RE: projects incubated by the incubator PMC
Jason, The short answer is that knowledge of the final destination is not a pre-requisite for entrance into the Incubator. The minimum requirement for entrance is sponsorship by a Member or Officer, and acceptance by the Incubator PMC. I've no particular care about where AltRMI or FtpServer end up, and can think of several suitable places for each. It seems that there is a community interested in collaborating on Ruper. With respect to where it could go, I don't know enough about it, yet, but I could imagine a project containing Ant, GUMP and Ruper. As I said to you last night, my personal preference is for multiple communities investing resources in the same problem domain to find a way to collaborate and merge their efforts. For example, I would like to see our container projects, such as HiveMind, Avalon, and the Geronimo kernel, find ways to collaborate. Similarly for Ant/Forrest/Gump/Centipede/Ruper and Maven/Continuum/Wagon. Why? In my view, collaboration is more effective, because people have a chance to combine their ideas and efforts. But, I don't know that we can do more than encourage people to collaborate. Geronimo and Directory are both intended to be top-level projects. Their exit requirements will include their ability to operate as such, as well as demonstrate the responsibility and trustworthiness that is a necessary condition for a top-level project. The phrase "going directly to TLP" is misleading, since those projects are in the Incubator, and won't exit until they are ready. Your comment implies that you feel that they should go from the Incubator to someplace else, e.g., Jakarta, and then graduate again to TLP status. I believe that sticking more large projects into umbrellas would be counterproductive. There is a proposal that for projects intended to go under an existing PMC, the Incubator's primary involvement would be ensuring that IP issues are resolved before the project's incubation status could be cleared, and the other PMC would be responsible for Community Building and teaching "The Apache Way." But for new projects included to stand on their own, the Incubator would be responsible for the Community issues. --- Noel - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
RE: projects incubated by the incubator PMC
On Sun, 2003-12-07 at 13:06, Noel J. Bergman wrote: > Jason, > > The short answer is that knowledge of the final destination is not a > pre-requisite for entrance into the Incubator. The minimum requirement for > entrance is sponsorship by a Member or Officer, and acceptance by the > Incubator PMC. I really think that should be changed. How can a project come here without a thought to its final destination? Don't you think that's important? That we could potentially end up with N projects just floating about here indefinitely is not a good thing. Additionally if no group in Apache expresses an interest in the project being incubated why are we incubating it? I think this all started with Tapestry and Jakarta was the eventual landing zone. What's going to happen with AltRMI and the FtpServer? They just sit in the incubator indefinitely? There's no IP issues as these came from within Apache anyway? If an incubated codebase has no slated home within then I would ask how it even landed here in the first place? My point in raising this discussion is to focus on Ruper as it's a small tool with no final landing destination as far as I can tell. I wasn't paying attention to lists, was away for a few days and came back to see it being mentored/helped by the Incubator itself. I voted +1 making the silly assumption that a destination was required. To my knowledge Nicola has made attempts to have projects like Centipede, Ruper be absorbed into Apache via Ant and those haven't been successful so I am assuming that's probably still the case (someone feel free to correct me if I'm wrong). Adam Jack contacted me personally just to shoot the shit about Ruper and I told him his code would be welcome in Maven. If the code has no final destination then I am a very strong -1 to the code being accepted. If Ant is interested that is good as it fits there. No one contacted the Maven PMC so obviously there is no interest there and I did make the offer of assistance to Adam. So that being said Ruper folks what the long-term plan? Additionally what's the long term plan for things like AltRMI and the FtpServer that are just floating at the moment. The projects can't operate out of the incubator and if no one asks where those projects are going it is likely they will just stagnate there. Projects that are free and clear of IP problems can't operate out of the Incubator can they? -- jvz. Jason van Zyl [EMAIL PROTECTED] http://tambora.zenplex.org In short, man creates for himself a new religion of a rational and technical order to justify his work and to be justified in it. -- Jacques Ellul, The Technological Society - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: projects incubated by the incubator PMC
There is no reason for a project to have a final destination until it has to go somewhere other than incubator, at which point it can decide whether it wants to be a TLP (calling for a board vote) or part of an existing project (calling for that project's pmc to vote). Maybe we should have a six month limit on incubation, where the result is promote or punt. We have no way to measure the worthiness of a project before it has even started, and generally speaking we are MUCH MUCH MUCH better off if a project gets started in Apache rather than on sourceforge. I don't care if there is some overlap with Wagon, nor do I care for any further discussion about which one is better -- if I can't find some objective criteria for evaluating software, then I obviously don't need that software. Once I need it, I can figure out for myself which one is better -- I don't need someone to assume that for me. It is easier for Apache to support both projects than to arbitrarily choose between the two. Roy - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: projects incubated by the incubator PMC
> Adam Jack contacted me personally just > to shoot the shit about Ruper and I told him his code would be welcome > in Maven. I was hoping that Apache was somehow Utopian and folks could co-operate, bury ego, and work for the greater good of softwaredom, yada yada. When you invited the Ruper code to become part of Wagon in Maven, you clearly explained that certain folks at Apache would not feel welcome there. That is disappointingly limiting to me, since in my world view, community is as valuable as code, and we are trying to satisfy [EMAIL PROTECTED] not less. Naive, or not, I still hold out hope of working with as many people as practical. I think it is great that Wagon is being developed as actively as it is, but it is far from mature & it/Ruper have different approaches to solving the problem. Introducing Ruper into Wagon or Wagon into Ruper, at this point, would probably confuse each's design & simply server to dissatisfy both teams/goals. I figure Ruper deserves a shot equal to the one you've granted to Wagon. It seems to me that having Ruper available, gives Apache another shot at solving the problem, and if/when a time for an equal opportunity merge arose, I'd be game to cooperate with it. regards, Adam - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: projects incubated by the incubator PMC
> I don't care if there is some overlap with Wagon, nor do I care for > any further discussion about which one is better -- if I can't find > some objective criteria for evaluating software, then I obviously > don't need that software. Once I need it, I can figure out for myself > which one is better -- I don't need someone to assume that for me. > It is easier for Apache to support both projects than to arbitrarily > choose between the two. Nicely put. Apache folks have clearly & unemotionally explained to me that some times it just makes sense for teams to agree to differ, and to work amicably/respectfully, yet apart. Since we have 2 (at least) code bases with different approaches to solving the problem, I think this might be just such a case. It is frustrating that (in this day and age) there isn't a simple solution for managing repositories of jars/artefacts, downloading upon demand and keeping environments healthy. Heck, this stuff isn't rocket science, it just needs to exist & fit openly into user environments. Two are better than none... regards, Adam - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: projects incubated by the incubator PMC
On Sun, 2003-12-07 at 18:33, Roy T. Fielding wrote: > There is no reason for a project to have a final destination until > it has to go somewhere other than incubator, at which point it can > decide whether it wants to be a TLP (calling for a board vote) or > part of an existing project (calling for that project's pmc to vote). > Maybe we should have a six month limit on incubation, where the > result is promote or punt. > > We have no way to measure the worthiness of a project before it has > even started - test coverage - clarity of documentation - established user base And those are three criterion that can easily be measured and I think that is a good start. Something that's far easier to pinpoint than the vague term "community". The three above listed criterion are things that can easily be faked. Tapestry satisfied these quite easily. > , and generally speaking we are MUCH MUCH MUCH better off > if a project gets started in Apache rather than on sourceforge. How so? I'm genuinely interested in why you think so. > I don't care if there is some overlap with Wagon, nor do I care for > any further discussion about which one is better -- if I can't find > some objective criteria for evaluating software, then I obviously > don't need that software. Once I need it, I can figure out for myself > which one is better -- I don't need someone to assume that for me. > It is easier for Apache to support both projects than to arbitrarily > choose between the two. It wouldn't bother me at all for other projects to come here that are overlapping partially or completely that's not my point. My point is the incubator shouldn't be a source code limbo. How did the decision come about that a complete path for incubation not be requisite as part of a proposal? Didn't the incubator start with Tapestry in which successful incubation meant the movement to a final destination? So theoretically, given the current situation, any number of codebases could land here and sit indefinitely? That we are actually encouraging people to use Apache's incubator as an alternative to SourceForge? > Roy > > > - > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] -- jvz. Jason van Zyl [EMAIL PROTECTED] http://tambora.zenplex.org In short, man creates for himself a new religion of a rational and technical order to justify his work and to be justified in it. -- Jacques Ellul, The Technological Society - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: projects incubated by the incubator PMC
On Sun, 2003-12-07 at 21:47, Jason van Zyl wrote: > On Sun, 2003-12-07 at 18:33, Roy T. Fielding wrote: > > There is no reason for a project to have a final destination until > > it has to go somewhere other than incubator, at which point it can > > decide whether it wants to be a TLP (calling for a board vote) or > > part of an existing project (calling for that project's pmc to vote). > > Maybe we should have a six month limit on incubation, where the > > result is promote or punt. > > > > We have no way to measure the worthiness of a project before it has > > even started > > - test coverage > - clarity of documentation > - established user base > > And those are three criterion that can easily be measured and I think > that is a good start. Something that's far easier to pinpoint than the > vague term "community". The three above listed criterion are things that > can easily be faked. Tapestry satisfied these quite easily. ^^^ ... things that can't be faked ... > > , and generally speaking we are MUCH MUCH MUCH better off > > if a project gets started in Apache rather than on sourceforge. > > How so? I'm genuinely interested in why you think so. > > > I don't care if there is some overlap with Wagon, nor do I care for > > any further discussion about which one is better -- if I can't find > > some objective criteria for evaluating software, then I obviously > > don't need that software. Once I need it, I can figure out for myself > > which one is better -- I don't need someone to assume that for me. > > It is easier for Apache to support both projects than to arbitrarily > > choose between the two. > > It wouldn't bother me at all for other projects to come here that are > overlapping partially or completely that's not my point. My point is the > incubator shouldn't be a source code limbo. > > How did the decision come about that a complete path for incubation not > be requisite as part of a proposal? Didn't the incubator start with > Tapestry in which successful incubation meant the movement to a final > destination? > > So theoretically, given the current situation, any number of codebases > could land here and sit indefinitely? That we are actually encouraging > people to use Apache's incubator as an alternative to SourceForge? > > > Roy > > > > > > - > > To unsubscribe, e-mail: [EMAIL PROTECTED] > > For additional commands, e-mail: [EMAIL PROTECTED] -- jvz. Jason van Zyl [EMAIL PROTECTED] http://tambora.zenplex.org In short, man creates for himself a new religion of a rational and technical order to justify his work and to be justified in it. -- Jacques Ellul, The Technological Society - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: projects incubated by the incubator PMC
On Sun, 2003-12-07 at 18:58, Adam R. B. Jack wrote: > > Adam Jack contacted me personally just > > to shoot the shit about Ruper and I told him his code would be welcome > > in Maven. > > I was hoping that Apache was somehow Utopian and folks could co-operate, > bury ego, and work for the greater good of softwaredom, yada yada. When you > invited the Ruper code to become part of Wagon in Maven, you clearly > explained that certain folks at Apache would not feel welcome there. Feel free to cite any part of the email I sent you that says anything about excluding anyone. You have my full permission. Even me with my fiercest of critisisms couldn't stop anyone from participating. The truth of the matter is that those bitching and whining most have done the least to try and cooperate. > That is > disappointingly limiting to me, since in my world view, community is as > valuable as code, and we are trying to satisfy [EMAIL PROTECTED] not less. > Naive, or not, I still hold out hope of working with as many people as > practical. In my world view the quality of code is paramount and the community is a natural by product of the process that made good code. You can have all the community you want but it's not going to make poor code any better. Communities form of their own accord and that is a good thing. Communities based on artifice which are used to prop up code that should otherwise die a natural death is bad. > I think it is great that Wagon is being developed as actively as it is, but > it is far from mature & it/Ruper have different approaches to solving the > problem. Introducing Ruper into Wagon or Wagon into Ruper, at this point, > would probably confuse each's design & simply server to dissatisfy both > teams/goals. I figure Ruper deserves a shot equal to the one you've granted > to Wagon. > > It seems to me that having Ruper available, gives Apache another shot at > solving the problem, and if/when a time for an equal opportunity merge > arose, I'd be game to cooperate with it. Although I find it somewhat suspect that code that has been rejected by other Apache projects is now coming through the incubator Wagon/Ruper is not the issue. The larger issue in my mind is the path of code through the incubator. I honestly don't understand the notion of letting code flounder around in the incubator indefinitely as serving any greater good. -- jvz. Jason van Zyl [EMAIL PROTECTED] http://tambora.zenplex.org In short, man creates for himself a new religion of a rational and technical order to justify his work and to be justified in it. -- Jacques Ellul, The Technological Society - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
