Sent from my mobile device (so please excuse typos) On 5 Jun 2011, at 11:14, Christian Grobmeier <grobme...@gmail.com> wrote:
>>>> Labs is a place for experimentation, not a place for developing software >>>> that is intended to be used. Your use case sounds more like software for >>>> use than an experiment. >>> >>> This is what I don't like. Imho Labs should be the place were >>> Committers can collaborate. >> >> It *is* a place where committers can collaborate. What is stopping you >> collaborating with committers? > >> What is it that you want labs to give you that it doesn't currently give> > > Sorry for my delay, but I had to think a little bit about labs. > > I identified the following scenarios. Lets imagine Max and Cheryl are > committers and want to collaborate. Thanks for taking the time. > 1) Max is committer at the Alpha-Project. Cheryl is committer at the > beta-project. Cheryl has an idea for the Alpha-Project, but since > Alpha does not allow here access to the SVN she cannot collaborate > with Max. Labs does not allow them to create a lab, b/c the change is > not really experimental but "usable software". They need to go to > github or apache-extras. If they intend to release usable software that requires a release they can go external as you suggest. Or they can work via patches and eventual commit access in the alpha project or they can make a proposal to the incubator, perhaps starting I labs if they so desire. I don't see the problem, perhaps others do. > 2) Max and Cheryl are committing somehow to Apache. They want to do > something completely new. There is no matching project at Apache so > far, so they can go into labs or to apache-extras Or incubator > 3) Max and Cheryl are both committing to Alpha. They have no need of > labs, because they have their projects sandbox to experimental stuff > which fits to their project. Yep. > In other terms, labs are useful if two committers are writing code > which is not related to an existing apache product. I don't see labs being limited to this type of project, but I do agree it supports it. > What I miss is option 1. Committers want to collaborate somehow but > not all have write access to the products sandbox. I don't see a > reason why they should go to apache-extras instead of using labs. If they don't want a release there's no problem. If they want a release then labs was never designed fr that and there are other vehicles by which the ASF provide support (the alpha project or incubator). Remember, one of the reasons that labs are not allowed to make releases is to prevent people routing round the incubator. I think to remove that limitation we need a stronger argument than this edge case. Others may think differently, let's hear from them. > >>> I understood Labs is no codedrop box. But I wish it would be a bit more of >>> that. >> >> What is a "codedrop box"? > > Sorry for my term. Its more a place were I can simply drop some code > ideas and leave it there, without caring what happens to them. A bit > like collecting snippets and ideas. Better we ignore my statement here No better we keep it because that is *exactly* what labs was created for. It is a "codedrop box". It is a place to play around with code without worrying about community. On two occasions that I recall I've inspected code here to get an understanding of how to achieve something. > >> Perhaps, in your case, we should be looking at making better use of >> apache-extras.org through the labs PMC (the ComDev PMC looks after it right >> now but we are deliberately hands-off on it). > > It seems you are considering apache-extras for my option 1. Only if option 1 needs to do a release and only if the incubator is not the right vehicle. > If yes, > apache-extras is somehow killing labs. Why should one use labs if he > can do everything with apache-extras more easy? It is not easier to give full access to all code in labs to all ASF committers in apache-extras. Extras is a public space where ASF rules don't apply. If a specific experiment does not wan to play by ASF rules (either labs or incubator) then extras might well be the rich place. The question is therefore: are the rules wrong? I'm not convinced by any of the arguments (so far) saying they are wrong. That doesn't mean I'm right though. Ross > > Cheers, > Christian > >> >> Ross >> >>> >>> --------------------------------------------------------------------- >>> To unsubscribe, e-mail: labs-unsubscr...@labs.apache.org >>> For additional commands, e-mail: labs-h...@labs.apache.org >>> >> >> >> --------------------------------------------------------------------- >> To unsubscribe, e-mail: labs-unsubscr...@labs.apache.org >> For additional commands, e-mail: labs-h...@labs.apache.org >> >> > > > > -- > http://www.grobmeier.de > > --------------------------------------------------------------------- > To unsubscribe, e-mail: labs-unsubscr...@labs.apache.org > For additional commands, e-mail: labs-h...@labs.apache.org > --------------------------------------------------------------------- To unsubscribe, e-mail: labs-unsubscr...@labs.apache.org For additional commands, e-mail: labs-h...@labs.apache.org
