Joel's argument is that Groove is a closed platform. That you have to pay to be part of it and if they don't give it away for free no one will write applications for it (this works for Jena). Dave's argument is that most platforms fail so you shouldn't bet that your application will become one. His solution seems to be get some money first and keep them coming back.
Joel: "...there is no free Groove runtime...nobody has Groove yet...That doomed the Groove idea from the start. I talked to some of the Groove "partners" who allegedly are developing software for Groove. "Was the Groove relationship worth anything?" I asked. "HA!" they said. "We paid $1500 and in exchange we get less than 10 clickthroughs a month from their web page. A waste of money. We couldn't even get Groove to share their customer lists with us.""
Dave: "Gassee had a sense of keeping the users entertained. He got it. Keep them coming back for new goodies. And maybe a breakthru. For that you need lots of risk-taking developers. That's how you get winners, not by hiring researchers to invent miracles for you...Apple did the opposite, invading competitive markets with void-creating white papers. "
Both Dave and Joel seem to care about the respective platforms in their articles and both felt the same way. Meanwhile, Ray just comments about how people are using Groove and marketing.
If anything, this is a lesson for anyone trying to write a piece of architecture. Listen to your developers and decide if you are a platform (building infrastructure) or an application. If you are a platform be open and free and hopefully charge for it later. If you're building an application, charge for it now and get any cent you can, because one day the big boys (like Microsoft or Apple or whoever) will come and give it away or integrate it into another product.