Le 12/17/13 3:17 PM, Gabriela Gibson a écrit : > Hi everyone, Hi Gabriela,
> > as you probably are aware, I had an idea fail to win an Apache Lab spot > recently. > > Being the incurable optimistic realist that I am, I counted no -1's, one +1 > vote, one committer who joined my project immediately, and the fact that > the idea wrote itself in about 16 hours flat. > > So on balance, I think as ideas go, this is a pretty good one, and whilst > Apache Lab technically failed for me, it is still a big fat profit in my > book :-) Starting a vote just before a week-end, 2 weeks before Xmas, is not specially a good timing :-) > > It was also recently suggested that Apache Labs could do with some ideas. > So, having some experience of the Apache Lab process, here is my personal > frank take and lemonade recipe, in no particular order and without > guarantee of validity. > > Enjoy the lemonade! :) > > =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= > > * People who are eligible to vote and the process > > Problem: Most of the Apache PMC's are busy, successful people and may not > have much spare time and there are only 72 hours for them to find out, > think about and approve an idea, and then X-mas weekend shopping + real > life happens. Absolutely. > > Solution: Scrap the vote, the mechanics are too failure prone. Or polish the proposal, and select a better date... > > An Apache Committer usually already has passed the community quality test > as person who can partake constructively in the community, so, they can be > trusted to use the Lab space productively if given a spot. If not, well, > this can be dealt with on a case-by-case basis. > > Open up the Apache Lab committee and appoint some Lab Guardians from the > wider community whose job it will be to facilitate the success of ideas > gifted to us. I'm with you. Im not sure the labs need some vote. Who is qualified to define what is a good idea ? > > > * What is a 'quality' idea? > > One that survives the ensuing fight. It might start small and ugly, but it > needs opinions to grow and find shape. A good idea can awe people into > silence (what could I possible add? Will my critique look mean?). In that > case, send praise! > > Even a bad idea can fire people up into joining the melee and at the end, > the idea often has morphed into something totally different, and shines > because it's been roughed up to a high polish. > > So, we can at best say 'I like', but, until the concept has been thoroughly > worked on by the community, 'quality' is relative. Maybe we can define it > best as 'quality ideas inspire people to participate in their formation > process'. > > But, the process makes or breaks the idea, not the other way round. +1 > > > * Purpose of Apache Labs > > I see Apache Labs as a prelude to Incubator -- it should be the nest where > the egg is laid before it can be placed for hatching. Not so sure... Not always. I see labs as a place I can experiment, and see if it brings something. More important to me, it's a collaborative place where I can share some code with a few other committers, without having to go through incubation. Whether it becomes something, or not, is irrelevant. Well, github is also a good place to start with, but, well, I'm an apacher ;-) > > Problem: This currently isn't happening, the mailing list is very quiet, > and part of the reason may be that Apache Labs can look like a 'tough gig' > and people who are looking for a space to play, might be intimidated by > seeing stuff get turned down. The initial stepping out and presentation is > quite a fragile and worrying process and takes a lot of guts on part of the > designer. I have started 2 labs : dungeon and Mavibot. One has stalled a long time ago (lack of time, even if I think it's an interesting idea), the other has graduated (directory.apache.org/mavibot). Another lab I know of has graduated too : Vysper (mina.apache.org/vysper-project). The labs that graduated were those gathering some committers, but not only. It was mainly about them going a bit farther than just an experiment. They required a lot more of effort, up to a point it was valuable to get them included into a TLP project. Well, some may question the reason they didn't went through incubation. But all in all, the people working on those projects are already apache people, and the umbrella projects accepted them because they were deeply connected. In some way, it was more a question of not polluting the TLP project at first... Ok, they were a bit specific, quite different from your proposal. It might have been easier for the labs PMC to vote them in for those reasons... > > Solution: Open up the playground. Invite & recruit --- the more the > merrier. Apache Labs should be a noisy, lively place that is easy to > enter, and could do with some guidance as to how the community should work > ideas. Due to the diversity of labs, I'm not sure it will ever be possible... And it's already easy to participate : you just have to be a committer ! (4000 potential people, and counting...) > > I do not see it as problem that non Apache members may not commit, this is > what the mailing list and patches are for, and if Lab does it's job > correctly it also functions as an entry process for potential new > committers who make friends in this informal place and hopefully get roped > into existing Apache projects. Basically, you are proposing to start a new GitHub/Sourceforge area. I'm not sure it's a good idea to try to mimic what already exists outside (NIH syndrom, as usal...) > > Additional thought: Some ideas are spanning several projects and the > originating project space may sometimes not be the right forum to test an > idea because the audience may be not be diverse enough. Labs to the rescue ! > > Apache Labs should be that space where projects can share ideas of this > kind with the rest of the community before deciding to take it into their > space or perhaps homing it elsewhere. Also, if projects bring ideas to Lab > and invite their members to go take a look, it will promote Lab in their > communities and we'll get more visitors. I would kindly suggest that you advertize your project before pushing a vote. If a few committers +1 the proposal, be sure that the labs PMC will consider that some valuable idea needs to be tested. And the labs low level traffic ML will help them : having 10 mails with 10 +1 on this ML will force them to react ;-) > > > * A culture for nurturing ideas > > An idea is always just a little bit of raw material. It can be smashed into > pieces and new ideas can be built from the fragments. It can grow into > something wonderful, or be a good demonstration of why we do not do things > that way. It's a teacher. And, good teachers are always ready to deliver > proof of what they claim when challenged. Teachers usually do some kind of preparation before throwing proofs to their students ;-) > > Of course this should and can be done politely, but in essence, the process > of working an idea into shape is always contentious to some degree. It's > like making sausages, the result is great, but the process is somewhat > mucky and needs some rules because it's etiquette lies somewhat out of the > normal social space, it's a melee by definition, but even melees have (and > need) engagement rules. > > Hence, Apache Labs needs a statement how ideas should be treated by the > community and we should have guidelines that help people make great > omelettes whilst being careful to compost the shells to make new fertile > compost to grow new stuff in. I sympathize with your concern, but here, I'm not sure that labs can do anything to help to make ideas becoming something real... People produce code from ideas, if they are interested in it. The lab is just an envelop, a nest. > > Apache Labs also needs a more cleared defined pathway of how to reach the > goal of entering the Incubator. > > > * People who design ideas > > Some designers only have one idea they've carefully worked a lot on, > others are literally plagued by a torrent of them and just make nice > sketches they want to show and share. > > Some want to keep control of their ideas, others are just happy to give > them away. > > However, all those creative people all have one thing in common: every > designer always defends their creation like a mother hen, because it's > usually their heartfelt opinion and best effort of creating the shape that > things should be. If it were not so, we would not have ideas! > > Some designers can fall into the trap of viewing their idea as a public > litmus for their capability, and some people may feel they have a > professional reputation to protect, and public failure feels > uncomfortable. This dread is more pronounced in designers who are new to > publishing their ideas, and it's totally natural and part of the growth > process of a designer. A good culture will go a long way of making that > daunting first step easy and comfortable to take. > > Apache Labs needs to be explicit about stating that an idea fail is not a > personal failure, and that it's better to have an idea fail than not be > presented, and that stepping out with the idea is a 'great job > done' as is --- if an idea that is gifted to Apache Labs fails or succeeds, > we all learn something valuable, but the community is the actual body that > is collectively responsible for the outcome, and, win or lose, we all have > part of the honour. > > Ideas command our respect, and part of this is creative destruction. I would say : don't be scared ! A lack of +1 does not mean "your idea sucks". What is important is to share those ideas, and see if some people aggregate. If it's the case, great ! Otherwise, well, ideas are not like gold mines : the vein never dries (but sometime, you find lead instead of gold ;-) ... > > * How to fail an Apache Lab idea gracefully and productively > > I view Apache Labs as an entry point to Incubator. It's where the egg is > conceived, and formed --- it arrives as a tiny bit of DNA instruction > (inspiration), grows into a big yolk(idea), the albumen (community) that > will feed the embryo (project) is established , and a pretty shell is > formed(shape) and finally the egg is laid, ready to be put into Incubator > for hatching. > > The more eggs we produce, more successful the Incubator will be, and the > more ASF will grow. Every idea we can win for our world is an addition to > our treasure hoard and a valuable chance for us. > > We cannot and should not control the creative process too much, it has it's > own momentum and at best we can facilitate the magic that happens, when it > does. There are of course techniques for conjuring up this happy event, > and using them enhances the chances =) AFAICT, labs's process is extremely lightweight ! 3 +1 is not really what I call a 'process', barely an approval... > > Let ideas fail naturally in the ensuing melee, a Lab is just a space where > things go *boom*, or serendipity turns mere mould into the saviour of > humanity. Again, sell your ideas ! If you weren't able to collect 3 +1, there are good chance that you didn't pushed far enough before going "public". Opening labs too much may lead to the exact opposite result : a flow of useless ideas, hiding the gems. > > It's a place where we discover things, not where already complete things > are delivered, and it's impossible to evaluate an idea properly before it's > been tested by the community. Film makers start with a Pitch : a 3 minutes presentation of their idea, nothing fancy, just an exposure of their "idea". This is the very first step, it costs nothing, but you *have* to go through a commity that let you turn the picth into a synopsis. The synopsis is a 10 minutes presentation of the story. Labs vote on pitches (ideas) and deliver synopsis. Becoming a TLP is the result. But the pitch has to be exciting at first ! > > An idea has failed when it's not growing or not rebuilt into a new idea. > This is a longish process and a community effort, after all, if the > designer could go solo they would, but they come to Apache Lab to gather a > community that turns their dream into reality. > > So, give every idea a year to be tested to destruction by all-comers, and > if it hasn't made it to Incubator by then, we have our answer and it's time > to run a script that vaults the idea's repository and stashes it into a > zip file complete with all the mails it attracted, ready to be delivered > back safely to it's designer. Automate that process -- a retired project > can always be resurrected, but hard deadlines tend to motivate people to > make pace because there is no 'manana' option. > > A failed idea also deserves an 'exit interview' in order to help the > designer to grow their creation-fu, and perhaps to point them into a more > productive direction, Apache Lab may just not be the right place for this > particular creation, despite everyone giving it their very best shot. > > So, the exit process is where all Apache Lab guardians should give a frank > assessment to the instigator, and this feedback should be the designers' > document and he gets to decide which > parts to publish on the Apache Lab mailing list. Some designers may > publish the entire thing, others may want to keep it private. It's their > personal reward for daring to step out into the lime light and > gifting the Apache Community a toy to play with, and a pretty good > consolation prize in it's own right. > > =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= > > Anyway, that's all the lemonade I got for now, I hope you liked my recipe, > if you did, go find some lemons and try it out :) :) :) I love lemonade ! Don't give up just because the first drop was too bitter : add some sugar in it, not too much, and try again ! Nobody will blame you or mock you ! -- Regards, Cordialement, Emmanuel Lécharny www.iktek.com --------------------------------------------------------------------- To unsubscribe, e-mail: labs-unsubscr...@labs.apache.org For additional commands, e-mail: labs-h...@labs.apache.org
