java Data Objects (JDO) will not be successful"Bad politics have been even more lethal than bad design.Accordingy I will keep my last summary of the technical flaws of JDO short:(1) JDOQL is awful. Strings must be omitted from a query API at all costs to allow compile-time checks and refactorings. I can't understand why "OO experts" tried to embed yet another string language instead of using the natural choice of langugages to create queries, the language which is already there: Java.(2) A standard for object persistence must define the full lifecycle of declarative persistence first, before moving on to transparent persistence.(3) Transparent persistence is very difficult to achieve in multi ...

Here is some enabling technology, that helps to make our company, db4objects Inc. work as good as it does:SkypeVNCSkype is perfect to talk to anyone in the world. Over the internet it's free and the sound quality is perfect. You can even dial out to normal phone lines at very low cost.VNC let's you share a computer with another remote user.We are doing pair programming by using these two systems together. The system even works great for conferences with multiple people, to explain algorithms to everyone at the same time.Our setup that has proved to work best for big sessions with many people: We run the VNC server on a rented webserver with 100 MBit backbone access and all of us connect to that server. The person that leads the discussion and has most to say invites everyone else with Skype, so he gets the best voice quality.The internet makes completely new distributed companies possible:The best experts from all over the world can work together on the same project at the same time....at even lower ...