Re: Podlings & Apache Project Maturity Model (was RE: [DISCUSS] Graduate Apache RocketMQ from podling to TLP)
> I'll be honest, I have no idea why they > think they have to do it, but they do it. I suspect it is because projects are motivated to graduate and believe that one or more people on the IPMC with -1 their graduation if they do not complete the model and put a check mark in every box. I suspect they believe this because they see other discussions or votes on this mailing list end up with -1 votes (or "fix it for your next release" statements) that they believe go beyond the official policy requirements of the issue at question. - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Podlings & Apache Project Maturity Model (was RE: [DISCUSS] Graduate Apache RocketMQ from podling to TLP)
Hi Bertrand, On Mon, Sep 4, 2017 at 4:54 AM Bertrand Delacretaz < bdelacre...@codeconsult.ch> wrote: > Hi John, > > On Fri, Aug 25, 2017 at 9:11 PM, John D. Ament > wrote: > > ...Its unfair for us to put some stake in the ground expecting podlings > to > > match up 100% on the questions. Many of the questions are subjective - > is > > the code easy to discover? respond to bug reports in a timely manner?... > > Ok, I think I understand your reluctance now: you don't want us to set > a gate for graduating podlings that many TLPs might not pass. > > I'm not sure how you got that from a comment about the subjective nature of something of the questions, but sure, that's definitely one problem. I'll also take this time to note that it is not easy to discovery the source code for the com dev website, nor the actual contents of the APMM webpage. > I agree with that, and although I'm a strong supporter of the Maturity > Model (having initiated it that's understandable ;-) I'm totally ok > with podlings graduating without fullfilling all of its requirements. > > In my view the model is: > > 1) A good tool to help discover areas that podlings might have neglected. > Yes. > > 2) A good tool to help podlings look at where they stand, and what > they might still need to improve after graduation. > Yes, and in a way to reflect upon their current state and see if it's going well or not. > > 3) A good tool to express our ideal way of doing things, in a concise > way, and evolve that definition over time. > Its also of use to look at an incoming podling and ask "are these things you're interested in achieving?" > > Based on this I will continue to push for podlings to come to > graduation with a self-assessment based on that model. > > OTOH I'm fine with us clarifying that it's not a requirement. > > The website already lists it as a recommendation only, not a requirement https://incubator.apache.org/guides/graduation.html#other_issues (I'm waiting on the build to fix the link). > -Bertrand > > - > To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org > For additional commands, e-mail: general-h...@incubator.apache.org > >
Re: Podlings & Apache Project Maturity Model (was RE: [DISCUSS] Graduate Apache RocketMQ from podling to TLP)
On Mon, Sep 4, 2017 at 4:34 PM, Jim Apple wrote: > ...I think the current Model has a number of vague statements that are > unlikely to be interpreted consistently without clarification. For > instance "The project is open and honest about the quality of its > code.".. That's a good example where a yes or no answer does not make much sense. What's useful is a response like "we include a list of know issues with each release" or "experimental or unstable modules are clearly labeled as such" that demonstrate that the project takes quality seriously and does not try to hide problems from their users. I think such items help projects take a look at their overall well-being, and at the same time they make it impossible to base graduation on a given score based on that model - because there are no black and white answers for those items. So IMO we're doing the right thing in recommending that projects do a self-assessment based on that model before graduation, yet not requiring specific results for graduation. -Bertrand - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Podlings & Apache Project Maturity Model (was RE: [DISCUSS] Graduate Apache RocketMQ from podling to TLP)
On Mon, Sep 4, 2017 at 7:26 AM Shane Curcuru wrote: > Bertrand Delacretaz wrote on 9/4/17 4:54 AM: > > Hi John, > > > > On Fri, Aug 25, 2017 at 9:11 PM, John D. Ament > wrote: > >> ...Its unfair for us to put some stake in the ground expecting podlings > to > >> match up 100% on the questions. Many of the questions are subjective - > is > >> the code easy to discover? respond to bug reports in a timely manner?... > > > > Ok, I think I understand your reluctance now: you don't want us to set > > a gate for graduating podlings that many TLPs might not pass. > > > > I agree with that, and although I'm a strong supporter of the Maturity > > Model (having initiated it that's understandable ;-) I'm totally ok > > with podlings graduating without fullfilling all of its requirements. > > > > In my view the model is: > > > > 1) A good tool to help discover areas that podlings might have neglected. > > > > 2) A good tool to help podlings look at where they stand, and what > > they might still need to improve after graduation. > > > > 3) A good tool to express our ideal way of doing things, in a concise > > way, and evolve that definition over time. > > > > Based on this I will continue to push for podlings to come to > > graduation with a self-assessment based on that model. > > > > OTOH I'm fine with us clarifying that it's not a requirement. > > Proposal: It should be a *requirement* for the podling to self-document > their maturity model answers in the [DISCUSS] thread before IPMC > graduation vote. The requirement is having done it, not passing it. > > To be clear - it is not a requirement today for podlings to complete the Apache Project Maturity Model. I'll be honest, I have no idea why they think they have to do it, but they do it. I don't want to stop them from doing it, but I want to stop them from incorrectly stating they pass everything. I also want to clarify that the answers should not be "Yes" or "No" but an "OK" and explaining their response, or perhaps an "N/A" and explain why that line doesn't apply to them yet (the most common issue is related to something Dave's brought up recently where projects are answer the security reporting questions). > It's *very* helpful to have podlings consider their growth using some > form of structured and consistent criteria, so IPMC (and board) can > consider how different podlings see themselves compared to past podling > history. > > It doesn't mean every podling has to say "Yes 100%" to every question, > just that they've considered each point and can describe their situation > there if not. I'd expect plenty of podlings would have some missing or > "we're not completely here" on some points, but still be healthy and > well-self-governing communities ready to graduate. > > > -- > > - Shane > https://www.apache.org/foundation/marks/resources > > - > To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org > For additional commands, e-mail: general-h...@incubator.apache.org > >
Re: Podlings & Apache Project Maturity Model (was RE: [DISCUSS] Graduate Apache RocketMQ from podling to TLP)
> It's *very* helpful to have podlings consider their growth using some > form of structured and consistent criteria, so IPMC (and board) can > consider how different podlings see themselves compared to past podling > history. I think the current Model has a number of vague statements that are unlikely to be interpreted consistently without clarification. For instance "The project is open and honest about the quality of its code." - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Podlings & Apache Project Maturity Model (was RE: [DISCUSS] Graduate Apache RocketMQ from podling to TLP)
Bertrand Delacretaz wrote on 9/4/17 4:54 AM: > Hi John, > > On Fri, Aug 25, 2017 at 9:11 PM, John D. Ament wrote: >> ...Its unfair for us to put some stake in the ground expecting podlings to >> match up 100% on the questions. Many of the questions are subjective - is >> the code easy to discover? respond to bug reports in a timely manner?... > > Ok, I think I understand your reluctance now: you don't want us to set > a gate for graduating podlings that many TLPs might not pass. > > I agree with that, and although I'm a strong supporter of the Maturity > Model (having initiated it that's understandable ;-) I'm totally ok > with podlings graduating without fullfilling all of its requirements. > > In my view the model is: > > 1) A good tool to help discover areas that podlings might have neglected. > > 2) A good tool to help podlings look at where they stand, and what > they might still need to improve after graduation. > > 3) A good tool to express our ideal way of doing things, in a concise > way, and evolve that definition over time. > > Based on this I will continue to push for podlings to come to > graduation with a self-assessment based on that model. > > OTOH I'm fine with us clarifying that it's not a requirement. Proposal: It should be a *requirement* for the podling to self-document their maturity model answers in the [DISCUSS] thread before IPMC graduation vote. The requirement is having done it, not passing it. It's *very* helpful to have podlings consider their growth using some form of structured and consistent criteria, so IPMC (and board) can consider how different podlings see themselves compared to past podling history. It doesn't mean every podling has to say "Yes 100%" to every question, just that they've considered each point and can describe their situation there if not. I'd expect plenty of podlings would have some missing or "we're not completely here" on some points, but still be healthy and well-self-governing communities ready to graduate. -- - Shane https://www.apache.org/foundation/marks/resources - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Podlings & Apache Project Maturity Model (was RE: [DISCUSS] Graduate Apache RocketMQ from podling to TLP)
Hi John, On Fri, Aug 25, 2017 at 9:11 PM, John D. Ament wrote: > ...Its unfair for us to put some stake in the ground expecting podlings to > match up 100% on the questions. Many of the questions are subjective - is > the code easy to discover? respond to bug reports in a timely manner?... Ok, I think I understand your reluctance now: you don't want us to set a gate for graduating podlings that many TLPs might not pass. I agree with that, and although I'm a strong supporter of the Maturity Model (having initiated it that's understandable ;-) I'm totally ok with podlings graduating without fullfilling all of its requirements. In my view the model is: 1) A good tool to help discover areas that podlings might have neglected. 2) A good tool to help podlings look at where they stand, and what they might still need to improve after graduation. 3) A good tool to express our ideal way of doing things, in a concise way, and evolve that definition over time. Based on this I will continue to push for podlings to come to graduation with a self-assessment based on that model. OTOH I'm fine with us clarifying that it's not a requirement. -Bertrand - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Podlings & Apache Project Maturity Model (was RE: [DISCUSS] Graduate Apache RocketMQ from podling to TLP)
Hi - I think that the model is a good measure and the development of it was an excellent example of the group dynamic in using a wiki. It is a list of best practices. If a podling goes through the process then we ought to treat it as a fair effort. I think that it would be worth applying a similar wiki based discussion to review the state of Podling intake and graduation check lists. I know your working on that and I would like to help when we bring Daffodil in. You bring up security reporting for Hadoop. A discussion of security requirements with the security team should be done, but not here because the examples I have are private. The problem is that (P)PMC and PMC need to monitor security issues with releases but under the current plan they often have only a few PMC members paying attention. This can lead to trouble with PMC oversight issues. The ASF gives projects substantial freedom, but in return there are norms around (in no particular order): - Foundation links. - Branding - Fundraising - Legal - Security - Infra - Community This is a lot to absorb and learn. Regards, Dave > On Aug 25, 2017, at 12:11 PM, John D. Ament wrote: > > (changing subjects to avoid confusion in RocketMQ's discussion) > > I've been pretty explicit about my disdain in the past over the use of the > Apache Project Maturity Model. The model describes an ideal world that all > projects should strive for, but I would be surprised if many projects > passed it. > > Its unfair for us to put some stake in the ground expecting podlings to > match up 100% on the questions. Many of the questions are subjective - is > the code easy to discover? respond to bug reports in a timely manner? > > My take is that if a podling can answer 1 question per section correctly, > and there's some validity to the answer (e.g. the IN section requires a > polygraph test) then they're on their way. For instance, figuring out how > to report a security issue around Apache Hadoop leads me to vendor websites > first, the first apache.org match is on the second page. This creates > violations in the CO, QU, and IN categories. > > John > > On Thu, Aug 24, 2017 at 3:30 PM Bertrand Delacretaz < > bdelacre...@codeconsult.ch> wrote: > >> On Thu, Aug 24, 2017 at 1:06 PM, John D. Ament >> wrote: >>> ...please understand that the Apache Maturity Model is something that >>> helps the com dev team evaluate TLPs against. Its relevance to a >>> graduating podling is extremely small... >> >> FWIW, I disagree...I think the maturity model is a great tool to help >> discover areas that podlings might have neglected in their work >> towards graduation. >> >> It's not THE single tool to evaluate TLP readiness, but I wouldn't >> qualify its relevance as "extremely small". >> >> (John - maybe we agree on the core, but I just reread the model and love >> it ;-) >> >> -Bertrand >> >> - >> To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org >> For additional commands, e-mail: general-h...@incubator.apache.org >> >> signature.asc Description: Message signed with OpenPGP
Podlings & Apache Project Maturity Model (was RE: [DISCUSS] Graduate Apache RocketMQ from podling to TLP)
(changing subjects to avoid confusion in RocketMQ's discussion) I've been pretty explicit about my disdain in the past over the use of the Apache Project Maturity Model. The model describes an ideal world that all projects should strive for, but I would be surprised if many projects passed it. Its unfair for us to put some stake in the ground expecting podlings to match up 100% on the questions. Many of the questions are subjective - is the code easy to discover? respond to bug reports in a timely manner? My take is that if a podling can answer 1 question per section correctly, and there's some validity to the answer (e.g. the IN section requires a polygraph test) then they're on their way. For instance, figuring out how to report a security issue around Apache Hadoop leads me to vendor websites first, the first apache.org match is on the second page. This creates violations in the CO, QU, and IN categories. John On Thu, Aug 24, 2017 at 3:30 PM Bertrand Delacretaz < bdelacre...@codeconsult.ch> wrote: > On Thu, Aug 24, 2017 at 1:06 PM, John D. Ament > wrote: > > ...please understand that the Apache Maturity Model is something that > > helps the com dev team evaluate TLPs against. Its relevance to a > > graduating podling is extremely small... > > FWIW, I disagree...I think the maturity model is a great tool to help > discover areas that podlings might have neglected in their work > towards graduation. > > It's not THE single tool to evaluate TLP readiness, but I wouldn't > qualify its relevance as "extremely small". > > (John - maybe we agree on the core, but I just reread the model and love > it ;-) > > -Bertrand > > - > To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org > For additional commands, e-mail: general-h...@incubator.apache.org > >