Hi all, After a long time of thinking, I decided to express my point of view hoping that it might incite a few ideas/solutions.
I see the whole problem of labs inactivity as two separate issues: 1. why current labs have low activity? 2. why aren't new labs created constantly? I am one of the committers that currently has a lab created and from my experience I find it normal that labs don't have activity from time to time. Being experimental projects, they are most probably developed in their contributor's free time. I am currently a graduating student and I have had periods of serveral months in which I simply couldn't find any free time to work on my lab... Assuming other committers have full time jobs and other open source projects where they are actively involved, free time to work on a lab can be *really* hard to find. However, to have a wider, realistic view of 1/ some input from current labs would be useful. One approach could be to ask lab creators to fill a short 3-question online form (or a wiki page) about the reason of inactivity. It shouldn't take longer than 2 minutes to fill it in. As labs are generally fail fast projects, I wouldn't find it unusual to do this every 6 months or so, asking contributors about the status of their project, if they need anything, etc. IMO, the Apache community/resources is the strongest item Labs hold over all other hosting options (at least I chose it for that). Anytime, labs contributors can ask for feedback or advice from related projects and mentors inside the community. I would emphasize and facilitate this in particular. The major problem seems to be 2/. The committers index shows somewhere around 2800 committers in the community at this moment. Given such a number, I would expect 2 or 3 lab proposals per month. This isn't happening because, as others stated, people don't know about labs. I, for one, luckily found out about Labs while chatting with Ant about my new project idea at the time when I was already a 6 month active committer... IMO the advertisement problem Labs is facing extends to the whole ASF ecosystem. Let me explain what I mean by that. One thing I noticed is that I'm approaching one year of being a committer and I am still finding out about many Apache projects I wasn't aware of. I have subscribed a long time ago to the announcements mailing list just to read the one paragraph description each project includes in the release announcement and still I can say I'm not aware of a lot of projects. Few incubator projects and no labs post there so no info about them. Subscribing to the incubator and labs lists isn't a way to go as most of the discussions there exceed my purpose. I've tried entering [1], [2], [3] to find out about projects but it is so painful to check each site and mailing list eventually (I quit after 10 projects). What I would find useful is a blog where Apache projects have the opportunity to write short descriptions (with code samples if they wish) about their project. It wouldn't have more than 1 post per week and projects can express their desire to write a post and will be scheduled in the next available week. All that's needed is a few people to manage the FIFO queue and work with the scheduled project on the post from the current week. A post on the foundation blog, a tweet from the ASF account, an email on the announcements@ and committers@ would spread the word about it and it would make a great resource for interested people to find out about Apache projects. I'm sure that projects looking for contributors would be more than happy to present their project there (especially incubator and labs projects). We could start with a description about Apache Labs. If you find this a good idea, I am willing to work on this. It shouldn't be too time-consuming, probably 2 or 3 people will make it (details TBD later if it would belong to infra, com-dev or a new project should be created). Feel free to add your opinions about it. Thank you for reading all the way through, Florian [1] http://apache.org/ [2] http://incubator.apache.org/ [3] http://labs.apache.org/labs.html On Wed, Jun 15, 2011 at 11:57 PM, Niall Pemberton <niall.pember...@gmail.com > wrote: > On Wed, Jun 15, 2011 at 9:32 AM, ant elder <ant.el...@gmail.com> wrote: > > On Wed, Jun 15, 2011 at 8:34 AM, Ross Gardler <rgard...@apache.org> > wrote: > >> On 15/06/2011 02:11, florent andré wrote: > >> > >> ... > >> > >>> B) Doing releases : > >>> Some express the need of doing release. Con expressed are related to > >>> "legal engagement" of Apache in such release and the necessary oversee > >>> of already really busy mentors. > >> > >> This misses the original point for not allowing releases. Labs is not > >> intended to be a way of "routing around" the incubator. > >> > >> It also misses the point that labs is not supposed to be producing > usable > >> code, it's a place for fast fail experiments. Usable code should go > through > >> incubation. > >> > >>> C) Include non Apache's commiter (let's call them Externals) : > >>> Ant propose this but the answer was "it's not possible" [1] because : > >>> - Externals have to sign a CLA > >>> - Will accept Externals as commiter > >>> > >>> Well, this is what is done is no problems during a new incubator > >>> project... Where is the problem here ? > >> > >> If we make the rules the same as the incubator why have labs at all? > This is > >> supposed to be a simple, no commitment playground for people who have > earned > >> merit. It is *explicitly* not a place for creating new communities. > >> > >>> Here comes two more personal POV, idea, whatever : > >>> > >>> - labs like an idea box : > >>> From my point of view - may be false - actual labs is more a "write > >>> code place" than an express, discussion, grow and develop ideas place. > >>> May a way to write some (visible from labs site) webpages, have a wiki > >>> for idea expression could be good... Some others tools can be imagined. > >> > >> It's not the incubator, why do we want to make it like the incubator? > >> > >> Personally I really like the fact there are almost no rules around here. > If > >> I want to do something with my fellow committers I just ask for a lab > and > >> its done (I've never owned a lab, but I have contributed to a couple and > >> used a couple). If I want to do something with a broader community then > I go > >> to the incubator or offsite. I really don't see the problem. > >> > >> I don't want to see labs go into the attic but I don't want to see it > gather > >> a whole raft of rules that prevent me people from freely experimenting > the > >> way it is intended. > >> > > > > The main purposes of the Incubator are about IP clearance and learning > > how to run a project in The Apache Way with community building etc. > > The Labs guideline that 'Labs is not intended to be a way of "routing > > around" the incubator.' to me relates to those two things. If one or > > more people work on a Lab they don't necessarily need to be doing it > > The Apache Way, and if they have submitted a CLA and aren't violating > > that then there shouldn't be IP issues, so if a Lab did a release I > > don't think it would necessarily imply "routing around" the incubator. > > > > I also don't see how following the normal Apache processes for things > > like committers or releases sees Labs "gather a whole raft of rules" > > it looks more like the opposite to me - Labs has come up with a whole > > raft of rules to prevent Labs doing things and thats one of (just one > > of) the reasons people are going elsewhere. > > > > IMHO if the Labs rules were simplified and more innovative or even > > experimental ways were found to fit in with the ASF then that would > > help make it more vibrant. Why not just give it a try! > > My question would be, if why not do what you want in the incubator > where there aren't the restrictions that Labs currently has? > > Niall > > > > ...ant > > > > --------------------------------------------------------------------- > > 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 > >
