Matthieu Riou wrote: > Hi Lab! > > I've been very pleased by the lab announcement, such an environment for > internal innovation has really been missing at the ASF and the lab seems to > fill that gap. And congratulation for the lab web site too, it's pretty > nice! > > Now come the criticisms (I hope constructive) :) Whereas I understand the > rational behind the "no release" policy, I really think this is the worst > solution to the problem. Innovative, one-man (or even a couple of men) > projects need feedback from users more than any other project. Prohibiting > releases limits drastically the opportunity for labs projects to be used, > both in and outside the ASF. Not that many people are comfortable with > building a project from the sources to use it. It breaks the principle of > "Release early, release often" so important in the early life of a project. > In short, I believe it's a good recipe for failure. > > An answer to this could be to develop projects in the lab and release them > outside (in SourceForge for example). I don't believe this is a real > solution. Why providing an environment for projects development if one of > its important parts must happen outside? > > My (much) preferred solution would be to have all labs project release with > a disclaimer saying that the ASF doesn't endorse the release and the > project > (or project owner) is solely responsible. I think it would make sense as it > would remain consistent with the "lab" portion of the story (we're > experimenting, can't guarantee what's inside). The labs projects would be > responsible of their release and what they put inside and again I think > this > would make sense as the projects' creators are all going to be ASF > committers. People who have a minimum understanding of what can and what > can't be released. Besides, as stated in the labs bylaws, the previous > alternative was to host outside of the ASF, which also means that the > project is responsible for its releases. > > I have a project that I cherish and I would love to see it in the lab. > However if I can't make any release, I know this is going to be a serious > issue for the project to ever become something bigger. And that's far more > risk than I can afford. > > So is this "no release" policy something that could be changed? Or is it > final?
It's carved in stone. The FAQ on the web site explains the rationale: labs is place where you, and eventually other interested committers can create a seed for something bigger, or fail in the process. Anything more 'sourceforgy' will result in a massive abuse and overlap with the incubator, which we don't want. The "release early and often" principle does not apply to software innovation but to community building. Simply put, labs have nothing to do with community building. If your users can't deal with "svn co -r 489893 http://labs.apache.org/asf/repos/labs/xxx/";, they should not be around here in the first place. If you think that you absolutely need an official ASF-stamped release, you can propose that your lab is promoted and undergoes incubation. If you think that your users are not going to like "svn co" as a way to get your software, you are free to package the stuff and release it from where ever you want. And, of course, you are free to avoid labs altogether and either do it somewhere else or propose your project for incubation. Long story short: labs is not, and will never become as long as I'm around, a lightweight incubator. -- Stefano. --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
