Roy T. Fielding wrote: > On Jan 19, 2006, at 12:36 AM, Raphaël Luta wrote: > >> We're doing loops here. My point in this thread is that initial code >> quality does matter in a code grant incubation because it is often >> burdened by backward compatibility with existing applications and >> thus major restructure may require a revolution which can hardly >> safely happen in the early months of the project open-source life. > > > And all those points are wrong. There is no burden of backward > compatibility because it must be an entirely new product -- all > of the names change anyway. A major restructure is a good idea; > that is, after all, why we founded Apache as a project to replace > NCSA httpd 1.3R, which was replaced by Shambhala within 6 months. > And it certainly doesn't have to happen "safely" -- the project is > going to be shooting for TLP status, which means about a year or more > under incubation before it can even do real releases, and the more > hard decisions the group has to make (in public), the better they > will learn how to collaborate. >
I fail to see how you can apply the httpd situation (standalone server, initial community with independant contributors, most downstream users relying on standard http/cgi) to the current proposal (development toolkit/"framework", community with hierarhical relationships and downstream users code-dependant on toolkit). You take the optimistic view that the community would work as expected; I have a pessimistic view that it will not without some cost to the ASF if at all. > Honestly, once the name is changed to something neutral like Kabuki, > none of your objections make any sense. Especially the ones about > code quality, since most of our projects started with code that > needed a serious re-arch almost immediately. I would love to see > three or four different ajax toolkits under the ASF, each with > its own architectural focus, and let them compete for developers, > but we can only approve one podling at a time. > +1 I just wish they would join with a least a nucleus of public community. > Meanwhile, I do think that any proposal to the Incubator needs at > least three active Apache committers involved, preferably members > that are willing to do infrastructure tasks. Incubator podlings > are seriously infrastructure dependent and the existing volunteers > are already tapped-out. > +1 -- Raphaël Luta - [EMAIL PROTECTED] Apache Portals - Enterprise Portal in Java http://portals.apache.org/ --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
