Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
I should have said I don't like the idea of the board receiving reports for podlings that need assistance. It already does. Its not the reporting that's a problem, its the support that's needed in a small number of cases. I'll expand on that in Chris' thread. I'll note that in this thread I answered the question of who Stratos should report to with the board, but I'll also note I don't expect the board to provide mentoring. That is a key difference between what I am proposing for pTLP and the original deconstruction proposal. Sent from a mobile device, please excuse mistakes and brevity On 15 Jun 2013 05:05, Greg Stein gst...@gmail.com wrote: On Fri, Jun 14, 2013 at 6:26 PM, Ross Gardler rgard...@opendirective.com wrote: ... Now, truth be told, I don't like the pTLP reporting to the board idea. I see no problem whatsoever with the suggested pTLP reporting. Let me throw out a hypothetical counterpoint here... The Incubator gathers reports from all of its podlings. It reviews them, discusses some aspects with those podlings, and then it files a report with the ASF Board. Three paragraphs stating, Hey. No issues. Everything is going great. Community is good. Legal is good. kthxbai. Would that fly with the ASF Board? Not a chance. The simple fact is that the Board *does* want to see reports from podlings. Those podlings will (hopefully) become part of the Foundation one day. The Board is *keenly* interested in what is going on, and how those podlings are doing. If you suggest a model of a total black box. Where *no* podling information escapes from the Incubator to the Board. And then one day... *poof!!* ... a graduation resolution appears before the Board. Do you honestly think the Board would just sign off on that? Again: not a chance. What this really means is: the Board wants to review podlings' progress and operations. I don't see how it can be argued any other way. So if that is true, then why does the IPMC need to be a middle man? Why not provide those reports from the podling directly to the Board? And why not get the podling directly engaged with the actual operation of the Foundation? About how to report to the Board? About shepherds, watching for commentary in the agenda, about committing to that agenda!, and about paying attention to board@ and its operations. If we want to teach new communities about how the ASF works, then why the artificial operation of the Incubator? Why not place them directly in contact with the *real* ASF? By all measures, Apache Subversion would have been a pTLP when it arrived at the Incubator. But we integrated very well into the ASF because there were Members, Directors, and other long-term Apache people who could answer huh? what is a PMC Member? how does that map to our 'full committer' status? what are these reports? ... and more. The close attention, and the direct integration with the Foundation, worked as well as you could expect. The Incubator did not provide much value, beyond what the extent Members were already providing (recall that we easily had a half-dozen at the time; I don't know the count offhand, but it was well past any normal podling). The Incubator may not provide value to certain projects that reach the pTLP bar (again: some thumbs-up definition of that is needed!), but it is *very* much required for projects/communities that are not as familiar with how we like to do things here. In this concrete case of Stratos, I personally (and as a voting Director) have every confidence in trying the pTLP approach. I outlined some areas that I believe the Board needs before accepting a pTLP, and so I'm looking forward to this experiment. I think it will turn out well. I do think we may be setting up some communities for anger, when the Board chooses to *not* grant pTLP status and refers the community to the Incubator. I really don't have a good answer there, especially around the future/obvious direction of pTLP is only for the Old Boys Club and other insiders. Sigh. Can't be helped, I think. Anyhow. To the original point: pTLP reporting to the Board is practically speaking a no-brainer. Podlings generally report direct to the Board today, minus some intermediary stuff. - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
That first sentence still doesn't parse, sorry ... I should have said I don't like the idea of the board taking responsibility. I have no problem with it receiving reports directly. Sent from a mobile device, please excuse mistakes and brevity On 15 Jun 2013 07:55, Ross Gardler rgard...@opendirective.com wrote: I should have said I don't like the idea of the board receiving reports for podlings that need assistance. It already does. Its not the reporting that's a problem, its the support that's needed in a small number of cases. I'll expand on that in Chris' thread. I'll note that in this thread I answered the question of who Stratos should report to with the board, but I'll also note I don't expect the board to provide mentoring. That is a key difference between what I am proposing for pTLP and the original deconstruction proposal. Sent from a mobile device, please excuse mistakes and brevity On 15 Jun 2013 05:05, Greg Stein gst...@gmail.com wrote: On Fri, Jun 14, 2013 at 6:26 PM, Ross Gardler rgard...@opendirective.com wrote: ... Now, truth be told, I don't like the pTLP reporting to the board idea. I see no problem whatsoever with the suggested pTLP reporting. Let me throw out a hypothetical counterpoint here... The Incubator gathers reports from all of its podlings. It reviews them, discusses some aspects with those podlings, and then it files a report with the ASF Board. Three paragraphs stating, Hey. No issues. Everything is going great. Community is good. Legal is good. kthxbai. Would that fly with the ASF Board? Not a chance. The simple fact is that the Board *does* want to see reports from podlings. Those podlings will (hopefully) become part of the Foundation one day. The Board is *keenly* interested in what is going on, and how those podlings are doing. If you suggest a model of a total black box. Where *no* podling information escapes from the Incubator to the Board. And then one day... *poof!!* ... a graduation resolution appears before the Board. Do you honestly think the Board would just sign off on that? Again: not a chance. What this really means is: the Board wants to review podlings' progress and operations. I don't see how it can be argued any other way. So if that is true, then why does the IPMC need to be a middle man? Why not provide those reports from the podling directly to the Board? And why not get the podling directly engaged with the actual operation of the Foundation? About how to report to the Board? About shepherds, watching for commentary in the agenda, about committing to that agenda!, and about paying attention to board@ and its operations. If we want to teach new communities about how the ASF works, then why the artificial operation of the Incubator? Why not place them directly in contact with the *real* ASF? By all measures, Apache Subversion would have been a pTLP when it arrived at the Incubator. But we integrated very well into the ASF because there were Members, Directors, and other long-term Apache people who could answer huh? what is a PMC Member? how does that map to our 'full committer' status? what are these reports? ... and more. The close attention, and the direct integration with the Foundation, worked as well as you could expect. The Incubator did not provide much value, beyond what the extent Members were already providing (recall that we easily had a half-dozen at the time; I don't know the count offhand, but it was well past any normal podling). The Incubator may not provide value to certain projects that reach the pTLP bar (again: some thumbs-up definition of that is needed!), but it is *very* much required for projects/communities that are not as familiar with how we like to do things here. In this concrete case of Stratos, I personally (and as a voting Director) have every confidence in trying the pTLP approach. I outlined some areas that I believe the Board needs before accepting a pTLP, and so I'm looking forward to this experiment. I think it will turn out well. I do think we may be setting up some communities for anger, when the Board chooses to *not* grant pTLP status and refers the community to the Incubator. I really don't have a good answer there, especially around the future/obvious direction of pTLP is only for the Old Boys Club and other insiders. Sigh. Can't be helped, I think. Anyhow. To the original point: pTLP reporting to the Board is practically speaking a no-brainer. Podlings generally report direct to the Board today, minus some intermediary stuff. - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
The Board is always the responsible party, but in the sense that you mean responsibility in finding a fix, then I fully agree. IMO, if a pTLP gets into the weeds, then the Board will just say fix yourself within six months, or we dismantle you. Cheers, -g On Jun 15, 2013 2:58 AM, Ross Gardler rgard...@opendirective.com wrote: That first sentence still doesn't parse, sorry ... I should have said I don't like the idea of the board taking responsibility. I have no problem with it receiving reports directly. Sent from a mobile device, please excuse mistakes and brevity On 15 Jun 2013 07:55, Ross Gardler rgard...@opendirective.com wrote: I should have said I don't like the idea of the board receiving reports for podlings that need assistance. It already does. Its not the reporting that's a problem, its the support that's needed in a small number of cases. I'll expand on that in Chris' thread. I'll note that in this thread I answered the question of who Stratos should report to with the board, but I'll also note I don't expect the board to provide mentoring. That is a key difference between what I am proposing for pTLP and the original deconstruction proposal. Sent from a mobile device, please excuse mistakes and brevity On 15 Jun 2013 05:05, Greg Stein gst...@gmail.com wrote: On Fri, Jun 14, 2013 at 6:26 PM, Ross Gardler rgard...@opendirective.com wrote: ... Now, truth be told, I don't like the pTLP reporting to the board idea. I see no problem whatsoever with the suggested pTLP reporting. Let me throw out a hypothetical counterpoint here... The Incubator gathers reports from all of its podlings. It reviews them, discusses some aspects with those podlings, and then it files a report with the ASF Board. Three paragraphs stating, Hey. No issues. Everything is going great. Community is good. Legal is good. kthxbai. Would that fly with the ASF Board? Not a chance. The simple fact is that the Board *does* want to see reports from podlings. Those podlings will (hopefully) become part of the Foundation one day. The Board is *keenly* interested in what is going on, and how those podlings are doing. If you suggest a model of a total black box. Where *no* podling information escapes from the Incubator to the Board. And then one day... *poof!!* ... a graduation resolution appears before the Board. Do you honestly think the Board would just sign off on that? Again: not a chance. What this really means is: the Board wants to review podlings' progress and operations. I don't see how it can be argued any other way. So if that is true, then why does the IPMC need to be a middle man? Why not provide those reports from the podling directly to the Board? And why not get the podling directly engaged with the actual operation of the Foundation? About how to report to the Board? About shepherds, watching for commentary in the agenda, about committing to that agenda!, and about paying attention to board@ and its operations. If we want to teach new communities about how the ASF works, then why the artificial operation of the Incubator? Why not place them directly in contact with the *real* ASF? By all measures, Apache Subversion would have been a pTLP when it arrived at the Incubator. But we integrated very well into the ASF because there were Members, Directors, and other long-term Apache people who could answer huh? what is a PMC Member? how does that map to our 'full committer' status? what are these reports? ... and more. The close attention, and the direct integration with the Foundation, worked as well as you could expect. The Incubator did not provide much value, beyond what the extent Members were already providing (recall that we easily had a half-dozen at the time; I don't know the count offhand, but it was well past any normal podling). The Incubator may not provide value to certain projects that reach the pTLP bar (again: some thumbs-up definition of that is needed!), but it is *very* much required for projects/communities that are not as familiar with how we like to do things here. In this concrete case of Stratos, I personally (and as a voting Director) have every confidence in trying the pTLP approach. I outlined some areas that I believe the Board needs before accepting a pTLP, and so I'm looking forward to this experiment. I think it will turn out well. I do think we may be setting up some communities for anger, when the Board chooses to *not* grant pTLP status and refers the community to the Incubator. I really don't have a good answer there, especially around the future/obvious direction of pTLP is only for the Old Boys Club and other insiders. Sigh. Can't be helped, I think. Anyhow. To the original point: pTLP reporting to the Board is practically speaking a no-brainer. Podlings generally report direct to the Board today, minus some
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
I'm also +1 (and excited!) on trying out this as a probationary TLPs, and with doing that using the approaches outlined by Ross and others in other emails on this thread (which is basically having a vote now to accept this as a podling so we can get started and then working up a probationary TLP proposal for it to submit to the board meeting). I also commit that as mentor i'll help try to make that work well while at the same time provide oversight so that any issue that might arise do get reported. ...ant On Thu, Jun 13, 2013 at 3:12 AM, Ross Gardler rgard...@opendirective.comwrote: So here's a thought... There have been many discussions about different ways to incubate projects. One of the most radical ideas is to dismantle the incubator and replace the podling concept with probationary TLPs reporting to the board. As readers of this list will know I do not support the idea of dismantling the IPMC. I believe it does a great job that is not easily replaced by a board of nine directors. However, I have always acknowledged that the idea has merit under a certain set of circumstances. For me those circumstances are present in the Apache Stratos proposal. That is there are sufficient mentors and initial committers who are ASF Members that we can be reasonably certain that this project will succeed here at the ASF. I would therefore like to propose that we use Apache Stratos as a test case for the probationary TLP idea. I've already talked to Chris (who is driving the deconstruct the IPMC case) and Ant (who is less keen on dismantling the IPMC but wants to see how a probationary TLP model will play out). Both have agreed to help with this experiment if the IPMC and the Board wish it to proceed. I have not, however, discussed it with all the initial comitters or even mentors - I'm expecting them to speak up now. For my part my intention is to get the project set-up and then dissolve into the background. I do not intend to monitor the project on a day-to-day basis. However, I do promise to help pick up the pieces if the experiment should go horribly wrong. Of course running a single experiment will only allow us to define the incubation process for probationary TLPs, It is not going to solve all the problems Chris sees in the IPMC. However it will give us an opportunity to define the process, ask the board to approve this process and thus lay the foundations for other projects wishing to follow this path. So, what do you think? Ross On 11 June 2013 10:10, Ross Gardler rgard...@opendirective.com wrote: It's with great pleasure that I invite the IPMC to review a new proposal [1] for the Apache Incubator. Please let us know if you have any questions or comments - as you will see there are plenty of people on the initial commit list ready and willing to answer your questions. I copy the full text of the proposal for your convenience: = Stratos - A PaaS Framework = == Abstract == Stratos will be a polyglot [[http://www.gartner.com/it-glossary/platform-as-a-service-paas|PaaS]] framework, providing developers a cloud-based environment for developing, testing, and running scalable applications, and IT providers high utilization rates, automated resource management, and platform-wide insight including monitoring and billing. == Proposal == The Stratos PaaS framework will encompass four layers: 1. An [[ http://www.gartner.com/it-glossary/infrastructure-as-a-service-iaas/|IaaS]]-agnostic layer that can interface with a wide variety of IaaS systems to provide elastic resources, and for multiple IaaS infrastructures to be automated at one time (hybrid clouds.) 2. A PaaS Controller with a cloud controller that automates and monitors IaaS runtime interactions, distributes artifacts to the underlying runtimes, deploys workloads, directs runtime traffic to the right runtimes using a tenant-aware elastic load balancer, and provides a portal for monitoring and provisioning of tenants on the system. 3. Foundational Services including security, logging, messaging, registry, storage (relational, file, and noSQL), task management, and billing. Foundational services will be loosely-coupled to allow swapping in alternate foundational services. 4. A Cartridge Architecture allowing frameworks, servers, and other runtimes to participate in the advantages of the system. The Cartridge Architecture must support multi-tenant workloads, and provide for various levels of tenant isolation and policy-based control over provisioning. Together these layers offer a foundational layer upon which applications and middleware frameworks can be deployed to speed time-to-market and simplify the development of scalable applications, as well as provide a high level of resource sharing and centralized management that can deliver lowest resource, infrastructure, and management costs. == Background == The Stratos
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
FWIW: IMO the Incubator still serves a crucial role in the social and legal aspects of *creating* an Apache project and community. The Probationary TLP opens us up to more risk for what I can see as no real benefit or reduction in effort or resources. Also, the unwash public is aware of the distinction between ASF Incubator projects and real PMCs and there is the understanding that Incubator podlings are still finding their way, and that the quality of their release and their community structure may not be on the same level as official TLPs. This distinction will now be lost. Again, I see risk with no real benefit or reason. Some podlings don't make it, and that's a shame, but it's a relatively painless situation. When a TLP, probationary or not, doesn't make, it's much more painful. And because it's a TLP, there is more incentive to keep on trying to make it work and it forces the board to be involved. The board is a hammer, not a scalpel. So I am +1 for Stratos as a regular podling, but -1 (NOT a veto) as a probationary TLP (since I am -1 for the concept in general). - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
Speaking as a Director, the Board will need a *definition* of probation. This is more than just a wiki page. I believe it needs to be a page laid down in www.a.o/dev/ that defines the constraints laid down upon a pTLP (I really like Ross' acronym there!). These differences/constraints should effectively match those we apply to podlings: release constraints, disclosures, etc. Next up will be a description/discussion of whether oversight is maintained with the move from IPMC to Board reporting (this is probably easy). The key point here is an argument on whether losing the IPMC oversight impacts the podling/pTLP, the Board, or the Foundation. There probably needs to be some kind of rough metric on what kind of podling makeup or proposal that could reasonably pass muster with the Board to become a pTLP. Frankly, the Board is going to be *very* subjective on what groups could become a pTLP. I believe there should be some kind of prose somewhere which states that a pTLP is going to be rare/subjective, and that $conditions are needed before even considering a proposal to the Board. Personally, I am hugely supportive of the pTLP concept, and am happy to see a candidate. Cheers, -g On Fri, Jun 14, 2013 at 4:23 AM, ant elder ant.el...@gmail.com wrote: I'm also +1 (and excited!) on trying out this as a probationary TLPs, and with doing that using the approaches outlined by Ross and others in other emails on this thread (which is basically having a vote now to accept this as a podling so we can get started and then working up a probationary TLP proposal for it to submit to the board meeting). I also commit that as mentor i'll help try to make that work well while at the same time provide oversight so that any issue that might arise do get reported. ...ant On Thu, Jun 13, 2013 at 3:12 AM, Ross Gardler rgard...@opendirective.comwrote: So here's a thought... There have been many discussions about different ways to incubate projects. One of the most radical ideas is to dismantle the incubator and replace the podling concept with probationary TLPs reporting to the board. As readers of this list will know I do not support the idea of dismantling the IPMC. I believe it does a great job that is not easily replaced by a board of nine directors. However, I have always acknowledged that the idea has merit under a certain set of circumstances. For me those circumstances are present in the Apache Stratos proposal. That is there are sufficient mentors and initial committers who are ASF Members that we can be reasonably certain that this project will succeed here at the ASF. I would therefore like to propose that we use Apache Stratos as a test case for the probationary TLP idea. I've already talked to Chris (who is driving the deconstruct the IPMC case) and Ant (who is less keen on dismantling the IPMC but wants to see how a probationary TLP model will play out). Both have agreed to help with this experiment if the IPMC and the Board wish it to proceed. I have not, however, discussed it with all the initial comitters or even mentors - I'm expecting them to speak up now. For my part my intention is to get the project set-up and then dissolve into the background. I do not intend to monitor the project on a day-to-day basis. However, I do promise to help pick up the pieces if the experiment should go horribly wrong. Of course running a single experiment will only allow us to define the incubation process for probationary TLPs, It is not going to solve all the problems Chris sees in the IPMC. However it will give us an opportunity to define the process, ask the board to approve this process and thus lay the foundations for other projects wishing to follow this path. So, what do you think? Ross On 11 June 2013 10:10, Ross Gardler rgard...@opendirective.com wrote: It's with great pleasure that I invite the IPMC to review a new proposal [1] for the Apache Incubator. Please let us know if you have any questions or comments - as you will see there are plenty of people on the initial commit list ready and willing to answer your questions. I copy the full text of the proposal for your convenience: = Stratos - A PaaS Framework = == Abstract == Stratos will be a polyglot [[http://www.gartner.com/it-glossary/platform-as-a-service-paas|PaaS]] framework, providing developers a cloud-based environment for developing, testing, and running scalable applications, and IT providers high utilization rates, automated resource management, and platform-wide insight including monitoring and billing. == Proposal == The Stratos PaaS framework will encompass four layers: 1. An [[ http://www.gartner.com/it-glossary/infrastructure-as-a-service-iaas/|IaaS]]-agnostic layer that can interface with a wide variety of IaaS systems to provide elastic resources, and for multiple IaaS infrastructures to be automated at one time (hybrid clouds.) 2. A PaaS
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
Thanks for your comments Jim. You will see from the archives that I share most of your concerns about probationary TLPs. However a number of IPMC members have argued strongly for the concept. I do believe there is merit in experimenting with different models of incubation and I do believe the probationary TLP has idea has some merit hiding under the covers. For me the problem is that it is presented as part of a larger dismantling the IPMC case. I do not believe the idea of dismantling the IPMC is a good one, but for most of us the dislike of this argument means the pTLP idea is not being considered. As you will have seen my proposal is to make the project a podling first then work with the IPMC to define the pTLP process. My intention is for this to address as many of the proposed solutions to issues collected on the IPMC wiki [1] as possible. One of those solutions is the pTLP idea. If we take the extreme view (as presented in Chris' original proposal) this means the pTLP will report to board. I intend to bring the proposal from the IPMC to the board for feedback (at least two directors have indicated value in the proposal when acting as IPMC members). Should the board reject this proposal we will have clearly defined a nice clean incubation process implementing many solutions offered in the wiki. We will see Stratos graduating in good healthy time as a result. It might help if I'm explicit about the benefit that I see in the pTLP proposal. For me I see that it addresses the common problem of mentor atrophy (Issue 01 on the wiki). For it to truly work we need podling members to have binding votes. The pTLP idea is one route to doing this (vote the podling members into the pTLP PMC) another way to do it is to vote them into the IPMC. Benson tried, for some time to address this but was caught in the cross-hairs of entrenched positions within the IPMC. My intention with this experiment is to bring both sides into a common place to work together on a solution rather than to spend all day throwing emails at this list. As you can see, regardless of the outcome, the IPMC will have gained clarity around one (vigorously supported) proposal which has been made 2-3 times over the last couple of years but has never been supported by action to test it in practice. This is an opportunity to test it. If the Board rejects the pTLP idea then we have to think about another way to bring the benefits of the proposal into the incubation process. I have my ideas, as do others, in the past they have always been shouted down. Once this experiment is complete and Stratos is a TLP I will turn my attention to testing the alternative solutions. On the other hand if it turns out this experiment finds a pTLP model that works we can just get on with implementing it. Either way we can get out of this period of stagnation we are (caused by Issue 03 on the wiki). Thank you for raising your concerns. Thank you for making your -1 clear and also for indicating it is not a veto (of course it is a good indicator of how you are likely to vote as a Director, this is very useful - thank you). The experiment can continue but we need to ensure that we address the concerns raised by Jim if this is ever to fly as a repeatable process. I propose we ensure that whatever we put before the board in July has an alternative course of action that does not disenfranchise the IPMC and does not push responsibility to the board. This way we can ensure that the Stratos project still gets a smooth passage through the incubation process. Ross [1] http://wiki.apache.org/incubator/IncubatorIssues2013 On 14 June 2013 13:45, Jim Jagielski j...@jagunet.com wrote: FWIW: IMO the Incubator still serves a crucial role in the social and legal aspects of *creating* an Apache project and community. The Probationary TLP opens us up to more risk for what I can see as no real benefit or reduction in effort or resources. Also, the unwash public is aware of the distinction between ASF Incubator projects and real PMCs and there is the understanding that Incubator podlings are still finding their way, and that the quality of their release and their community structure may not be on the same level as official TLPs. This distinction will now be lost. Again, I see risk with no real benefit or reason. Some podlings don't make it, and that's a shame, but it's a relatively painless situation. When a TLP, probationary or not, doesn't make, it's much more painful. And because it's a TLP, there is more incentive to keep on trying to make it work and it forces the board to be involved. The board is a hammer, not a scalpel. So I am +1 for Stratos as a regular podling, but -1 (NOT a veto) as a probationary TLP (since I am -1 for the concept in general). - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On Fri, Jun 14, 2013 at 9:31 AM, Ross Gardler rgard...@opendirective.com wrote: ... I propose we ensure that whatever we put before the board in July has an alternative course of action that does not disenfranchise the IPMC It is perfectly fine to disenfranchise the IPMC. It does not get a blank pass. It must provide value. Competing ideas are allowed and encouraged. The IPMC does not have a monopoly; it is subject to the Board, which may decide other models may be workable; thus, the IPMC needs to accept that fact. Cheers, -g - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
Thanks Greg, Your pointers (as a director) are useful. As an IPMC member who supports the concept of pTLP I hope you will assist, either during discussion or during review, in ensuring we have sufficient clarity around these items prior to submission to the board (yes I realise that's what this mail is - thanks for the good start ;-) Ross On 14 June 2013 14:32, Greg Stein gst...@gmail.com wrote: Speaking as a Director, the Board will need a *definition* of probation. This is more than just a wiki page. I believe it needs to be a page laid down in www.a.o/dev/ that defines the constraints laid down upon a pTLP (I really like Ross' acronym there!). These differences/constraints should effectively match those we apply to podlings: release constraints, disclosures, etc. Next up will be a description/discussion of whether oversight is maintained with the move from IPMC to Board reporting (this is probably easy). The key point here is an argument on whether losing the IPMC oversight impacts the podling/pTLP, the Board, or the Foundation. There probably needs to be some kind of rough metric on what kind of podling makeup or proposal that could reasonably pass muster with the Board to become a pTLP. Frankly, the Board is going to be *very* subjective on what groups could become a pTLP. I believe there should be some kind of prose somewhere which states that a pTLP is going to be rare/subjective, and that $conditions are needed before even considering a proposal to the Board. Personally, I am hugely supportive of the pTLP concept, and am happy to see a candidate. Cheers, -g On Fri, Jun 14, 2013 at 4:23 AM, ant elder ant.el...@gmail.com wrote: I'm also +1 (and excited!) on trying out this as a probationary TLPs, and with doing that using the approaches outlined by Ross and others in other emails on this thread (which is basically having a vote now to accept this as a podling so we can get started and then working up a probationary TLP proposal for it to submit to the board meeting). I also commit that as mentor i'll help try to make that work well while at the same time provide oversight so that any issue that might arise do get reported. ...ant On Thu, Jun 13, 2013 at 3:12 AM, Ross Gardler rgard...@opendirective.comwrote: So here's a thought... There have been many discussions about different ways to incubate projects. One of the most radical ideas is to dismantle the incubator and replace the podling concept with probationary TLPs reporting to the board. As readers of this list will know I do not support the idea of dismantling the IPMC. I believe it does a great job that is not easily replaced by a board of nine directors. However, I have always acknowledged that the idea has merit under a certain set of circumstances. For me those circumstances are present in the Apache Stratos proposal. That is there are sufficient mentors and initial committers who are ASF Members that we can be reasonably certain that this project will succeed here at the ASF. I would therefore like to propose that we use Apache Stratos as a test case for the probationary TLP idea. I've already talked to Chris (who is driving the deconstruct the IPMC case) and Ant (who is less keen on dismantling the IPMC but wants to see how a probationary TLP model will play out). Both have agreed to help with this experiment if the IPMC and the Board wish it to proceed. I have not, however, discussed it with all the initial comitters or even mentors - I'm expecting them to speak up now. For my part my intention is to get the project set-up and then dissolve into the background. I do not intend to monitor the project on a day-to-day basis. However, I do promise to help pick up the pieces if the experiment should go horribly wrong. Of course running a single experiment will only allow us to define the incubation process for probationary TLPs, It is not going to solve all the problems Chris sees in the IPMC. However it will give us an opportunity to define the process, ask the board to approve this process and thus lay the foundations for other projects wishing to follow this path. So, what do you think? Ross On 11 June 2013 10:10, Ross Gardler rgard...@opendirective.com wrote: It's with great pleasure that I invite the IPMC to review a new proposal [1] for the Apache Incubator. Please let us know if you have any questions or comments - as you will see there are plenty of people on the initial commit list ready and willing to answer your questions. I copy the full text of the proposal for your convenience: = Stratos - A PaaS Framework = == Abstract == Stratos will be a polyglot [[http://www.gartner.com/it-glossary/platform-as-a-service-paas|PaaS]] framework, providing developers a cloud-based environment for developing, testing, and running scalable applications, and IT providers high utilization rates, automated resource management, and
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On Jun 14, 2013, at 9:31 AM, Ross Gardler rgard...@opendirective.com wrote: Thanks for your comments Jim. You will see from the archives that I share most of your concerns about probationary TLPs. However a number of IPMC members have argued strongly for the concept. To be clear, the only info that seems official about pTLPs is the Wiki page proposal: http://wiki.apache.org/incubator/IncubatorDeconstructionProposal IMO, it simply doesn't provide enough meat, details and rationale for me to be able to commit to it enough; there's a difference between the concept and the implementation of that concept. - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On 14 June 2013 14:40, Greg Stein gst...@gmail.com wrote: On Fri, Jun 14, 2013 at 9:31 AM, Ross Gardler rgard...@opendirective.com wrote: ... I propose we ensure that whatever we put before the board in July has an alternative course of action that does not disenfranchise the IPMC It is perfectly fine to disenfranchise the IPMC. It does not get a blank pass. It must provide value. Competing ideas are allowed and encouraged. The IPMC does not have a monopoly; it is subject to the Board, which may decide other models may be workable; thus, the IPMC needs to accept that fact. Yep. I said *alternative* What I want to do here is provide a resolution and supporting materials for the pTLP concept. My concern is that if the board says no (and it might) we should have an alternative that will allow the IPMC to pull value from the pTLP idea. With these two things together the board can evaluate the merit of both competing ideas. What I do not want to happen is we put all this effort into proposing the pTLP idea, then have the board reject it and the Stratos project being a month behind as a result. Some aspects of the pTLP idea are dependent on the board saying yes. Ross Cheers, -g - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On 14 June 2013 14:45, Jim Jagielski j...@jagunet.com wrote: On Jun 14, 2013, at 9:31 AM, Ross Gardler rgard...@opendirective.com wrote: Thanks for your comments Jim. You will see from the archives that I share most of your concerns about probationary TLPs. However a number of IPMC members have argued strongly for the concept. To be clear, the only info that seems official about pTLPs is the Wiki page proposal: http://wiki.apache.org/incubator/IncubatorDeconstructionProposal IMO, it simply doesn't provide enough meat, details and rationale for me to be able to commit to it enough; there's a difference between the concept and the implementation of that concept. Agreed (and partially discussed elsewhere in this thread). Greg outlines a minimum he wants to see from the proposal to the board. I agree with his minimum and will be looking to two supporters of the pTLP concept to coordinate delivery of that minimum (I discussed this with both of them prior to make this suggestion, both agreed). Ross - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
Of course, I'm fully supportive of the pTLP concept as I believe it begins to implement parts of the proposal I threw up. BTW, note there aren't a ton of options besides the recent one on the wiki proposed by Bertrand, and added to by Alan, Ant, Ross and a lot of others. Those are concerns/issues with the Incubator. They aren't a proposal of what to do. Someone please point me to something that's been as discussed; worked on; and that contains a responsibility transition matrix. A *lot* of thought has went into this. I'm not just whining throw out the Incubator; it sucks -- I'm saying here's a set of incremental actions that taken as a whole, or individually towards an eventual whole will still allow projects to come into our Foundation, but to do so in a way that's a more natural fit for where we want these projects to end up anyways. I would urge the board to consider: * I've spent a great deal of time discussing the benefits and pitfalls of the approach with people at length for nearly 2 years now. So, no need to rehash those, I request folks to do research in the archives for the threads Incubation yes, Incubator no by Bill Rowe, anything that references the Incubator Deconstruction Proposal, etc. * pTLP are nothing different than what existed before there was an Incubator. Yes we have more projects now. So what. We'll continue to have more projects and those will eventually graduate to TLP, so we'll be in the same place. * This is an incremental step in the deconstruction proposal. We aren't saying implement the whole thing at once -- we are taking some constructive steps here (including what happened RE: discussion on the ComDev side of inheriting the docs). Let's get some data points here, report back, etc., in the vein of Apache with the small incremental, reversible change part. * The board needs to [DISCUSS] this. It's an important issue and I have a few podlings in mind already that would fit the pTLP concept. Kudos to Ross for having the courage to push this forward. Note, Ross was one of the most vocal discussers, seeing both the merit and potential pitfalls of my proposal. In short, if I've got Ross convinced enough to at least try it, then give it a chance. Cheers, Chris ++ Chris Mattmann, Ph.D. Senior Computer Scientist NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA Office: 171-266B, Mailstop: 171-246 Email: chris.a.mattm...@nasa.gov WWW: http://sunset.usc.edu/~mattmann/ ++ Adjunct Assistant Professor, Computer Science Department University of Southern California, Los Angeles, CA 90089 USA ++ -Original Message- From: Ross Gardler rgard...@opendirective.com Reply-To: general@incubator.apache.org general@incubator.apache.org Date: Friday, June 14, 2013 6:48 AM To: general@incubator.apache.org general@incubator.apache.org Subject: Re: [DISCUSS] Accept Stratos as an Apache Incubation Project On 14 June 2013 14:45, Jim Jagielski j...@jagunet.com wrote: On Jun 14, 2013, at 9:31 AM, Ross Gardler rgard...@opendirective.com wrote: Thanks for your comments Jim. You will see from the archives that I share most of your concerns about probationary TLPs. However a number of IPMC members have argued strongly for the concept. To be clear, the only info that seems official about pTLPs is the Wiki page proposal: http://wiki.apache.org/incubator/IncubatorDeconstructionProposal IMO, it simply doesn't provide enough meat, details and rationale for me to be able to commit to it enough; there's a difference between the concept and the implementation of that concept. Agreed (and partially discussed elsewhere in this thread). Greg outlines a minimum he wants to see from the proposal to the board. I agree with his minimum and will be looking to two supporters of the pTLP concept to coordinate delivery of that minimum (I discussed this with both of them prior to make this suggestion, both agreed). Ross - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
One comment that I wish I could express more clearly: On 6/14/2013 10:42 AM, Mattmann, Chris A (398J) wrote: ... * pTLP are nothing different than what existed before there was an Incubator. Yes we have more projects now. So what. We'll continue to have more projects and those will eventually graduate to TLP, so we'll be in the same place. Actually, from the *outside* perspective, it's completely different. The Incubator has podlings, this weird thing that most people don't quite understand, but know that it's weird, so they look at the Incubator site enough to understand Oh, it's kind of at Apache, but it's not an *Apache project*. pTLPs are something that... well, the p is lowercase, so who cares - they're just TLPs which is pretty easy to google to top level project, which means - a-ha! - here's a brand new *Apache project*! And look - they're totally run by SoAndSo Co - what, is Apache now pay to play, allowing an Apache project to work like that? This distinction - as seen in the view of the general computer using public, and especially in the view of those in positions of authority at corporations that use/build/create software - is *very* important to helping to define the Apache brand. It's taken a while to educate enough of the press and other open source watchers to understand what podlings are, and the whole concept not only of probation, but also the concept of incubation: i.e. not just waiting until the new podling does the right kinds of things, but actively helping and guiding the community to learning and embracing The Apache Way of doing things... at which point, they graduate and then become an Apache project. As VP, Brand, I believe this is an important issue. But I often don't know how to explain it, precisely because we all here know what podlings and pTLPs and so on are - what is important to me is the perception of the *rest* of the world who *doesn't* know what they are. Nor does the rest of the world care about the details - what they care about is: is it an Apache project? If so, it's got good community, a safe license, will be around for a while, and I know I can contribute to it safely without being abused by competitors/bigcos/whoever. That's what the Apache brand means. - Shane - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On 14 June 2013 15:42, Mattmann, Chris A (398J) chris.a.mattm...@jpl.nasa.gov wrote: Those are concerns/issues with the Incubator. They aren't a proposal of what to do. For concrete suggestions about how to address the issues in the wiki take a look at the solutions section under each issue. All issue have at least one suggestion, many have more. * pTLP are nothing different than what existed before there was an Incubator. Sorry Chris, I disagree. It is *very* different. Where it isn't so different is in cases where there are plenty of experienced and active mentors (I believe Stratos is one such case which is why I proposed it as a test case). * This is an incremental step in the deconstruction proposal. Not for me it isn't. It's an incremental step towards finding the merits in your proposal for a specific type of project. I have never supported the deconstruction proposal and I trust people (including the board when the IPMC makes its recommendations) are clear on this. My championing of this proposal should not be seen as a championing of the whole deconstruction idea. Note, Ross was one of the most vocal discussers, seeing both the merit and potential pitfalls of my proposal. In short, if I've got Ross convinced enough to at least try it, then give it a chance. The only thing you have convinced me of is that there is no opportunity to find the merits while this pTLP idea is wrapped up in the larger deconstruction proposal. I made this clear in the initial proposal of the idea and I made it clear in my response to Jim. I'm making it clear again here. I want to expose the merits and avoid the pitfalls of your larger proposal. I am *not* adding my weight to your larger proposal. I am merely moving past an incomplete wiki page with practical activity. I will do the same with many of the other proposals that have merit (e.g. Ant has started to build momentum behind his tooling suggestions, |I hope to help there too). Ross Cheers, Chris ++ Chris Mattmann, Ph.D. Senior Computer Scientist NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA Office: 171-266B, Mailstop: 171-246 Email: chris.a.mattm...@nasa.gov WWW: http://sunset.usc.edu/~mattmann/ ++ Adjunct Assistant Professor, Computer Science Department University of Southern California, Los Angeles, CA 90089 USA ++ -Original Message- From: Ross Gardler rgard...@opendirective.com Reply-To: general@incubator.apache.org general@incubator.apache.org Date: Friday, June 14, 2013 6:48 AM To: general@incubator.apache.org general@incubator.apache.org Subject: Re: [DISCUSS] Accept Stratos as an Apache Incubation Project On 14 June 2013 14:45, Jim Jagielski j...@jagunet.com wrote: On Jun 14, 2013, at 9:31 AM, Ross Gardler rgard...@opendirective.com wrote: Thanks for your comments Jim. You will see from the archives that I share most of your concerns about probationary TLPs. However a number of IPMC members have argued strongly for the concept. To be clear, the only info that seems official about pTLPs is the Wiki page proposal: http://wiki.apache.org/incubator/IncubatorDeconstructionProposal IMO, it simply doesn't provide enough meat, details and rationale for me to be able to commit to it enough; there's a difference between the concept and the implementation of that concept. Agreed (and partially discussed elsewhere in this thread). Greg outlines a minimum he wants to see from the proposal to the board. I agree with his minimum and will be looking to two supporters of the pTLP concept to coordinate delivery of that minimum (I discussed this with both of them prior to make this suggestion, both agreed). Ross - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
I'm putting my replies to this in a separate thread, and going to combine responses to Shane and Jim in it too since pTLP is not only related to accepting Stratos as an Apache Incubation project. ++ Chris Mattmann, Ph.D. Senior Computer Scientist NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA Office: 171-266B, Mailstop: 171-246 Email: chris.a.mattm...@nasa.gov WWW: http://sunset.usc.edu/~mattmann/ ++ Adjunct Assistant Professor, Computer Science Department University of Southern California, Los Angeles, CA 90089 USA ++ -Original Message- From: Ross Gardler rgard...@opendirective.com Reply-To: general@incubator.apache.org general@incubator.apache.org Date: Friday, June 14, 2013 8:38 AM To: general@incubator.apache.org general@incubator.apache.org Subject: Re: [DISCUSS] Accept Stratos as an Apache Incubation Project On 14 June 2013 15:42, Mattmann, Chris A (398J) chris.a.mattm...@jpl.nasa.gov wrote: Those are concerns/issues with the Incubator. They aren't a proposal of what to do. For concrete suggestions about how to address the issues in the wiki take a look at the solutions section under each issue. All issue have at least one suggestion, many have more. * pTLP are nothing different than what existed before there was an Incubator. Sorry Chris, I disagree. It is *very* different. Where it isn't so different is in cases where there are plenty of experienced and active mentors (I believe Stratos is one such case which is why I proposed it as a test case). * This is an incremental step in the deconstruction proposal. Not for me it isn't. It's an incremental step towards finding the merits in your proposal for a specific type of project. I have never supported the deconstruction proposal and I trust people (including the board when the IPMC makes its recommendations) are clear on this. My championing of this proposal should not be seen as a championing of the whole deconstruction idea. Note, Ross was one of the most vocal discussers, seeing both the merit and potential pitfalls of my proposal. In short, if I've got Ross convinced enough to at least try it, then give it a chance. The only thing you have convinced me of is that there is no opportunity to find the merits while this pTLP idea is wrapped up in the larger deconstruction proposal. I made this clear in the initial proposal of the idea and I made it clear in my response to Jim. I'm making it clear again here. I want to expose the merits and avoid the pitfalls of your larger proposal. I am *not* adding my weight to your larger proposal. I am merely moving past an incomplete wiki page with practical activity. I will do the same with many of the other proposals that have merit (e.g. Ant has started to build momentum behind his tooling suggestions, |I hope to help there too). Ross Cheers, Chris ++ Chris Mattmann, Ph.D. Senior Computer Scientist NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA Office: 171-266B, Mailstop: 171-246 Email: chris.a.mattm...@nasa.gov WWW: http://sunset.usc.edu/~mattmann/ ++ Adjunct Assistant Professor, Computer Science Department University of Southern California, Los Angeles, CA 90089 USA ++ -Original Message- From: Ross Gardler rgard...@opendirective.com Reply-To: general@incubator.apache.org general@incubator.apache.org Date: Friday, June 14, 2013 6:48 AM To: general@incubator.apache.org general@incubator.apache.org Subject: Re: [DISCUSS] Accept Stratos as an Apache Incubation Project On 14 June 2013 14:45, Jim Jagielski j...@jagunet.com wrote: On Jun 14, 2013, at 9:31 AM, Ross Gardler rgard...@opendirective.com wrote: Thanks for your comments Jim. You will see from the archives that I share most of your concerns about probationary TLPs. However a number of IPMC members have argued strongly for the concept. To be clear, the only info that seems official about pTLPs is the Wiki page proposal: http://wiki.apache.org/incubator/IncubatorDeconstructionProposal IMO, it simply doesn't provide enough meat, details and rationale for me to be able to commit to it enough; there's a difference between the concept and the implementation of that concept. Agreed (and partially discussed elsewhere in this thread). Greg outlines a minimum he wants to see from the proposal to the board. I agree with his minimum and will be looking to two supporters of the pTLP concept to coordinate delivery of that minimum (I discussed this with both of them prior to make this suggestion, both agreed). Ross
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On Fri, Jun 14, 2013 at 5:45 AM, Jim Jagielski j...@jagunet.com wrote: The Probationary TLP opens us up to more risk for what I can see as no real benefit or reduction in effort or resources. I expect the experiment to yield a specific benefit to Stratos: a more intensive incubation experience and a more cultured, better educated PMC on graduation. Making it possible to reward contributors who have demonstrated a thorough understanding of the Apache Way *during incubation*, *as individuals*, will motivate people to learn the Incubator curriculum. These standout individuals will subsequently be motivated to teach their peers by their strong attachment to their project and desire to see it succeed, working alongside IPMC Mentors too bootstrap additional individual contributors in the Apache Way one-by-one. Ideally, the habits they acquire will persist beyond graduation when IPMC members peel away, leaving the top level project with an elevated level of expertise on average when compared with a podling that goes through our standard incubation process. These topics were discussed in the Mentoring individuals as well as projects thread: http://markmail.org/message/5snqt5bwwivzk33h So I am +1 for Stratos as a regular podling, but -1 (NOT a veto) as a probationary TLP (since I am -1 for the concept in general). Like Ross, I also agree with many of your critiques and am not enamored of every detail in the wiki. However, I accept that my foresight is imperfect and that my skepticism may not be warranted in some cases. I'm OK with a bit of wildness because Stratos seems likely to succeed, because there are so many experienced people to catch problems and limit potential damage, and because it's being run as an experiment rather than baked in permanently. Marvin Humphrey - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On Jun 14, 2013, at 10:42 AM, Mattmann, Chris A (398J) chris.a.mattm...@jpl.nasa.gov wrote: * pTLP are nothing different than what existed before there was an Incubator. Yes we have more projects now. So what. We'll continue to have more projects and those will eventually graduate to TLP, so we'll be in the same place. That kind of ignores the reason why we created the Incubator in the 1st place. If what existed before worked, then we wouldn't be discussing an Incubator now :) For me, the main 2 goals of the Incubator are legal clearance and oversight. I doubt the board will be happy with having additional oversight duties related to a pTLP. It's not what the board is for, or why it exists. - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
+1 to all Greg's comments below. Thanks dude. Cheers, Chris ++ Chris Mattmann, Ph.D. Senior Computer Scientist NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA Office: 171-266B, Mailstop: 171-246 Email: chris.a.mattm...@nasa.gov WWW: http://sunset.usc.edu/~mattmann/ ++ Adjunct Assistant Professor, Computer Science Department University of Southern California, Los Angeles, CA 90089 USA ++ -Original Message- From: Greg Stein gst...@gmail.com Reply-To: general@incubator.apache.org general@incubator.apache.org Date: Friday, June 14, 2013 6:32 AM To: general@incubator.apache.org general@incubator.apache.org Subject: Re: [DISCUSS] Accept Stratos as an Apache Incubation Project Speaking as a Director, the Board will need a *definition* of probation. This is more than just a wiki page. I believe it needs to be a page laid down in www.a.o/dev/ that defines the constraints laid down upon a pTLP (I really like Ross' acronym there!). These differences/constraints should effectively match those we apply to podlings: release constraints, disclosures, etc. Next up will be a description/discussion of whether oversight is maintained with the move from IPMC to Board reporting (this is probably easy). The key point here is an argument on whether losing the IPMC oversight impacts the podling/pTLP, the Board, or the Foundation. There probably needs to be some kind of rough metric on what kind of podling makeup or proposal that could reasonably pass muster with the Board to become a pTLP. Frankly, the Board is going to be *very* subjective on what groups could become a pTLP. I believe there should be some kind of prose somewhere which states that a pTLP is going to be rare/subjective, and that $conditions are needed before even considering a proposal to the Board. Personally, I am hugely supportive of the pTLP concept, and am happy to see a candidate. Cheers, -g On Fri, Jun 14, 2013 at 4:23 AM, ant elder ant.el...@gmail.com wrote: I'm also +1 (and excited!) on trying out this as a probationary TLPs, and with doing that using the approaches outlined by Ross and others in other emails on this thread (which is basically having a vote now to accept this as a podling so we can get started and then working up a probationary TLP proposal for it to submit to the board meeting). I also commit that as mentor i'll help try to make that work well while at the same time provide oversight so that any issue that might arise do get reported. ...ant On Thu, Jun 13, 2013 at 3:12 AM, Ross Gardler rgard...@opendirective.comwrote: So here's a thought... There have been many discussions about different ways to incubate projects. One of the most radical ideas is to dismantle the incubator and replace the podling concept with probationary TLPs reporting to the board. As readers of this list will know I do not support the idea of dismantling the IPMC. I believe it does a great job that is not easily replaced by a board of nine directors. However, I have always acknowledged that the idea has merit under a certain set of circumstances. For me those circumstances are present in the Apache Stratos proposal. That is there are sufficient mentors and initial committers who are ASF Members that we can be reasonably certain that this project will succeed here at the ASF. I would therefore like to propose that we use Apache Stratos as a test case for the probationary TLP idea. I've already talked to Chris (who is driving the deconstruct the IPMC case) and Ant (who is less keen on dismantling the IPMC but wants to see how a probationary TLP model will play out). Both have agreed to help with this experiment if the IPMC and the Board wish it to proceed. I have not, however, discussed it with all the initial comitters or even mentors - I'm expecting them to speak up now. For my part my intention is to get the project set-up and then dissolve into the background. I do not intend to monitor the project on a day-to-day basis. However, I do promise to help pick up the pieces if the experiment should go horribly wrong. Of course running a single experiment will only allow us to define the incubation process for probationary TLPs, It is not going to solve all the problems Chris sees in the IPMC. However it will give us an opportunity to define the process, ask the board to approve this process and thus lay the foundations for other projects wishing to follow this path. So, what do you think? Ross On 11 June 2013 10:10, Ross Gardler rgard...@opendirective.com wrote: It's with great pleasure that I invite the IPMC to review a new proposal [1] for the Apache Incubator. Please let us know if you have any questions or comments - as you will see there are plenty of people on the initial commit
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On 14 June 2013 17:19, Mattmann, Chris A (398J) chris.a.mattm...@jpl.nasa.gov wrote: I'm putting my replies to this in a separate thread, and going to combine responses to Shane and Jim in it too since pTLP is not only related to accepting Stratos as an Apache Incubation project. Good move. I'll call the vote on the Stratos podling (i.e. not pTLP now) Ross ++ Chris Mattmann, Ph.D. Senior Computer Scientist NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA Office: 171-266B, Mailstop: 171-246 Email: chris.a.mattm...@nasa.gov WWW: http://sunset.usc.edu/~mattmann/ ++ Adjunct Assistant Professor, Computer Science Department University of Southern California, Los Angeles, CA 90089 USA ++ -Original Message- From: Ross Gardler rgard...@opendirective.com Reply-To: general@incubator.apache.org general@incubator.apache.org Date: Friday, June 14, 2013 8:38 AM To: general@incubator.apache.org general@incubator.apache.org Subject: Re: [DISCUSS] Accept Stratos as an Apache Incubation Project On 14 June 2013 15:42, Mattmann, Chris A (398J) chris.a.mattm...@jpl.nasa.gov wrote: Those are concerns/issues with the Incubator. They aren't a proposal of what to do. For concrete suggestions about how to address the issues in the wiki take a look at the solutions section under each issue. All issue have at least one suggestion, many have more. * pTLP are nothing different than what existed before there was an Incubator. Sorry Chris, I disagree. It is *very* different. Where it isn't so different is in cases where there are plenty of experienced and active mentors (I believe Stratos is one such case which is why I proposed it as a test case). * This is an incremental step in the deconstruction proposal. Not for me it isn't. It's an incremental step towards finding the merits in your proposal for a specific type of project. I have never supported the deconstruction proposal and I trust people (including the board when the IPMC makes its recommendations) are clear on this. My championing of this proposal should not be seen as a championing of the whole deconstruction idea. Note, Ross was one of the most vocal discussers, seeing both the merit and potential pitfalls of my proposal. In short, if I've got Ross convinced enough to at least try it, then give it a chance. The only thing you have convinced me of is that there is no opportunity to find the merits while this pTLP idea is wrapped up in the larger deconstruction proposal. I made this clear in the initial proposal of the idea and I made it clear in my response to Jim. I'm making it clear again here. I want to expose the merits and avoid the pitfalls of your larger proposal. I am *not* adding my weight to your larger proposal. I am merely moving past an incomplete wiki page with practical activity. I will do the same with many of the other proposals that have merit (e.g. Ant has started to build momentum behind his tooling suggestions, |I hope to help there too). Ross Cheers, Chris ++ Chris Mattmann, Ph.D. Senior Computer Scientist NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA Office: 171-266B, Mailstop: 171-246 Email: chris.a.mattm...@nasa.gov WWW: http://sunset.usc.edu/~mattmann/ ++ Adjunct Assistant Professor, Computer Science Department University of Southern California, Los Angeles, CA 90089 USA ++ -Original Message- From: Ross Gardler rgard...@opendirective.com Reply-To: general@incubator.apache.org general@incubator.apache.org Date: Friday, June 14, 2013 6:48 AM To: general@incubator.apache.org general@incubator.apache.org Subject: Re: [DISCUSS] Accept Stratos as an Apache Incubation Project On 14 June 2013 14:45, Jim Jagielski j...@jagunet.com wrote: On Jun 14, 2013, at 9:31 AM, Ross Gardler rgard...@opendirective.com wrote: Thanks for your comments Jim. You will see from the archives that I share most of your concerns about probationary TLPs. However a number of IPMC members have argued strongly for the concept. To be clear, the only info that seems official about pTLPs is the Wiki page proposal: http://wiki.apache.org/incubator/IncubatorDeconstructionProposal IMO, it simply doesn't provide enough meat, details and rationale for me to be able to commit to it enough; there's a difference between the concept and the implementation of that concept. Agreed (and partially discussed elsewhere in this thread). Greg outlines a minimum he wants to see from the proposal to the board. I agree with his minimum and will be looking to two supporters of the pTLP concept
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On 14 June 2013 18:16, Jim Jagielski j...@jagunet.com wrote: On Jun 14, 2013, at 10:42 AM, Mattmann, Chris A (398J) chris.a.mattm...@jpl.nasa.gov wrote: * pTLP are nothing different than what existed before there was an Incubator. Yes we have more projects now. So what. We'll continue to have more projects and those will eventually graduate to TLP, so we'll be in the same place. That kind of ignores the reason why we created the Incubator in the 1st place. If what existed before worked, then we wouldn't be discussing an Incubator now :) For me, the main 2 goals of the Incubator are legal clearance and oversight. I doubt the board will be happy with having additional oversight duties related to a pTLP. It's not what the board is for, or why it exists. Jim, I want to reassure you that I, and others, have expressed this precise concern before, many times. The proposal as it stands in the wiki does not address this issue, neither has it been addressed to my satisfaction in discussion on this list. For me one of the largest problems faced in the IPMC is ISSUE 03 in the wiki, expressed as The IPMC is a large and noisy place where success are not celebrated and problems are quickly amplified due to size and varied approaches of IPMC membership. If podlings never needed to turn to the IPMC for assistance this problem would not exist. However, ISSUE 01 (Some podlings have mostly inactive mentors, and/or a stale list of mentors that doesn't reflect reality) means that some projects need the support of the IPMC. The advantage, that I see, of a pTLP concept is that all incubating projects *must* have three active PMC members - starting with mentors only but moving quickly to include active members of the initial commit list (note this is a subtle difference from Chris' proposal). Doing this bypasses ISSUE 01, at least to some extent, as the pool of individuals able to do do legal clearance and oversight is grown quickly and thus the need to come to the IPMC is quickly reduced. Now, truth be told, I don't like the pTLP reporting to the board idea. I have objected vocally to this on the grounds you provide above. However, I do think that bringing the right committers on podlings into the IPMC so that they have a binding vote is a good idea. I've put this forwards as a proposal but was shouted down by those who perceived this as adding a new layer to the model (since it requires a some kind of subset of the IPMC to actually run affairs). My intention with this experiment is to provide an environment in which we can draw out the benefits hidden in the pTLP idea whilst addressing the concerns that it avoids the oversight and legal clearance. To be honest, I think we already achieved a fair bit in the last couple of days. I feel I have a good understanding of what the proposal needs to look like. However, I promised Ant that I'd get out of his way once the vote was completed so that he can inject his own ideas. I know Chris will also be present. The good news is that I'm in the process of packing my house into a shipping container and my family into a plane ready for an international relocation in a couple of weeks. there will be plenty of space left for these people to think. Ross - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On Fri, Jun 14, 2013 at 6:26 PM, Ross Gardler rgard...@opendirective.com wrote: ... Now, truth be told, I don't like the pTLP reporting to the board idea. I see no problem whatsoever with the suggested pTLP reporting. Let me throw out a hypothetical counterpoint here... The Incubator gathers reports from all of its podlings. It reviews them, discusses some aspects with those podlings, and then it files a report with the ASF Board. Three paragraphs stating, Hey. No issues. Everything is going great. Community is good. Legal is good. kthxbai. Would that fly with the ASF Board? Not a chance. The simple fact is that the Board *does* want to see reports from podlings. Those podlings will (hopefully) become part of the Foundation one day. The Board is *keenly* interested in what is going on, and how those podlings are doing. If you suggest a model of a total black box. Where *no* podling information escapes from the Incubator to the Board. And then one day... *poof!!* ... a graduation resolution appears before the Board. Do you honestly think the Board would just sign off on that? Again: not a chance. What this really means is: the Board wants to review podlings' progress and operations. I don't see how it can be argued any other way. So if that is true, then why does the IPMC need to be a middle man? Why not provide those reports from the podling directly to the Board? And why not get the podling directly engaged with the actual operation of the Foundation? About how to report to the Board? About shepherds, watching for commentary in the agenda, about committing to that agenda!, and about paying attention to board@ and its operations. If we want to teach new communities about how the ASF works, then why the artificial operation of the Incubator? Why not place them directly in contact with the *real* ASF? By all measures, Apache Subversion would have been a pTLP when it arrived at the Incubator. But we integrated very well into the ASF because there were Members, Directors, and other long-term Apache people who could answer huh? what is a PMC Member? how does that map to our 'full committer' status? what are these reports? ... and more. The close attention, and the direct integration with the Foundation, worked as well as you could expect. The Incubator did not provide much value, beyond what the extent Members were already providing (recall that we easily had a half-dozen at the time; I don't know the count offhand, but it was well past any normal podling). The Incubator may not provide value to certain projects that reach the pTLP bar (again: some thumbs-up definition of that is needed!), but it is *very* much required for projects/communities that are not as familiar with how we like to do things here. In this concrete case of Stratos, I personally (and as a voting Director) have every confidence in trying the pTLP approach. I outlined some areas that I believe the Board needs before accepting a pTLP, and so I'm looking forward to this experiment. I think it will turn out well. I do think we may be setting up some communities for anger, when the Board chooses to *not* grant pTLP status and refers the community to the Incubator. I really don't have a good answer there, especially around the future/obvious direction of pTLP is only for the Old Boys Club and other insiders. Sigh. Can't be helped, I think. Anyhow. To the original point: pTLP reporting to the Board is practically speaking a no-brainer. Podlings generally report direct to the Board today, minus some intermediary stuff. - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
+1 of course from me, with a commitment to doing work to help out and making sure this doesn't fail; instead that it succeeds which I'm sure it will. I'll also sign up for a round of beers at the next ApacheCon. For sure! :) Cheers, Chris ++ Chris Mattmann, Ph.D. Senior Computer Scientist NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA Office: 171-266B, Mailstop: 171-246 Email: chris.a.mattm...@nasa.gov WWW: http://sunset.usc.edu/~mattmann/ ++ Adjunct Assistant Professor, Computer Science Department University of Southern California, Los Angeles, CA 90089 USA ++ -Original Message- From: Ross Gardler rgard...@opendirective.com Reply-To: general@incubator.apache.org general@incubator.apache.org Date: Wednesday, June 12, 2013 7:12 PM To: general@incubator.apache.org general@incubator.apache.org Subject: Re: [DISCUSS] Accept Stratos as an Apache Incubation Project So here's a thought... There have been many discussions about different ways to incubate projects. One of the most radical ideas is to dismantle the incubator and replace the podling concept with probationary TLPs reporting to the board. As readers of this list will know I do not support the idea of dismantling the IPMC. I believe it does a great job that is not easily replaced by a board of nine directors. However, I have always acknowledged that the idea has merit under a certain set of circumstances. For me those circumstances are present in the Apache Stratos proposal. That is there are sufficient mentors and initial committers who are ASF Members that we can be reasonably certain that this project will succeed here at the ASF. I would therefore like to propose that we use Apache Stratos as a test case for the probationary TLP idea. I've already talked to Chris (who is driving the deconstruct the IPMC case) and Ant (who is less keen on dismantling the IPMC but wants to see how a probationary TLP model will play out). Both have agreed to help with this experiment if the IPMC and the Board wish it to proceed. I have not, however, discussed it with all the initial comitters or even mentors - I'm expecting them to speak up now. For my part my intention is to get the project set-up and then dissolve into the background. I do not intend to monitor the project on a day-to-day basis. However, I do promise to help pick up the pieces if the experiment should go horribly wrong. Of course running a single experiment will only allow us to define the incubation process for probationary TLPs, It is not going to solve all the problems Chris sees in the IPMC. However it will give us an opportunity to define the process, ask the board to approve this process and thus lay the foundations for other projects wishing to follow this path. So, what do you think? Ross On 11 June 2013 10:10, Ross Gardler rgard...@opendirective.com wrote: It's with great pleasure that I invite the IPMC to review a new proposal [1] for the Apache Incubator. Please let us know if you have any questions or comments - as you will see there are plenty of people on the initial commit list ready and willing to answer your questions. I copy the full text of the proposal for your convenience: = Stratos - A PaaS Framework = == Abstract == Stratos will be a polyglot [[http://www.gartner.com/it-glossary/platform-as-a-service-paas|PaaS]] framework, providing developers a cloud-based environment for developing, testing, and running scalable applications, and IT providers high utilization rates, automated resource management, and platform-wide insight including monitoring and billing. == Proposal == The Stratos PaaS framework will encompass four layers: 1. An [[http://www.gartner.com/it-glossary/infrastructure-as-a-service-iaas/|Ia aS]]-agnostic layer that can interface with a wide variety of IaaS systems to provide elastic resources, and for multiple IaaS infrastructures to be automated at one time (hybrid clouds.) 2. A PaaS Controller with a cloud controller that automates and monitors IaaS runtime interactions, distributes artifacts to the underlying runtimes, deploys workloads, directs runtime traffic to the right runtimes using a tenant-aware elastic load balancer, and provides a portal for monitoring and provisioning of tenants on the system. 3. Foundational Services including security, logging, messaging, registry, storage (relational, file, and noSQL), task management, and billing. Foundational services will be loosely-coupled to allow swapping in alternate foundational services. 4. A Cartridge Architecture allowing frameworks, servers, and other runtimes to participate in the advantages of the system. The Cartridge Architecture must support multi-tenant workloads, and provide for various levels of tenant isolation and policy-based
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On 13 June 2013 04:56, Alan Cabrera l...@toolazydogs.com wrote: On Jun 12, 2013, at 7:12 PM, Ross Gardler rgard...@opendirective.com wrote: So here's a thought... ... I would therefore like to propose that we use Apache Stratos as a test case for the probationary TLP idea. I've already talked to Chris (who is driving the deconstruct the IPMC case) and Ant (who is less keen on dismantling the IPMC but wants to see how a probationary TLP model will play out). Both have agreed to help with this experiment if the IPMC and the Board wish it to proceed. I have not, however, discussed it with all the initial comitters or even mentors - I'm expecting them to speak up now. ... So, what do you think? I don't see the need to force Stratos through the Incubator given the current proposed membership. Some points: Who's responsible for monitoring the probation, the IPMC or the board? I think it should be the IPMC. I think we should come up with a concrete plan then go to the board. If the board is OK with taking it on then it should be board as this will be closer to Chris' defined end goal. In either case I undertake, as I noted in my original mail, to be the one that steps up to fix things if it all goes wrong. That's true whether it is IPMC or Board. What bits must absolutely be done before probation begins? That needs to be defined. Given the fact the next board meeting is only a week away I suggest we first make this a podling to allow us to start the project here at the ASF. We can then work with the various committees to work out what the right set-up process is (i.e. don't set up as a podling, set up as a pTLP). We can then shoot for submitting a board resolution next month. I have already made it clear to the proposers of the project that taking this route will result in a slightly longer set-up period (because of the need to define new policies along the way). They are comfortable trading slower set-up for potentially faster graduation. What minimum criteria does a probationary TLP have to meet to stay in good graces? Exactly the same as any other TLP. What happens if the probationary TLP is not in good graces? Exactly the same as any other TLP. The board says fix it. If it isn't fixed the board kicks out the problem element(s) and invites remaining PMC to fix it. If that failes the pTLP is sent packing. What bits must absolutely be done before probation completes? Same as graduation from the Incubator (a release, demonstration of a healthy community, approval of the board) Fleshing out these and, I'm sure, others' concerns on a wiki, as Joe pointed out, would be a great idea. Yes, but please note my proposal to do this as a standard podling rather than in this discussion phase. I don't think we need everything in a row before the team can get to work. If it should prove impossible to find a sensible process then we can simply leave the project as a standard podling. So to recap the proposed timeline: - IPMC votes on accepting the podling with the intention of moving it to a pTLP - mentors (with Chris' assistance) guide project committers in working with the various committees to define incubation/probation process - submit a board resolution in July to create the pTLP - if project is not ready to do so this can be delayed until August - If the board are unhappy with the project then I am called in to clear up the mess I made - If the board are happy with progress submit a resolution to become a TLP in 12 months (target 6 months) Ross Regards, Alan - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On 13 June 2013 04:04, Sanjiva Weerawarana sanj...@wso2.com wrote: On Thu, Jun 13, 2013 at 8:27 AM, David Nalley da...@gnsa.us wrote: ccing trademarks@ Note the mix of public and private lists. Yeah we also thought about that .. if StratosLive is too close to home we can certainly change it. Bit painful but not impossible. Better to get an answer to the question up front IMO. +1. Better yes - required before we can vote - no. My reasoning is that WSO2 have already agreed they will use a different name if VP Branding requires it. In fact they won't have any choice since the proposal clearly indicates the Stratos trademark will be assigned to the ASF. That said I encourage someone on the project commit list to mail trademarks@ to understand what would be acceptable. Ross Ross The thinking is to not give that a specific name but rather call it the WSO2 distribution of Apache Stratos (or something like that). I don't think that's one of the 'sanctioned' uses of marks from: In fact, per the below page, it's explicitly forbidden. http://www.apache.org/foundation/marks/faq/#products Hmm that idea was based on how Cloudera is distributing Hadoop: http://www.cloudera.com/content/cloudera/en/products/cdh.html Is that also incorrect then? IIRC they've been doing that for years and presumably ASF is aware of it? Sanjiva. -- Sanjiva Weerawarana, Ph.D. Founder, Chairman CEO; WSO2, Inc.; http://wso2.com/ email: sanj...@wso2.com; phone: +94 11 763 9614; cell: +94 77 787 6880 | +1 650 265 8311 blog: http://sanjiva.weerawarana.org/ Lean . Enterprise . Middleware - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
Ross, thanks for bringing this up! I'm happy to be a part of this experiment. On 13 June 2013 09:13, Ross Gardler rgard...@opendirective.com wrote: Better yes - required before we can vote - no. My reasoning is that WSO2 have already agreed they will use a different name if VP Branding requires it. In fact they won't have any choice since the proposal clearly indicates the Stratos trademark will be assigned to the ASF. While owning the Stratos trademark would certainly put us in a position of being able to ask WSO2 to stop using StratosLive. But there's no guarantee that we would be successful. Case in point: Apache CouchDB and Couchbase. We are very uncomfortable with the name Couchbase, but there's not much we can do about it. Certainly, asking them to rebrand is not going to work. (Full disclosure: I am on the PMC and have been dealing with this mess for over a year.) Couchbase causes us problems because CouchDB's progenitor left CouchDB and founded Couchbase, and made a public statement to the effect that Couchbase was the future of CouchDB. This caused a lot of market confusion, with many people (understandably) thinking that Couchbase replaced CouchDB. You still see the effects of that today. Our support channels are littered with confused Couchbase customers. And I am sure many people who would have been CouchDB users end up downloading and installing Couchbase products. So my concern with StratosLive is: would there be consumer confusion between Apache Stratos and WSO2 StratosLive? My gut tells me: yes. In many people's minds, there will already be some uncertainty about provenance, and so the name will only compound that confusion. I learnt two lessons from the Couchbase fiasco: 1. You need to be mindful of third-parties using your trademark in a compound name. Many people use couch as a sort of adjective. Couch-this and Couch-that. And in one sense, this is great, because couch is this term that is bigger than Apache CouchDB. But at the same time, the meaning is diffuse. And in many ways, the CouchDB community has little control over it. We have plans to remedy that situation, but they are not relevant here. 2. You need to apply branding rules consistently. We allowed Couchbase (and others) to share our brand because they were seen as friendly to the community. (And indeed, for many years, they were hugely beneficial for the project.) Unfortunately, that sets precedent. And it's hard to rewind precedent. It also leaves you vulnerable to the possibility that they won't always be friendly. At which point, you're gonna be SOL. On 13 June 2013 04:04, Sanjiva Weerawarana sanj...@wso2.com wrote: The thinking is to not give that a specific name but rather call it the WSO2 distribution of Apache Stratos (or something like that). I don't think that's one of the 'sanctioned' uses of marks from: In fact, per the below page, it's explicitly forbidden. http://www.apache.org/foundation/marks/faq/#products Hmm that idea was based on how Cloudera is distributing Hadoop: http://www.cloudera.com/content/cloudera/en/products/cdh.html Is that also incorrect then? IIRC they've been doing that for years and presumably ASF is aware of it? The ASF is aware of it, but I don't think we're happy with it. Shane will have more thoughts on this. -- NS
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On 13 June 2013 10:56, Noah Slater nsla...@apache.org wrote: Ross, thanks for bringing this up! I'm happy to be a part of this experiment. On 13 June 2013 09:13, Ross Gardler rgard...@opendirective.com wrote: Better yes - required before we can vote - no. My reasoning is that WSO2 have already agreed they will use a different name if VP Branding requires it. In fact they won't have any choice since the proposal clearly indicates the Stratos trademark will be assigned to the ASF. While owning the Stratos trademark would certainly put us in a position of being able to ask WSO2 to stop using StratosLive. But there's no guarantee that we would be successful. That's true, but there is no change in that risk even if we give WSO2 an answer before the vote. What is more important (in my non-legal opinion) is a publicly archived statement from the WSO2 CEO stating they have no intention of abusing the Stratos mark that will be donated to the ASF. We already have that. It's only about the required order, not about the end game. Note, it is normal practice for pre-existing marks to be formally donated to the ASF during incubation, usually just before graduation. The former owner does not (usually) want to assign a mark that may become useless if the project does not graduate. Such graduation is not wholly under the control of the trademark owner. 1. You need to be mindful of third-parties using your trademark in a compound name. This is already encoded in the ASF trademarks policies - WSO2 have agreed to conform to those policies. 2. You need to apply branding rules consistently. WSO2 have agreed to do so. Sanjiva - perhaps you can edit the proposal to this effect in order to help allay any fears. That is to include the statement you made earlier: if StratosLive is too close to home we can certainly change it. Bit painful but not impossible. I don't want to hold up the vote for entry into the incubator on this issue. At the same time I don't want VP Branding to be in a position of having to come to a quick decision. We will have plenty of time during incubation to resolve everything to our satisfaction. Ultimately VP Branding will be able to object to graduation if the issue has not been adequately addressed during incubation. We allowed Couchbase (and others) to share our brand because they were seen as friendly to the community. This is the root of why CouchDB has a problem today. I believe you a projecting that problem and its cause onto a different issue here. This is not about allowing WSO2 an exception to the existing policy. It is about giving WSO2, the Stratos project community (which includes you) and VP Branding time to work on a satisfactory solution *during* incubation rather than prior to a vote. In my opinion WSO2 have demonstrated they are willing to play by the rules. Ross - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On Thu, Jun 13, 2013 at 3:26 PM, Noah Slater nsla...@apache.org wrote: Ross, thanks for bringing this up! I'm happy to be a part of this experiment. Cool. While owning the Stratos trademark would certainly put us in a position of being able to ask WSO2 to stop using StratosLive. But there's no guarantee that we would be successful. Case in point: Apache CouchDB and As long as I'm running WSO2 that would be successful. However my job is not tenured :-). I learnt two lessons from the Couchbase fiasco: Thanks for your well-reasoned explanation. We will go ahead and change StratosLive as well .. it will take a bit of time but we will do it. (BTW the current StratosLive is not Stratos 2.0 (what we're donating) based.) Hmm that idea was based on how Cloudera is distributing Hadoop: http://www.cloudera.com/content/cloudera/en/products/cdh.html Is that also incorrect then? IIRC they've been doing that for years and presumably ASF is aware of it? The ASF is aware of it, but I don't think we're happy with it. Shane will have more thoughts on this. I understand that but, using your own metric (would there be consumer confusion between Apache Stratos and WSO2 XXX), I would argue that there will be no consumer confusion as to whether what they're downloading is an Apache product or a WSO2 one. Anyway, we are not going to play games with ASF brands; that's not the way WSO2 does stuff. Lets see what Shane says as well. Sanjiva. -- Sanjiva Weerawarana, Ph.D. Founder, Chairman CEO; WSO2, Inc.; http://wso2.com/ email: sanj...@wso2.com; phone: +94 11 763 9614; cell: +94 77 787 6880 | +1 650 265 8311 blog: http://sanjiva.weerawarana.org/ Lean . Enterprise . Middleware
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
I'm just happy to have PaaS as a part of Apache. Thanks to all who are doing this. -- With best regards / с наилучшими пожеланиями, Alexei Fedotov / Алексей Федотов, http://dataved.ru/ +7 916 562 8095 On Thu, Jun 13, 2013 at 2:21 PM, Ross Gardler rgard...@opendirective.com wrote: On 13 June 2013 10:56, Noah Slater nsla...@apache.org wrote: Ross, thanks for bringing this up! I'm happy to be a part of this experiment. On 13 June 2013 09:13, Ross Gardler rgard...@opendirective.com wrote: Better yes - required before we can vote - no. My reasoning is that WSO2 have already agreed they will use a different name if VP Branding requires it. In fact they won't have any choice since the proposal clearly indicates the Stratos trademark will be assigned to the ASF. While owning the Stratos trademark would certainly put us in a position of being able to ask WSO2 to stop using StratosLive. But there's no guarantee that we would be successful. That's true, but there is no change in that risk even if we give WSO2 an answer before the vote. What is more important (in my non-legal opinion) is a publicly archived statement from the WSO2 CEO stating they have no intention of abusing the Stratos mark that will be donated to the ASF. We already have that. It's only about the required order, not about the end game. Note, it is normal practice for pre-existing marks to be formally donated to the ASF during incubation, usually just before graduation. The former owner does not (usually) want to assign a mark that may become useless if the project does not graduate. Such graduation is not wholly under the control of the trademark owner. 1. You need to be mindful of third-parties using your trademark in a compound name. This is already encoded in the ASF trademarks policies - WSO2 have agreed to conform to those policies. 2. You need to apply branding rules consistently. WSO2 have agreed to do so. Sanjiva - perhaps you can edit the proposal to this effect in order to help allay any fears. That is to include the statement you made earlier: if StratosLive is too close to home we can certainly change it. Bit painful but not impossible. I don't want to hold up the vote for entry into the incubator on this issue. At the same time I don't want VP Branding to be in a position of having to come to a quick decision. We will have plenty of time during incubation to resolve everything to our satisfaction. Ultimately VP Branding will be able to object to graduation if the issue has not been adequately addressed during incubation. We allowed Couchbase (and others) to share our brand because they were seen as friendly to the community. This is the root of why CouchDB has a problem today. I believe you a projecting that problem and its cause onto a different issue here. This is not about allowing WSO2 an exception to the existing policy. It is about giving WSO2, the Stratos project community (which includes you) and VP Branding time to work on a satisfactory solution *during* incubation rather than prior to a vote. In my opinion WSO2 have demonstrated they are willing to play by the rules. Ross - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On Thu, Jun 13, 2013 at 3:51 PM, Ross Gardler rgard...@opendirective.comwrote: 2. You need to apply branding rules consistently. WSO2 have agreed to do so. Sanjiva - perhaps you can edit the proposal to this effect in order to help allay any fears. That is to include the statement you made earlier: if StratosLive is too close to home we can certainly change it. Bit painful but not impossible. Will do - I will go further and state that by the time the project graduates we will deprecate WSO2 StratosLive. Sanjiva. -- Sanjiva Weerawarana, Ph.D. Founder, Chairman CEO; WSO2, Inc.; http://wso2.com/ email: sanj...@wso2.com; phone: +94 11 763 9614; cell: +94 77 787 6880 | +1 650 265 8311 blog: http://sanjiva.weerawarana.org/ Lean . Enterprise . Middleware
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On 13 June 2013 11:26, Sanjiva Weerawarana sanj...@wso2.com wrote: On Thu, Jun 13, 2013 at 3:51 PM, Ross Gardler rgard...@opendirective.comwrote: 2. You need to apply branding rules consistently. WSO2 have agreed to do so. Sanjiva - perhaps you can edit the proposal to this effect in order to help allay any fears. That is to include the statement you made earlier: if StratosLive is too close to home we can certainly change it. Bit painful but not impossible. Will do - I will go further and state that by the time the project graduates we will deprecate WSO2 StratosLive. :-D This is perfect thank you. Ross Sanjiva. -- Sanjiva Weerawarana, Ph.D. Founder, Chairman CEO; WSO2, Inc.; http://wso2.com/ email: sanj...@wso2.com; phone: +94 11 763 9614; cell: +94 77 787 6880 | +1 650 265 8311 blog: http://sanjiva.weerawarana.org/ Lean . Enterprise . Middleware - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On 13 June 2013 11:21, Ross Gardler rgard...@opendirective.com wrote: This is the root of why CouchDB has a problem today. I believe you a projecting that problem and its cause onto a different issue here. This is not about allowing WSO2 an exception to the existing policy. It is about giving WSO2, the Stratos project community (which includes you) and VP Branding time to work on a satisfactory solution *during* incubation rather than prior to a vote. In my opinion WSO2 have demonstrated they are willing to play by the rules. Quite possibly. Once bitten... ;) In any case, we are in agreement. To clarify my intent: I only meant to share my experience so that we're mindful of the territory we're venturing into. I don't think the proposal needs to be delayed. We will have plenty of time to work on these things. Also. Thanks for your follow-up emails, Sanjiva's! That's good news. -- NS
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On Jun 12, 2013, at 7:32 PM, Marvin Humphrey mar...@rectangular.com wrote: On Tue, Jun 11, 2013 at 2:10 AM, Ross Gardler rgard...@opendirective.com wrote: It's with great pleasure that I invite the IPMC to review a new proposal [1] for the Apache Incubator. Please let us know if you have any questions or comments - as you will see there are plenty of people on the initial commit list ready and willing to answer your questions. It is certainly a large group and a thoroughly prepared proposal with a lot of resources behind it. == Known Risks == Stratos has largely been developed by sponsored developers employed at a single organization - WSO2. Seeking a broader community of contributors is a top goal of contributing Stratos to Apache. WSO2 plans to continue to offer services and commercial support packages for Stratos, so there is a financial incentive to broaden Stratos’ appeal. This may provide the misinterpretation that Stratos remains merely a WSO2 technology. However, WSO2’s main business strategy is to build and support higher level PaaS offerings (including the WSO2 middleware stack) on top of a common PaaS framework, as provided by Stratos. This includes a WSO2 StratosLive option which is a public PaaS based on WSO2 Stratos. Kudos for the honest self-analysis. I wonder whether Apache Brand Management would cry foul on a trademark like StratosLive if such a product were to appear later because of the confusingly similar name guideline. We've seen project founders leave and compete with ASF products while using confusingly similar names before, e.g. CouchBase. I agree. I would assume that Stratos would become a trademark of the ASF and, as such, WSO2's use of StratosLive as a product name would not be allowed. - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
(Please note that this conversation is happening on general@, a publicly archived list, so there is a limit to what level of details of brand strategy I will engage in on this thread.) To be brief: - This issue should not gate a [VOTE] to begin incubation. - Given both WSO2's history and the many Apache members involved in the project, I'm confident the right thing will end up happening. - The requirement is that the branding for a podling must be formally donated to the ASF before it can graduate. We are happy to sign agreements earlier or later; if we sign earlier we're happy to include language such that any trademark assignment will revert to the original owner if graduation fails for some reason. - Unofficially (meaning: I've only spent 5 minutes reading this all so far) I do *not* believe it would be tenable to have a future Apache Stratos and a future WSO2 StratosLive. Many thanks in advance to Sanjiva and company for being willing to change this! - Shane On 6/13/2013 6:21 AM, Ross Gardler wrote: On 13 June 2013 10:56, Noah Slater nsla...@apache.org wrote: Ross, thanks for bringing this up! I'm happy to be a part of this experiment. On 13 June 2013 09:13, Ross Gardler rgard...@opendirective.com wrote: Better yes - required before we can vote - no. My reasoning is that WSO2 have already agreed they will use a different name if VP Branding requires it. In fact they won't have any choice since the proposal clearly indicates the Stratos trademark will be assigned to the ASF. While owning the Stratos trademark would certainly put us in a position of being able to ask WSO2 to stop using StratosLive. But there's no guarantee that we would be successful. That's true, but there is no change in that risk even if we give WSO2 an answer before the vote. What is more important (in my non-legal opinion) is a publicly archived statement from the WSO2 CEO stating they have no intention of abusing the Stratos mark that will be donated to the ASF. We already have that. It's only about the required order, not about the end game. Note, it is normal practice for pre-existing marks to be formally donated to the ASF during incubation, usually just before graduation. The former owner does not (usually) want to assign a mark that may become useless if the project does not graduate. Such graduation is not wholly under the control of the trademark owner. 1. You need to be mindful of third-parties using your trademark in a compound name. This is already encoded in the ASF trademarks policies - WSO2 have agreed to conform to those policies. 2. You need to apply branding rules consistently. WSO2 have agreed to do so. Sanjiva - perhaps you can edit the proposal to this effect in order to help allay any fears. That is to include the statement you made earlier: if StratosLive is too close to home we can certainly change it. Bit painful but not impossible. I don't want to hold up the vote for entry into the incubator on this issue. At the same time I don't want VP Branding to be in a position of having to come to a quick decision. We will have plenty of time during incubation to resolve everything to our satisfaction. Ultimately VP Branding will be able to object to graduation if the issue has not been adequately addressed during incubation. We allowed Couchbase (and others) to share our brand because they were seen as friendly to the community. This is the root of why CouchDB has a problem today. I believe you a projecting that problem and its cause onto a different issue here. This is not about allowing WSO2 an exception to the existing policy. It is about giving WSO2, the Stratos project community (which includes you) and VP Branding time to work on a satisfactory solution *during* incubation rather than prior to a vote. In my opinion WSO2 have demonstrated they are willing to play by the rules. Ross - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On Jun 13, 2013, at 1:10 AM, Ross Gardler rgard...@opendirective.com wrote: On 13 June 2013 04:56, Alan Cabrera l...@toolazydogs.com wrote: On Jun 12, 2013, at 7:12 PM, Ross Gardler rgard...@opendirective.com wrote: So here's a thought... ... I would therefore like to propose that we use Apache Stratos as a test case for the probationary TLP idea. I've already talked to Chris (who is driving the deconstruct the IPMC case) and Ant (who is less keen on dismantling the IPMC but wants to see how a probationary TLP model will play out). Both have agreed to help with this experiment if the IPMC and the Board wish it to proceed. I have not, however, discussed it with all the initial comitters or even mentors - I'm expecting them to speak up now. ... So, what do you think? I don't see the need to force Stratos through the Incubator given the current proposed membership. Some points: Who's responsible for monitoring the probation, the IPMC or the board? I think it should be the IPMC. I think we should come up with a concrete plan then go to the board. If the board is OK with taking it on then it should be board as this will be closer to Chris' defined end goal. In either case I undertake, as I noted in my original mail, to be the one that steps up to fix things if it all goes wrong. That's true whether it is IPMC or Board. I guess the details of how this governance will work, what are the roles, and who will fill them, will need to be ironed out. What bits must absolutely be done before probation begins? That needs to be defined. Given the fact the next board meeting is only a week away I suggest we first make this a podling to allow us to start the project here at the ASF. We can then work with the various committees to work out what the right set-up process is (i.e. don't set up as a podling, set up as a pTLP). We can then shoot for submitting a board resolution next month. I have already made it clear to the proposers of the project that taking this route will result in a slightly longer set-up period (because of the need to define new policies along the way). They are comfortable trading slower set-up for potentially faster graduation. It would probably be good to be clear on what are the exact characteristics that make this podling pTLP worthy for the future. For example, the number of ASF veterans in its ranks. What minimum criteria does a probationary TLP have to meet to stay in good graces? Exactly the same as any other TLP. What happens if the probationary TLP is not in good graces? Exactly the same as any other TLP. The board says fix it. If it isn't fixed the board kicks out the problem element(s) and invites remaining PMC to fix it. If that failes the pTLP is sent packing. What bits must absolutely be done before probation completes? Same as graduation from the Incubator (a release, demonstration of a healthy community, approval of the board) Nice and simple. Fleshing out these and, I'm sure, others' concerns on a wiki, as Joe pointed out, would be a great idea. Yes, but please note my proposal to do this as a standard podling rather than in this discussion phase. I don't think we need everything in a row before the team can get to work. If it should prove impossible to find a sensible process then we can simply leave the project as a standard podling. Makes sense. So to recap the proposed timeline: - IPMC votes on accepting the podling with the intention of moving it to a pTLP - mentors (with Chris' assistance) guide project committers in working with the various committees to define incubation/probation process - submit a board resolution in July to create the pTLP - if project is not ready to do so this can be delayed until August - If the board are unhappy with the project then I am called in to clear up the mess I made - If the board are happy with progress submit a resolution to become a TLP in 12 months (target 6 months) +1 Though I wouldn't put a date on TLP; keep things simple. We don't for podlings and since the pTLP will be filled with trustworthy ASF members we can trust they will do the right thing. Regards, Alan - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On 13 June 2013 14:12, Alan Cabrera l...@toolazydogs.com wrote: On Jun 13, 2013, at 1:10 AM, Ross Gardler rgard...@opendirective.com wrote: On 13 June 2013 04:56, Alan Cabrera l...@toolazydogs.com wrote: On Jun 12, 2013, at 7:12 PM, Ross Gardler rgard...@opendirective.com wrote: So here's a thought... ... I would therefore like to propose that we use Apache Stratos as a test case for the probationary TLP idea. I've already talked to Chris (who is driving the deconstruct the IPMC case) and Ant (who is less keen on dismantling the IPMC but wants to see how a probationary TLP model will play out). Both have agreed to help with this experiment if the IPMC and the Board wish it to proceed. I have not, however, discussed it with all the initial comitters or even mentors - I'm expecting them to speak up now. ... So, what do you think? I don't see the need to force Stratos through the Incubator given the current proposed membership. Some points: Who's responsible for monitoring the probation, the IPMC or the board? I think it should be the IPMC. I think we should come up with a concrete plan then go to the board. If the board is OK with taking it on then it should be board as this will be closer to Chris' defined end goal. In either case I undertake, as I noted in my original mail, to be the one that steps up to fix things if it all goes wrong. That's true whether it is IPMC or Board. I guess the details of how this governance will work, what are the roles, and who will fill them, will need to be ironed out. Yes. Of course in this case I'm proposing a period as a podling to give us time to iron those details out. However, here's my starting suggestion: This is just a TLP so we need to identify is committers, PMC, PMC chair. My starting suggestion is: - commiters (see proposal) - PMC members (I suggest initial membership is the mentors, the mentors seek to vote initial committers into the PMC as quickly as possible) - PMC chair (I would suggest the chair is the Champion until the PMC is confident enough to elect one from their own ranks - should be done ASAP, but certainly before graduation) What bits must absolutely be done before probation begins? That needs to be defined. Given the fact the next board meeting is only a week away I suggest we first make this a podling to allow us to start the project here at the ASF. We can then work with the various committees to work out what the right set-up process is (i.e. don't set up as a podling, set up as a pTLP). We can then shoot for submitting a board resolution next month. I have already made it clear to the proposers of the project that taking this route will result in a slightly longer set-up period (because of the need to define new policies along the way). They are comfortable trading slower set-up for potentially faster graduation. It would probably be good to be clear on what are the exact characteristics that make this podling pTLP worthy for the future. For example, the number of ASF veterans in its ranks. The board expects a TLP to be able to make releases. That requires 3 +1 votes. That implies 3 initial PMC members. According to my starting proposal above this means 3 mentors minimum. This in turn matches what has come to be common practice in the IPMC. So to recap the proposed timeline: - IPMC votes on accepting the podling with the intention of moving it to a pTLP - mentors (with Chris' assistance) guide project committers in working with the various committees to define incubation/probation process - submit a board resolution in July to create the pTLP - if project is not ready to do so this can be delayed until August - If the board are unhappy with the project then I am called in to clear up the mess I made - If the board are happy with progress submit a resolution to become a TLP in 12 months (target 6 months) +1 Though I wouldn't put a date on TLP; keep things simple. We don't for podlings and since the pTLP will be filled with trustworthy ASF members we can trust they will do the right thing. Yes, I did wonder about that when I was writing this. I kept adding and removing it. I would like a date in there as targets are always something to aim for. The fact we don't have such a target for podlings is one of the items that some people suggest needs fixing. That said, the board is sensible enough to give a pTLP longer than the target if it is clear things are moving in the right direction. I would suggest we keep the date but make sure it is only a guideline. Ross - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On Thu, Jun 13, 2013 at 05:39:24AM +0530, Sanjiva Weerawarana wrote: On Thu, Jun 13, 2013 at 5:02 AM, Marvin Humphrey mar...@rectangular.comwrote: == Known Risks == Stratos has largely been developed by sponsored developers employed at a single organization - WSO2. Seeking a broader community of contributors is a top goal of contributing Stratos to Apache. WSO2 plans to continue to offer services and commercial support packages for Stratos, so there is a financial incentive to broaden Stratos’ appeal. This may provide the misinterpretation that Stratos remains merely a WSO2 technology. However, WSO2’s main business strategy is to build and support higher level PaaS offerings (including the WSO2 middleware stack) on top of a common PaaS framework, as provided by Stratos. This includes a WSO2 StratosLive option which is a public PaaS based on WSO2 Stratos. Kudos for the honest self-analysis. Thanks :-). I wonder whether Apache Brand Management would cry foul on a trademark like StratosLive if such a product were to appear later because of the confusingly similar name guideline. We've seen project founders leave and compete with ASF products while using confusingly similar names before, e.g. CouchBase. Yeah we also thought about that .. if StratosLive is too close to home we can certainly change it. Bit painful but not impossible. Alternatively, if WS02 wanted to keep using the Stratos term, the podling (to be) could re-brand the project itself to something other than Stratos. - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On Thu, Jun 13, 2013, at 08:58 AM, Chip Childers wrote: Alternatively, if WS02 wanted to keep using the Stratos term, the podling (to be) could re-brand the project itself to something other than Stratos. I think this would be a better way to go, honestly. If there was one thing I could do over again about CloudStack, it would have been to have let Citrix keep the name for their commercial offering and choose a different name for the Apache project. Best, jzb -- Joe Brockmeier j...@zonker.net Twitter: @jzb http://www.dissociatedpress.net/ - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On Wed, Jun 12, 2013 at 7:12 PM, Ross Gardler rgard...@opendirective.com wrote: I would therefore like to propose that we use Apache Stratos as a test case for the probationary TLP idea. If the Stratos folks are amenable, +1. It's an incremental, reversible step forward. With so many experienced Members and Mentors involved, Stratos is assured of getting everything they would get out of an ordinary incubation and more, even if there are some bumps. And I hope there will be a payoff for Stratos by sidestepping what is IMO the Incubator's flawed approach to meritocracy[1], where true responsibility arrives all at once on graduation. On general@, we'll probably get the usual storms of email. But we would have gotten those anyway, and at least we'll be talking about real data rather than hypotheticals. Marvin Humphrey [1] http://s.apache.org/nzm - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
+1 for the probationary TLP idea. Azeez On Thu, Jun 13, 2013 at 7:42 AM, Ross Gardler rgard...@opendirective.comwrote: So here's a thought... There have been many discussions about different ways to incubate projects. One of the most radical ideas is to dismantle the incubator and replace the podling concept with probationary TLPs reporting to the board. As readers of this list will know I do not support the idea of dismantling the IPMC. I believe it does a great job that is not easily replaced by a board of nine directors. However, I have always acknowledged that the idea has merit under a certain set of circumstances. For me those circumstances are present in the Apache Stratos proposal. That is there are sufficient mentors and initial committers who are ASF Members that we can be reasonably certain that this project will succeed here at the ASF. I would therefore like to propose that we use Apache Stratos as a test case for the probationary TLP idea. I've already talked to Chris (who is driving the deconstruct the IPMC case) and Ant (who is less keen on dismantling the IPMC but wants to see how a probationary TLP model will play out). Both have agreed to help with this experiment if the IPMC and the Board wish it to proceed. I have not, however, discussed it with all the initial comitters or even mentors - I'm expecting them to speak up now. For my part my intention is to get the project set-up and then dissolve into the background. I do not intend to monitor the project on a day-to-day basis. However, I do promise to help pick up the pieces if the experiment should go horribly wrong. Of course running a single experiment will only allow us to define the incubation process for probationary TLPs, It is not going to solve all the problems Chris sees in the IPMC. However it will give us an opportunity to define the process, ask the board to approve this process and thus lay the foundations for other projects wishing to follow this path. So, what do you think? Ross On 11 June 2013 10:10, Ross Gardler rgard...@opendirective.com wrote: It's with great pleasure that I invite the IPMC to review a new proposal [1] for the Apache Incubator. Please let us know if you have any questions or comments - as you will see there are plenty of people on the initial commit list ready and willing to answer your questions. I copy the full text of the proposal for your convenience: = Stratos - A PaaS Framework = == Abstract == Stratos will be a polyglot [[http://www.gartner.com/it-glossary/platform-as-a-service-paas|PaaS]] framework, providing developers a cloud-based environment for developing, testing, and running scalable applications, and IT providers high utilization rates, automated resource management, and platform-wide insight including monitoring and billing. == Proposal == The Stratos PaaS framework will encompass four layers: 1. An [[ http://www.gartner.com/it-glossary/infrastructure-as-a-service-iaas/|IaaS]]-agnostic layer that can interface with a wide variety of IaaS systems to provide elastic resources, and for multiple IaaS infrastructures to be automated at one time (hybrid clouds.) 2. A PaaS Controller with a cloud controller that automates and monitors IaaS runtime interactions, distributes artifacts to the underlying runtimes, deploys workloads, directs runtime traffic to the right runtimes using a tenant-aware elastic load balancer, and provides a portal for monitoring and provisioning of tenants on the system. 3. Foundational Services including security, logging, messaging, registry, storage (relational, file, and noSQL), task management, and billing. Foundational services will be loosely-coupled to allow swapping in alternate foundational services. 4. A Cartridge Architecture allowing frameworks, servers, and other runtimes to participate in the advantages of the system. The Cartridge Architecture must support multi-tenant workloads, and provide for various levels of tenant isolation and policy-based control over provisioning. Together these layers offer a foundational layer upon which applications and middleware frameworks can be deployed to speed time-to-market and simplify the development of scalable applications, as well as provide a high level of resource sharing and centralized management that can deliver lowest resource, infrastructure, and management costs. == Background == The Stratos Project has been under development[a] at http://wso2.org under the Apache 2.0 license and the Apache Way governance model since 2010. It initially was focussed on providing PaaS benefits to the users of WSO2 Carbon middleware platform. In version 2.0, to be released in summer 2013, extensive work has been done to clearly separate out the PaaS framework from the products (cartridges) that run on top of it. Stratos now has the ability to run arbitrary
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
Hi The proposal is well detailed and very informative, the purpose of the projects fits ASF's Cloud Ecosystem and the community is very promising already reminds me with the one of ACS. Very looking forward to start rolling the podling tasks On Tue, Jun 11, 2013 at 11:46 AM, Christian Grobmeier grobme...@gmail.comwrote: Impressive proposal. Reminds me on RH OpenShift actually. Please note the first two links are broken in gmail. They are: * http://www.gartner.com/it-glossary/platform-as-a-service-paas * http://www.gartner.com/it-glossary/infrastructure-as-a-service-iaas Anyway, proposal looks really good for me. On Tue, Jun 11, 2013 at 11:10 AM, Ross Gardler rgard...@opendirective.com wrote: It's with great pleasure that I invite the IPMC to review a new proposal [1] for the Apache Incubator. Please let us know if you have any questions or comments - as you will see there are plenty of people on the initial commit list ready and willing to answer your questions. I copy the full text of the proposal for your convenience: = Stratos - A PaaS Framework = == Abstract == Stratos will be a polyglot [[http://www.gartner.com/it-glossary/platform-as-a-service-paas|PaaS]] framework, providing developers a cloud-based environment for developing, testing, and running scalable applications, and IT providers high utilization rates, automated resource management, and platform-wide insight including monitoring and billing. == Proposal == The Stratos PaaS framework will encompass four layers: 1. An [[ http://www.gartner.com/it-glossary/infrastructure-as-a-service-iaas/|IaaS]]-agnostic layer that can interface with a wide variety of IaaS systems to provide elastic resources, and for multiple IaaS infrastructures to be automated at one time (hybrid clouds.) 2. A PaaS Controller with a cloud controller that automates and monitors IaaS runtime interactions, distributes artifacts to the underlying runtimes, deploys workloads, directs runtime traffic to the right runtimes using a tenant-aware elastic load balancer, and provides a portal for monitoring and provisioning of tenants on the system. 3. Foundational Services including security, logging, messaging, registry, storage (relational, file, and noSQL), task management, and billing. Foundational services will be loosely-coupled to allow swapping in alternate foundational services. 4. A Cartridge Architecture allowing frameworks, servers, and other runtimes to participate in the advantages of the system. The Cartridge Architecture must support multi-tenant workloads, and provide for various levels of tenant isolation and policy-based control over provisioning. Together these layers offer a foundational layer upon which applications and middleware frameworks can be deployed to speed time-to-market and simplify the development of scalable applications, as well as provide a high level of resource sharing and centralized management that can deliver lowest resource, infrastructure, and management costs. == Background == The Stratos Project has been under development[a] at http://wso2.org under the Apache 2.0 license and the Apache Way governance model since 2010. It initially was focussed on providing PaaS benefits to the users of WSO2 Carbon middleware platform. In version 2.0, to be released in summer 2013, extensive work has been done to clearly separate out the PaaS framework from the products (cartridges) that run on top of it. Stratos now has the ability to run arbitrary workloads, including Java, PHP, MySQL, Jetty, Tomcat, and many more. == Rationale == PaaS is in demand by enterprises and organizations of all sizes. The drive towards instance provisioning, high resource utilization and thus low cost, combined with a wide platform of general-purpose services to build on, PaaS has the opportunity to accelerate the development cycle and innovation index of a new class of applications, services, and business models. PaaS offerings are widely diversified but largely associated with powerful corporate interests. With the commencement of the Stratos project at Apache, vendors and users will have a neutral community free from corporate governance restrictions, with which to collaborate and accelerate the development of a platform that provides wide benefits across the industry. As a flexible framework, we expect a wide variety of platforms to leverage the technology to fill specific niches and needs. == Current Status == Stratos has been in development since 2010 at WSO2, under the Apache License and under the Apache Way. Contribution to Apache, from which many of the core components are sourced, should be very straightforward. == Meritocracy == The contributors have a longstanding commitment and practice of meritocracy in their personal and professional capacities. Many of the committers on the
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On Wed, Jun 12, 2013 at 09:53:22PM +0200, Mohammad Nour El-Din wrote: Hi The proposal is well detailed and very informative, the purpose of the projects fits ASF's Cloud Ecosystem and the community is very promising already reminds me with the one of ACS. Very looking forward to start rolling the podling tasks +1 - looking forward to seeing the VOTE start, and for us to get going on building an ACS provider plugin ;-) - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On Tue, Jun 11, 2013 at 2:10 AM, Ross Gardler rgard...@opendirective.com wrote: It's with great pleasure that I invite the IPMC to review a new proposal [1] for the Apache Incubator. Please let us know if you have any questions or comments - as you will see there are plenty of people on the initial commit list ready and willing to answer your questions. It is certainly a large group and a thoroughly prepared proposal with a lot of resources behind it. == Known Risks == Stratos has largely been developed by sponsored developers employed at a single organization - WSO2. Seeking a broader community of contributors is a top goal of contributing Stratos to Apache. WSO2 plans to continue to offer services and commercial support packages for Stratos, so there is a financial incentive to broaden Stratos’ appeal. This may provide the misinterpretation that Stratos remains merely a WSO2 technology. However, WSO2’s main business strategy is to build and support higher level PaaS offerings (including the WSO2 middleware stack) on top of a common PaaS framework, as provided by Stratos. This includes a WSO2 StratosLive option which is a public PaaS based on WSO2 Stratos. Kudos for the honest self-analysis. I wonder whether Apache Brand Management would cry foul on a trademark like StratosLive if such a product were to appear later because of the confusingly similar name guideline. We've seen project founders leave and compete with ASF products while using confusingly similar names before, e.g. CouchBase. If a top goal is seeking a broader community of contributors, the project might benefit if it could find a way to reassure potential contributors on this matter. Perhaps the name for the WS02 product should be vetted in advance? http://www.apache.org/foundation/marks/ === Mailing Lists === 1. d...@stratos.incubator.apache.org - for developer/user discussions, JIRA change notifications continuous build/test notifications 2. comm...@stratos.incubator.apache.org - for commit mails It's a technicality, but the podling would also need a private@ list. Since there's only a combined dev/user list to start with though, obviously some thought has gone into this. :) Marvin Humphrey - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On Thu, Jun 13, 2013 at 5:02 AM, Marvin Humphrey mar...@rectangular.comwrote: == Known Risks == Stratos has largely been developed by sponsored developers employed at a single organization - WSO2. Seeking a broader community of contributors is a top goal of contributing Stratos to Apache. WSO2 plans to continue to offer services and commercial support packages for Stratos, so there is a financial incentive to broaden Stratos’ appeal. This may provide the misinterpretation that Stratos remains merely a WSO2 technology. However, WSO2’s main business strategy is to build and support higher level PaaS offerings (including the WSO2 middleware stack) on top of a common PaaS framework, as provided by Stratos. This includes a WSO2 StratosLive option which is a public PaaS based on WSO2 Stratos. Kudos for the honest self-analysis. Thanks :-). I wonder whether Apache Brand Management would cry foul on a trademark like StratosLive if such a product were to appear later because of the confusingly similar name guideline. We've seen project founders leave and compete with ASF products while using confusingly similar names before, e.g. CouchBase. Yeah we also thought about that .. if StratosLive is too close to home we can certainly change it. Bit painful but not impossible. If a top goal is seeking a broader community of contributors, the project might benefit if it could find a way to reassure potential contributors on this matter. Perhaps the name for the WS02 product should be vetted in advance? Yep can do - our current thinking is to have a WSO2 distribution that contains Apache Stratos and includes all of our middleware products (app server, esb, etc. etc.). That is, a PaaS built on the PaaS framework from Apache. The thinking is to not give that a specific name but rather call it the WSO2 distribution of Apache Stratos (or something like that). Shane do you have thoughts on that? That idea is based on what some others do and it seems to me is clear enough that its not just the Apache code being redistributed but has more stuff. Analogous to Redhat being a Linux distribution. Cheers, Sanjiva. -- Sanjiva Weerawarana, Ph.D. Founder, Chairman CEO; WSO2, Inc.; http://wso2.com/ email: sanj...@wso2.com; phone: +94 11 763 9614; cell: +94 77 787 6880 | +1 650 265 8311 blog: http://sanjiva.weerawarana.org/ Lean . Enterprise . Middleware
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
+1 -- dims On Tue, Jun 11, 2013 at 5:10 AM, Ross Gardler rgard...@opendirective.com wrote: It's with great pleasure that I invite the IPMC to review a new proposal [1] for the Apache Incubator. Please let us know if you have any questions or comments - as you will see there are plenty of people on the initial commit list ready and willing to answer your questions. I copy the full text of the proposal for your convenience: = Stratos - A PaaS Framework = == Abstract == Stratos will be a polyglot [[http://www.gartner.com/it-glossary/platform-as-a-service-paas|PaaS]] framework, providing developers a cloud-based environment for developing, testing, and running scalable applications, and IT providers high utilization rates, automated resource management, and platform-wide insight including monitoring and billing. == Proposal == The Stratos PaaS framework will encompass four layers: 1. An [[http://www.gartner.com/it-glossary/infrastructure-as-a-service-iaas/|IaaS]]-agnostic layer that can interface with a wide variety of IaaS systems to provide elastic resources, and for multiple IaaS infrastructures to be automated at one time (hybrid clouds.) 2. A PaaS Controller with a cloud controller that automates and monitors IaaS runtime interactions, distributes artifacts to the underlying runtimes, deploys workloads, directs runtime traffic to the right runtimes using a tenant-aware elastic load balancer, and provides a portal for monitoring and provisioning of tenants on the system. 3. Foundational Services including security, logging, messaging, registry, storage (relational, file, and noSQL), task management, and billing. Foundational services will be loosely-coupled to allow swapping in alternate foundational services. 4. A Cartridge Architecture allowing frameworks, servers, and other runtimes to participate in the advantages of the system. The Cartridge Architecture must support multi-tenant workloads, and provide for various levels of tenant isolation and policy-based control over provisioning. Together these layers offer a foundational layer upon which applications and middleware frameworks can be deployed to speed time-to-market and simplify the development of scalable applications, as well as provide a high level of resource sharing and centralized management that can deliver lowest resource, infrastructure, and management costs. == Background == The Stratos Project has been under development[a] at http://wso2.org under the Apache 2.0 license and the Apache Way governance model since 2010. It initially was focussed on providing PaaS benefits to the users of WSO2 Carbon middleware platform. In version 2.0, to be released in summer 2013, extensive work has been done to clearly separate out the PaaS framework from the products (cartridges) that run on top of it. Stratos now has the ability to run arbitrary workloads, including Java, PHP, MySQL, Jetty, Tomcat, and many more. == Rationale == PaaS is in demand by enterprises and organizations of all sizes. The drive towards instance provisioning, high resource utilization and thus low cost, combined with a wide platform of general-purpose services to build on, PaaS has the opportunity to accelerate the development cycle and innovation index of a new class of applications, services, and business models. PaaS offerings are widely diversified but largely associated with powerful corporate interests. With the commencement of the Stratos project at Apache, vendors and users will have a neutral community free from corporate governance restrictions, with which to collaborate and accelerate the development of a platform that provides wide benefits across the industry. As a flexible framework, we expect a wide variety of platforms to leverage the technology to fill specific niches and needs. == Current Status == Stratos has been in development since 2010 at WSO2, under the Apache License and under the Apache Way. Contribution to Apache, from which many of the core components are sourced, should be very straightforward. == Meritocracy == The contributors have a longstanding commitment and practice of meritocracy in their personal and professional capacities. Many of the committers on the existing project are already Apache Committers. == Community == The committers recognize the need to support more significant contributions from a broad swath of the industry and community. Apache is the obvious choice for a project that already embodies Apache values and is driven by developers who are committers on many other Apache projects. Stratos has appeal, and should attract community members, from among a number of constituencies: * Private PaaS deployment within an enterprise to benefit enterprise application deployment. * PaaS providers wishing to leverage a PaaS toolkit to build a PaaS customized to their particular vertical industry, ecosystem, or internal development
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
Great to see WSO2 Stratos coming to Apache, and I would love to join/help this project. Deepal On Tue, Jun 11, 2013 at 5:10 AM, Ross Gardler rgard...@opendirective.comwrote: It's with great pleasure that I invite the IPMC to review a new proposal [1] for the Apache Incubator. Please let us know if you have any questions or comments - as you will see there are plenty of people on the initial commit list ready and willing to answer your questions. I copy the full text of the proposal for your convenience: = Stratos - A PaaS Framework = == Abstract == Stratos will be a polyglot [[http://www.gartner.com/it-glossary/platform-as-a-service-paas|PaaS]] framework, providing developers a cloud-based environment for developing, testing, and running scalable applications, and IT providers high utilization rates, automated resource management, and platform-wide insight including monitoring and billing. == Proposal == The Stratos PaaS framework will encompass four layers: 1. An [[ http://www.gartner.com/it-glossary/infrastructure-as-a-service-iaas/|IaaS]]-agnostic layer that can interface with a wide variety of IaaS systems to provide elastic resources, and for multiple IaaS infrastructures to be automated at one time (hybrid clouds.) 2. A PaaS Controller with a cloud controller that automates and monitors IaaS runtime interactions, distributes artifacts to the underlying runtimes, deploys workloads, directs runtime traffic to the right runtimes using a tenant-aware elastic load balancer, and provides a portal for monitoring and provisioning of tenants on the system. 3. Foundational Services including security, logging, messaging, registry, storage (relational, file, and noSQL), task management, and billing. Foundational services will be loosely-coupled to allow swapping in alternate foundational services. 4. A Cartridge Architecture allowing frameworks, servers, and other runtimes to participate in the advantages of the system. The Cartridge Architecture must support multi-tenant workloads, and provide for various levels of tenant isolation and policy-based control over provisioning. Together these layers offer a foundational layer upon which applications and middleware frameworks can be deployed to speed time-to-market and simplify the development of scalable applications, as well as provide a high level of resource sharing and centralized management that can deliver lowest resource, infrastructure, and management costs. == Background == The Stratos Project has been under development[a] at http://wso2.org under the Apache 2.0 license and the Apache Way governance model since 2010. It initially was focussed on providing PaaS benefits to the users of WSO2 Carbon middleware platform. In version 2.0, to be released in summer 2013, extensive work has been done to clearly separate out the PaaS framework from the products (cartridges) that run on top of it. Stratos now has the ability to run arbitrary workloads, including Java, PHP, MySQL, Jetty, Tomcat, and many more. == Rationale == PaaS is in demand by enterprises and organizations of all sizes. The drive towards instance provisioning, high resource utilization and thus low cost, combined with a wide platform of general-purpose services to build on, PaaS has the opportunity to accelerate the development cycle and innovation index of a new class of applications, services, and business models. PaaS offerings are widely diversified but largely associated with powerful corporate interests. With the commencement of the Stratos project at Apache, vendors and users will have a neutral community free from corporate governance restrictions, with which to collaborate and accelerate the development of a platform that provides wide benefits across the industry. As a flexible framework, we expect a wide variety of platforms to leverage the technology to fill specific niches and needs. == Current Status == Stratos has been in development since 2010 at WSO2, under the Apache License and under the Apache Way. Contribution to Apache, from which many of the core components are sourced, should be very straightforward. == Meritocracy == The contributors have a longstanding commitment and practice of meritocracy in their personal and professional capacities. Many of the committers on the existing project are already Apache Committers. == Community == The committers recognize the need to support more significant contributions from a broad swath of the industry and community. Apache is the obvious choice for a project that already embodies Apache values and is driven by developers who are committers on many other Apache projects. Stratos has appeal, and should attract community members, from among a number of constituencies: * Private PaaS deployment within an enterprise to benefit enterprise application deployment. * PaaS providers wishing to leverage a PaaS toolkit to build a PaaS
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On Thu, Jun 13, 2013 at 5:02 AM, Marvin Humphrey mar...@rectangular.comwrote: === Mailing Lists === 1. d...@stratos.incubator.apache.org - for developer/user discussions, JIRA change notifications continuous build/test notifications 2. comm...@stratos.incubator.apache.org - for commit mails It's a technicality, but the podling would also need a private@ list. Since there's only a combined dev/user list to start with though, obviously some thought has gone into this. :) When we initially created the list of required mailing lists, we requested for user dev lists. We spoke to a few people who have experience with bringing projects into the incubator, and were advised to start with a combined list so that all discussions happen in one place, and when there is a significant user community involvement, we could decide to create a user list. I will go ahead add a private@ list. Thanks Azeez * *
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
Good catch on the missing private@ list - thanks Ross On 13 June 2013 01:43, Afkham Azeez afk...@gmail.com wrote: On Thu, Jun 13, 2013 at 5:02 AM, Marvin Humphrey mar...@rectangular.comwrote: === Mailing Lists === 1. d...@stratos.incubator.apache.org - for developer/user discussions, JIRA change notifications continuous build/test notifications 2. comm...@stratos.incubator.apache.org - for commit mails It's a technicality, but the podling would also need a private@ list. Since there's only a combined dev/user list to start with though, obviously some thought has gone into this. :) When we initially created the list of required mailing lists, we requested for user dev lists. We spoke to a few people who have experience with bringing projects into the incubator, and were advised to start with a combined list so that all discussions happen in one place, and when there is a significant user community involvement, we could decide to create a user list. I will go ahead add a private@ list. Thanks Azeez * * - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
So here's a thought... There have been many discussions about different ways to incubate projects. One of the most radical ideas is to dismantle the incubator and replace the podling concept with probationary TLPs reporting to the board. As readers of this list will know I do not support the idea of dismantling the IPMC. I believe it does a great job that is not easily replaced by a board of nine directors. However, I have always acknowledged that the idea has merit under a certain set of circumstances. For me those circumstances are present in the Apache Stratos proposal. That is there are sufficient mentors and initial committers who are ASF Members that we can be reasonably certain that this project will succeed here at the ASF. I would therefore like to propose that we use Apache Stratos as a test case for the probationary TLP idea. I've already talked to Chris (who is driving the deconstruct the IPMC case) and Ant (who is less keen on dismantling the IPMC but wants to see how a probationary TLP model will play out). Both have agreed to help with this experiment if the IPMC and the Board wish it to proceed. I have not, however, discussed it with all the initial comitters or even mentors - I'm expecting them to speak up now. For my part my intention is to get the project set-up and then dissolve into the background. I do not intend to monitor the project on a day-to-day basis. However, I do promise to help pick up the pieces if the experiment should go horribly wrong. Of course running a single experiment will only allow us to define the incubation process for probationary TLPs, It is not going to solve all the problems Chris sees in the IPMC. However it will give us an opportunity to define the process, ask the board to approve this process and thus lay the foundations for other projects wishing to follow this path. So, what do you think? Ross On 11 June 2013 10:10, Ross Gardler rgard...@opendirective.com wrote: It's with great pleasure that I invite the IPMC to review a new proposal [1] for the Apache Incubator. Please let us know if you have any questions or comments - as you will see there are plenty of people on the initial commit list ready and willing to answer your questions. I copy the full text of the proposal for your convenience: = Stratos - A PaaS Framework = == Abstract == Stratos will be a polyglot [[http://www.gartner.com/it-glossary/platform-as-a-service-paas|PaaS]] framework, providing developers a cloud-based environment for developing, testing, and running scalable applications, and IT providers high utilization rates, automated resource management, and platform-wide insight including monitoring and billing. == Proposal == The Stratos PaaS framework will encompass four layers: 1. An [[http://www.gartner.com/it-glossary/infrastructure-as-a-service-iaas/|IaaS]]-agnostic layer that can interface with a wide variety of IaaS systems to provide elastic resources, and for multiple IaaS infrastructures to be automated at one time (hybrid clouds.) 2. A PaaS Controller with a cloud controller that automates and monitors IaaS runtime interactions, distributes artifacts to the underlying runtimes, deploys workloads, directs runtime traffic to the right runtimes using a tenant-aware elastic load balancer, and provides a portal for monitoring and provisioning of tenants on the system. 3. Foundational Services including security, logging, messaging, registry, storage (relational, file, and noSQL), task management, and billing. Foundational services will be loosely-coupled to allow swapping in alternate foundational services. 4. A Cartridge Architecture allowing frameworks, servers, and other runtimes to participate in the advantages of the system. The Cartridge Architecture must support multi-tenant workloads, and provide for various levels of tenant isolation and policy-based control over provisioning. Together these layers offer a foundational layer upon which applications and middleware frameworks can be deployed to speed time-to-market and simplify the development of scalable applications, as well as provide a high level of resource sharing and centralized management that can deliver lowest resource, infrastructure, and management costs. == Background == The Stratos Project has been under development[a] at http://wso2.org under the Apache 2.0 license and the Apache Way governance model since 2010. It initially was focussed on providing PaaS benefits to the users of WSO2 Carbon middleware platform. In version 2.0, to be released in summer 2013, extensive work has been done to clearly separate out the PaaS framework from the products (cartridges) that run on top of it. Stratos now has the ability to run arbitrary workloads, including Java, PHP, MySQL, Jetty, Tomcat, and many more. == Rationale == PaaS is in demand by enterprises and organizations of all sizes. The drive towards instance provisioning, high resource
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
It'd help to know concretely what is meant by a probationary TLP, particularly what is different about it from normal incubation. I am not looking for yet another email discussion, but an URL to a wiki page would be nice. From: Ross Gardler rgard...@opendirective.com To: general@incubator.apache.org Sent: Wednesday, June 12, 2013 10:12 PM Subject: Re: [DISCUSS] Accept Stratos as an Apache Incubation Project So here's a thought... There have been many discussions about different ways to incubate projects. One of the most radical ideas is to dismantle the incubator and replace the podling concept with probationary TLPs reporting to the board. As readers of this list will know I do not support the idea of dismantling the IPMC. I believe it does a great job that is not easily replaced by a board of nine directors. However, I have always acknowledged that the idea has merit under a certain set of circumstances. For me those circumstances are present in the Apache Stratos proposal. That is there are sufficient mentors and initial committers who are ASF Members that we can be reasonably certain that this project will succeed here at the ASF. I would therefore like to propose that we use Apache Stratos as a test case for the probationary TLP idea. I've already talked to Chris (who is driving the deconstruct the IPMC case) and Ant (who is less keen on dismantling the IPMC but wants to see how a probationary TLP model will play out). Both have agreed to help with this experiment if the IPMC and the Board wish it to proceed. I have not, however, discussed it with all the initial comitters or even mentors - I'm expecting them to speak up now. For my part my intention is to get the project set-up and then dissolve into the background. I do not intend to monitor the project on a day-to-day basis. However, I do promise to help pick up the pieces if the experiment should go horribly wrong. Of course running a single experiment will only allow us to define the incubation process for probationary TLPs, It is not going to solve all the problems Chris sees in the IPMC. However it will give us an opportunity to define the process, ask the board to approve this process and thus lay the foundations for other projects wishing to follow this path. So, what do you think? Ross On 11 June 2013 10:10, Ross Gardler rgard...@opendirective.com wrote: It's with great pleasure that I invite the IPMC to review a new proposal [1] for the Apache Incubator. Please let us know if you have any questions or comments - as you will see there are plenty of people on the initial commit list ready and willing to answer your questions. I copy the full text of the proposal for your convenience: = Stratos - A PaaS Framework = == Abstract == Stratos will be a polyglot [[http://www.gartner.com/it-glossary/platform-as-a-service-paas|PaaS]] framework, providing developers a cloud-based environment for developing, testing, and running scalable applications, and IT providers high utilization rates, automated resource management, and platform-wide insight including monitoring and billing. == Proposal == The Stratos PaaS framework will encompass four layers: 1. An [[http://www.gartner.com/it-glossary/infrastructure-as-a-service-iaas/|IaaS]]-agnostic layer that can interface with a wide variety of IaaS systems to provide elastic resources, and for multiple IaaS infrastructures to be automated at one time (hybrid clouds.) 2. A PaaS Controller with a cloud controller that automates and monitors IaaS runtime interactions, distributes artifacts to the underlying runtimes, deploys workloads, directs runtime traffic to the right runtimes using a tenant-aware elastic load balancer, and provides a portal for monitoring and provisioning of tenants on the system. 3. Foundational Services including security, logging, messaging, registry, storage (relational, file, and noSQL), task management, and billing. Foundational services will be loosely-coupled to allow swapping in alternate foundational services. 4. A Cartridge Architecture allowing frameworks, servers, and other runtimes to participate in the advantages of the system. The Cartridge Architecture must support multi-tenant workloads, and provide for various levels of tenant isolation and policy-based control over provisioning. Together these layers offer a foundational layer upon which applications and middleware frameworks can be deployed to speed time-to-market and simplify the development of scalable applications, as well as provide a high level of resource sharing and centralized management that can deliver lowest resource, infrastructure, and management costs. == Background == The Stratos Project has been under development[a] at http://wso2.org under the Apache 2.0 license and the Apache Way governance model since 2010. It initially was focussed on providing PaaS benefits to the users of WSO2 Carbon
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
Chris' proposal is at http://wiki.apache.org/incubator/IncubatorDeconstructionProposal This will need some work before it can be made real but I think there is enough meat on the bones to understand the concept. Ross On 13 June 2013 03:16, Joe Schaefer joe_schae...@yahoo.com wrote: It'd help to know concretely what is meant by a probationary TLP, particularly what is different about it from normal incubation. I am not looking for yet another email discussion, but an URL to a wiki page would be nice. From: Ross Gardler rgard...@opendirective.com To: general@incubator.apache.org Sent: Wednesday, June 12, 2013 10:12 PM Subject: Re: [DISCUSS] Accept Stratos as an Apache Incubation Project So here's a thought... There have been many discussions about different ways to incubate projects. One of the most radical ideas is to dismantle the incubator and replace the podling concept with probationary TLPs reporting to the board. As readers of this list will know I do not support the idea of dismantling the IPMC. I believe it does a great job that is not easily replaced by a board of nine directors. However, I have always acknowledged that the idea has merit under a certain set of circumstances. For me those circumstances are present in the Apache Stratos proposal. That is there are sufficient mentors and initial committers who are ASF Members that we can be reasonably certain that this project will succeed here at the ASF. I would therefore like to propose that we use Apache Stratos as a test case for the probationary TLP idea. I've already talked to Chris (who is driving the deconstruct the IPMC case) and Ant (who is less keen on dismantling the IPMC but wants to see how a probationary TLP model will play out). Both have agreed to help with this experiment if the IPMC and the Board wish it to proceed. I have not, however, discussed it with all the initial comitters or even mentors - I'm expecting them to speak up now. For my part my intention is to get the project set-up and then dissolve into the background. I do not intend to monitor the project on a day-to-day basis. However, I do promise to help pick up the pieces if the experiment should go horribly wrong. Of course running a single experiment will only allow us to define the incubation process for probationary TLPs, It is not going to solve all the problems Chris sees in the IPMC. However it will give us an opportunity to define the process, ask the board to approve this process and thus lay the foundations for other projects wishing to follow this path. So, what do you think? Ross On 11 June 2013 10:10, Ross Gardler rgard...@opendirective.com wrote: It's with great pleasure that I invite the IPMC to review a new proposal [1] for the Apache Incubator. Please let us know if you have any questions or comments - as you will see there are plenty of people on the initial commit list ready and willing to answer your questions. I copy the full text of the proposal for your convenience: = Stratos - A PaaS Framework = == Abstract == Stratos will be a polyglot [[http://www.gartner.com/it-glossary/platform-as-a-service-paas|PaaS]] framework, providing developers a cloud-based environment for developing, testing, and running scalable applications, and IT providers high utilization rates, automated resource management, and platform-wide insight including monitoring and billing. == Proposal == The Stratos PaaS framework will encompass four layers: 1. An [[http://www.gartner.com/it-glossary/infrastructure-as-a-service-iaas/|IaaS]]-agnostic layer that can interface with a wide variety of IaaS systems to provide elastic resources, and for multiple IaaS infrastructures to be automated at one time (hybrid clouds.) 2. A PaaS Controller with a cloud controller that automates and monitors IaaS runtime interactions, distributes artifacts to the underlying runtimes, deploys workloads, directs runtime traffic to the right runtimes using a tenant-aware elastic load balancer, and provides a portal for monitoring and provisioning of tenants on the system. 3. Foundational Services including security, logging, messaging, registry, storage (relational, file, and noSQL), task management, and billing. Foundational services will be loosely-coupled to allow swapping in alternate foundational services. 4. A Cartridge Architecture allowing frameworks, servers, and other runtimes to participate in the advantages of the system. The Cartridge Architecture must support multi-tenant workloads, and provide for various levels of tenant isolation and policy-based control over provisioning. Together these layers offer a foundational layer upon which applications and middleware frameworks can be deployed to speed time-to-market and simplify the development of scalable applications, as well as provide a high level of resource sharing and centralized management that can deliver lowest resource
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
Hi Ross, + 1 for considering this project for the probationary TLP experiment. Here are my 2 cents: * The initial committers list on the proposal demonstrates diversity. * The proposal building process was inclusive (a glimpse of wiki history shows some of it). With enough outsiders (who were not part of the code donation organization), I can see the community building already. * Along with 7 nominated mentors there are significant number of ASF and IPMC members on the initial committers list, so I can trust the PMC to get the releases right. Ofcourse there will be few tasks like podling name search which needs to be done upfront, but other wise I see no issues wit this experiment. I can foresee this project will quickly cruise incubation, so if the board agrees why not get it jumpstarted? Cheers, Suresh On Jun 12, 2013, at 10:12 PM, Ross Gardler rgard...@opendirective.com wrote: So here's a thought... There have been many discussions about different ways to incubate projects. One of the most radical ideas is to dismantle the incubator and replace the podling concept with probationary TLPs reporting to the board. As readers of this list will know I do not support the idea of dismantling the IPMC. I believe it does a great job that is not easily replaced by a board of nine directors. However, I have always acknowledged that the idea has merit under a certain set of circumstances. For me those circumstances are present in the Apache Stratos proposal. That is there are sufficient mentors and initial committers who are ASF Members that we can be reasonably certain that this project will succeed here at the ASF. I would therefore like to propose that we use Apache Stratos as a test case for the probationary TLP idea. I've already talked to Chris (who is driving the deconstruct the IPMC case) and Ant (who is less keen on dismantling the IPMC but wants to see how a probationary TLP model will play out). Both have agreed to help with this experiment if the IPMC and the Board wish it to proceed. I have not, however, discussed it with all the initial comitters or even mentors - I'm expecting them to speak up now. For my part my intention is to get the project set-up and then dissolve into the background. I do not intend to monitor the project on a day-to-day basis. However, I do promise to help pick up the pieces if the experiment should go horribly wrong. Of course running a single experiment will only allow us to define the incubation process for probationary TLPs, It is not going to solve all the problems Chris sees in the IPMC. However it will give us an opportunity to define the process, ask the board to approve this process and thus lay the foundations for other projects wishing to follow this path. So, what do you think? Ross On 11 June 2013 10:10, Ross Gardler rgard...@opendirective.com wrote: It's with great pleasure that I invite the IPMC to review a new proposal [1] for the Apache Incubator. Please let us know if you have any questions or comments - as you will see there are plenty of people on the initial commit list ready and willing to answer your questions. I copy the full text of the proposal for your convenience: = Stratos - A PaaS Framework = == Abstract == Stratos will be a polyglot [[http://www.gartner.com/it-glossary/platform-as-a-service-paas|PaaS]] framework, providing developers a cloud-based environment for developing, testing, and running scalable applications, and IT providers high utilization rates, automated resource management, and platform-wide insight including monitoring and billing. == Proposal == The Stratos PaaS framework will encompass four layers: 1. An [[http://www.gartner.com/it-glossary/infrastructure-as-a-service-iaas/|IaaS]]-agnostic layer that can interface with a wide variety of IaaS systems to provide elastic resources, and for multiple IaaS infrastructures to be automated at one time (hybrid clouds.) 2. A PaaS Controller with a cloud controller that automates and monitors IaaS runtime interactions, distributes artifacts to the underlying runtimes, deploys workloads, directs runtime traffic to the right runtimes using a tenant-aware elastic load balancer, and provides a portal for monitoring and provisioning of tenants on the system. 3. Foundational Services including security, logging, messaging, registry, storage (relational, file, and noSQL), task management, and billing. Foundational services will be loosely-coupled to allow swapping in alternate foundational services. 4. A Cartridge Architecture allowing frameworks, servers, and other runtimes to participate in the advantages of the system. The Cartridge Architecture must support multi-tenant workloads, and provide for various levels of tenant isolation and policy-based control over provisioning. Together these layers offer a foundational layer upon which applications and middleware frameworks can be
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
ccing trademarks@ Note the mix of public and private lists. On Wed, Jun 12, 2013 at 8:09 PM, Sanjiva Weerawarana sanj...@wso2.com wrote: On Thu, Jun 13, 2013 at 5:02 AM, Marvin Humphrey mar...@rectangular.comwrote: == Known Risks == Stratos has largely been developed by sponsored developers employed at a single organization - WSO2. Seeking a broader community of contributors is a top goal of contributing Stratos to Apache. WSO2 plans to continue to offer services and commercial support packages for Stratos, so there is a financial incentive to broaden Stratos’ appeal. This may provide the misinterpretation that Stratos remains merely a WSO2 technology. However, WSO2’s main business strategy is to build and support higher level PaaS offerings (including the WSO2 middleware stack) on top of a common PaaS framework, as provided by Stratos. This includes a WSO2 StratosLive option which is a public PaaS based on WSO2 Stratos. Kudos for the honest self-analysis. Thanks :-). I wonder whether Apache Brand Management would cry foul on a trademark like StratosLive if such a product were to appear later because of the confusingly similar name guideline. We've seen project founders leave and compete with ASF products while using confusingly similar names before, e.g. CouchBase. Yeah we also thought about that .. if StratosLive is too close to home we can certainly change it. Bit painful but not impossible. Better to get an answer to the question up front IMO. If a top goal is seeking a broader community of contributors, the project might benefit if it could find a way to reassure potential contributors on this matter. Perhaps the name for the WS02 product should be vetted in advance? Yep can do - our current thinking is to have a WSO2 distribution that contains Apache Stratos and includes all of our middleware products (app server, esb, etc. etc.). That is, a PaaS built on the PaaS framework from Apache. The thinking is to not give that a specific name but rather call it the WSO2 distribution of Apache Stratos (or something like that). I don't think that's one of the 'sanctioned' uses of marks from: In fact, per the below page, it's explicitly forbidden. http://www.apache.org/foundation/marks/faq/#products Shane do you have thoughts on that? That idea is based on what some others do and it seems to me is clear enough that its not just the Apache code being redistributed but has more stuff. Analogous to Redhat being a Linux distribution. Cheers, Sanjiva. -- Sanjiva Weerawarana, Ph.D. Founder, Chairman CEO; WSO2, Inc.; http://wso2.com/ email: sanj...@wso2.com; phone: +94 11 763 9614; cell: +94 77 787 6880 | +1 650 265 8311 blog: http://sanjiva.weerawarana.org/ Lean . Enterprise . Middleware --David - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On Thu, Jun 13, 2013 at 8:27 AM, David Nalley da...@gnsa.us wrote: ccing trademarks@ Note the mix of public and private lists. Yeah we also thought about that .. if StratosLive is too close to home we can certainly change it. Bit painful but not impossible. Better to get an answer to the question up front IMO. +1. The thinking is to not give that a specific name but rather call it the WSO2 distribution of Apache Stratos (or something like that). I don't think that's one of the 'sanctioned' uses of marks from: In fact, per the below page, it's explicitly forbidden. http://www.apache.org/foundation/marks/faq/#products Hmm that idea was based on how Cloudera is distributing Hadoop: http://www.cloudera.com/content/cloudera/en/products/cdh.html Is that also incorrect then? IIRC they've been doing that for years and presumably ASF is aware of it? Sanjiva. -- Sanjiva Weerawarana, Ph.D. Founder, Chairman CEO; WSO2, Inc.; http://wso2.com/ email: sanj...@wso2.com; phone: +94 11 763 9614; cell: +94 77 787 6880 | +1 650 265 8311 blog: http://sanjiva.weerawarana.org/ Lean . Enterprise . Middleware
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
On Jun 12, 2013, at 7:12 PM, Ross Gardler rgard...@opendirective.com wrote: So here's a thought... There have been many discussions about different ways to incubate projects. One of the most radical ideas is to dismantle the incubator and replace the podling concept with probationary TLPs reporting to the board. As readers of this list will know I do not support the idea of dismantling the IPMC. I believe it does a great job that is not easily replaced by a board of nine directors. However, I have always acknowledged that the idea has merit under a certain set of circumstances. For me those circumstances are present in the Apache Stratos proposal. That is there are sufficient mentors and initial committers who are ASF Members that we can be reasonably certain that this project will succeed here at the ASF. I would therefore like to propose that we use Apache Stratos as a test case for the probationary TLP idea. I've already talked to Chris (who is driving the deconstruct the IPMC case) and Ant (who is less keen on dismantling the IPMC but wants to see how a probationary TLP model will play out). Both have agreed to help with this experiment if the IPMC and the Board wish it to proceed. I have not, however, discussed it with all the initial comitters or even mentors - I'm expecting them to speak up now. For my part my intention is to get the project set-up and then dissolve into the background. I do not intend to monitor the project on a day-to-day basis. However, I do promise to help pick up the pieces if the experiment should go horribly wrong. Of course running a single experiment will only allow us to define the incubation process for probationary TLPs, It is not going to solve all the problems Chris sees in the IPMC. However it will give us an opportunity to define the process, ask the board to approve this process and thus lay the foundations for other projects wishing to follow this path. So, what do you think? I don't see the need to force Stratos through the Incubator given the current proposed membership. Some points: Who's responsible for monitoring the probation, the IPMC or the board? I think it should be the IPMC. What bits must absolutely be done before probation begins? What minimum criteria does a probationary TLP have to meet to stay in good graces? What happens if the probationary TLP is not in good graces? What bits must absolutely be done before probation completes? Fleshing out these and, I'm sure, others' concerns on a wiki, as Joe pointed out, would be a great idea. Regards, Alan
[DISCUSS] Accept Stratos as an Apache Incubation Project
It's with great pleasure that I invite the IPMC to review a new proposal [1] for the Apache Incubator. Please let us know if you have any questions or comments - as you will see there are plenty of people on the initial commit list ready and willing to answer your questions. I copy the full text of the proposal for your convenience: = Stratos - A PaaS Framework = == Abstract == Stratos will be a polyglot [[http://www.gartner.com/it-glossary/platform-as-a-service-paas|PaaS]] framework, providing developers a cloud-based environment for developing, testing, and running scalable applications, and IT providers high utilization rates, automated resource management, and platform-wide insight including monitoring and billing. == Proposal == The Stratos PaaS framework will encompass four layers: 1. An [[http://www.gartner.com/it-glossary/infrastructure-as-a-service-iaas/|IaaS]]-agnostic layer that can interface with a wide variety of IaaS systems to provide elastic resources, and for multiple IaaS infrastructures to be automated at one time (hybrid clouds.) 2. A PaaS Controller with a cloud controller that automates and monitors IaaS runtime interactions, distributes artifacts to the underlying runtimes, deploys workloads, directs runtime traffic to the right runtimes using a tenant-aware elastic load balancer, and provides a portal for monitoring and provisioning of tenants on the system. 3. Foundational Services including security, logging, messaging, registry, storage (relational, file, and noSQL), task management, and billing. Foundational services will be loosely-coupled to allow swapping in alternate foundational services. 4. A Cartridge Architecture allowing frameworks, servers, and other runtimes to participate in the advantages of the system. The Cartridge Architecture must support multi-tenant workloads, and provide for various levels of tenant isolation and policy-based control over provisioning. Together these layers offer a foundational layer upon which applications and middleware frameworks can be deployed to speed time-to-market and simplify the development of scalable applications, as well as provide a high level of resource sharing and centralized management that can deliver lowest resource, infrastructure, and management costs. == Background == The Stratos Project has been under development[a] at http://wso2.org under the Apache 2.0 license and the Apache Way governance model since 2010. It initially was focussed on providing PaaS benefits to the users of WSO2 Carbon middleware platform. In version 2.0, to be released in summer 2013, extensive work has been done to clearly separate out the PaaS framework from the products (cartridges) that run on top of it. Stratos now has the ability to run arbitrary workloads, including Java, PHP, MySQL, Jetty, Tomcat, and many more. == Rationale == PaaS is in demand by enterprises and organizations of all sizes. The drive towards instance provisioning, high resource utilization and thus low cost, combined with a wide platform of general-purpose services to build on, PaaS has the opportunity to accelerate the development cycle and innovation index of a new class of applications, services, and business models. PaaS offerings are widely diversified but largely associated with powerful corporate interests. With the commencement of the Stratos project at Apache, vendors and users will have a neutral community free from corporate governance restrictions, with which to collaborate and accelerate the development of a platform that provides wide benefits across the industry. As a flexible framework, we expect a wide variety of platforms to leverage the technology to fill specific niches and needs. == Current Status == Stratos has been in development since 2010 at WSO2, under the Apache License and under the Apache Way. Contribution to Apache, from which many of the core components are sourced, should be very straightforward. == Meritocracy == The contributors have a longstanding commitment and practice of meritocracy in their personal and professional capacities. Many of the committers on the existing project are already Apache Committers. == Community == The committers recognize the need to support more significant contributions from a broad swath of the industry and community. Apache is the obvious choice for a project that already embodies Apache values and is driven by developers who are committers on many other Apache projects. Stratos has appeal, and should attract community members, from among a number of constituencies: * Private PaaS deployment within an enterprise to benefit enterprise application deployment. * PaaS providers wishing to leverage a PaaS toolkit to build a PaaS customized to their particular vertical industry, ecosystem, or internal development processes. * SaaS providers wishing to build upon a widely-deployed and supported elastic, multi-tenant platform. * ISPs wishing to offer hosted application services with higher
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
Apologies. I just realise I never added Ant Elder to the mentor list. I've added him to the wiki, please be aware he is not in the list of mentors on the proposal pasted below. Ross On 11 June 2013 10:10, Ross Gardler rgard...@opendirective.com wrote: It's with great pleasure that I invite the IPMC to review a new proposal [1] for the Apache Incubator. Please let us know if you have any questions or comments - as you will see there are plenty of people on the initial commit list ready and willing to answer your questions. I copy the full text of the proposal for your convenience: = Stratos - A PaaS Framework = == Abstract == Stratos will be a polyglot [[http://www.gartner.com/it-glossary/platform-as-a-service-paas|PaaS]] framework, providing developers a cloud-based environment for developing, testing, and running scalable applications, and IT providers high utilization rates, automated resource management, and platform-wide insight including monitoring and billing. == Proposal == The Stratos PaaS framework will encompass four layers: 1. An [[http://www.gartner.com/it-glossary/infrastructure-as-a-service-iaas/|IaaS]]-agnostic layer that can interface with a wide variety of IaaS systems to provide elastic resources, and for multiple IaaS infrastructures to be automated at one time (hybrid clouds.) 2. A PaaS Controller with a cloud controller that automates and monitors IaaS runtime interactions, distributes artifacts to the underlying runtimes, deploys workloads, directs runtime traffic to the right runtimes using a tenant-aware elastic load balancer, and provides a portal for monitoring and provisioning of tenants on the system. 3. Foundational Services including security, logging, messaging, registry, storage (relational, file, and noSQL), task management, and billing. Foundational services will be loosely-coupled to allow swapping in alternate foundational services. 4. A Cartridge Architecture allowing frameworks, servers, and other runtimes to participate in the advantages of the system. The Cartridge Architecture must support multi-tenant workloads, and provide for various levels of tenant isolation and policy-based control over provisioning. Together these layers offer a foundational layer upon which applications and middleware frameworks can be deployed to speed time-to-market and simplify the development of scalable applications, as well as provide a high level of resource sharing and centralized management that can deliver lowest resource, infrastructure, and management costs. == Background == The Stratos Project has been under development[a] at http://wso2.org under the Apache 2.0 license and the Apache Way governance model since 2010. It initially was focussed on providing PaaS benefits to the users of WSO2 Carbon middleware platform. In version 2.0, to be released in summer 2013, extensive work has been done to clearly separate out the PaaS framework from the products (cartridges) that run on top of it. Stratos now has the ability to run arbitrary workloads, including Java, PHP, MySQL, Jetty, Tomcat, and many more. == Rationale == PaaS is in demand by enterprises and organizations of all sizes. The drive towards instance provisioning, high resource utilization and thus low cost, combined with a wide platform of general-purpose services to build on, PaaS has the opportunity to accelerate the development cycle and innovation index of a new class of applications, services, and business models. PaaS offerings are widely diversified but largely associated with powerful corporate interests. With the commencement of the Stratos project at Apache, vendors and users will have a neutral community free from corporate governance restrictions, with which to collaborate and accelerate the development of a platform that provides wide benefits across the industry. As a flexible framework, we expect a wide variety of platforms to leverage the technology to fill specific niches and needs. == Current Status == Stratos has been in development since 2010 at WSO2, under the Apache License and under the Apache Way. Contribution to Apache, from which many of the core components are sourced, should be very straightforward. == Meritocracy == The contributors have a longstanding commitment and practice of meritocracy in their personal and professional capacities. Many of the committers on the existing project are already Apache Committers. == Community == The committers recognize the need to support more significant contributions from a broad swath of the industry and community. Apache is the obvious choice for a project that already embodies Apache values and is driven by developers who are committers on many other Apache projects. Stratos has appeal, and should attract community members, from among a number of constituencies: * Private PaaS deployment within an enterprise to benefit enterprise application
Re: [DISCUSS] Accept Stratos as an Apache Incubation Project
Impressive proposal. Reminds me on RH OpenShift actually. Please note the first two links are broken in gmail. They are: * http://www.gartner.com/it-glossary/platform-as-a-service-paas * http://www.gartner.com/it-glossary/infrastructure-as-a-service-iaas Anyway, proposal looks really good for me. On Tue, Jun 11, 2013 at 11:10 AM, Ross Gardler rgard...@opendirective.com wrote: It's with great pleasure that I invite the IPMC to review a new proposal [1] for the Apache Incubator. Please let us know if you have any questions or comments - as you will see there are plenty of people on the initial commit list ready and willing to answer your questions. I copy the full text of the proposal for your convenience: = Stratos - A PaaS Framework = == Abstract == Stratos will be a polyglot [[http://www.gartner.com/it-glossary/platform-as-a-service-paas|PaaS]] framework, providing developers a cloud-based environment for developing, testing, and running scalable applications, and IT providers high utilization rates, automated resource management, and platform-wide insight including monitoring and billing. == Proposal == The Stratos PaaS framework will encompass four layers: 1. An [[http://www.gartner.com/it-glossary/infrastructure-as-a-service-iaas/|IaaS]]-agnostic layer that can interface with a wide variety of IaaS systems to provide elastic resources, and for multiple IaaS infrastructures to be automated at one time (hybrid clouds.) 2. A PaaS Controller with a cloud controller that automates and monitors IaaS runtime interactions, distributes artifacts to the underlying runtimes, deploys workloads, directs runtime traffic to the right runtimes using a tenant-aware elastic load balancer, and provides a portal for monitoring and provisioning of tenants on the system. 3. Foundational Services including security, logging, messaging, registry, storage (relational, file, and noSQL), task management, and billing. Foundational services will be loosely-coupled to allow swapping in alternate foundational services. 4. A Cartridge Architecture allowing frameworks, servers, and other runtimes to participate in the advantages of the system. The Cartridge Architecture must support multi-tenant workloads, and provide for various levels of tenant isolation and policy-based control over provisioning. Together these layers offer a foundational layer upon which applications and middleware frameworks can be deployed to speed time-to-market and simplify the development of scalable applications, as well as provide a high level of resource sharing and centralized management that can deliver lowest resource, infrastructure, and management costs. == Background == The Stratos Project has been under development[a] at http://wso2.org under the Apache 2.0 license and the Apache Way governance model since 2010. It initially was focussed on providing PaaS benefits to the users of WSO2 Carbon middleware platform. In version 2.0, to be released in summer 2013, extensive work has been done to clearly separate out the PaaS framework from the products (cartridges) that run on top of it. Stratos now has the ability to run arbitrary workloads, including Java, PHP, MySQL, Jetty, Tomcat, and many more. == Rationale == PaaS is in demand by enterprises and organizations of all sizes. The drive towards instance provisioning, high resource utilization and thus low cost, combined with a wide platform of general-purpose services to build on, PaaS has the opportunity to accelerate the development cycle and innovation index of a new class of applications, services, and business models. PaaS offerings are widely diversified but largely associated with powerful corporate interests. With the commencement of the Stratos project at Apache, vendors and users will have a neutral community free from corporate governance restrictions, with which to collaborate and accelerate the development of a platform that provides wide benefits across the industry. As a flexible framework, we expect a wide variety of platforms to leverage the technology to fill specific niches and needs. == Current Status == Stratos has been in development since 2010 at WSO2, under the Apache License and under the Apache Way. Contribution to Apache, from which many of the core components are sourced, should be very straightforward. == Meritocracy == The contributors have a longstanding commitment and practice of meritocracy in their personal and professional capacities. Many of the committers on the existing project are already Apache Committers. == Community == The committers recognize the need to support more significant contributions from a broad swath of the industry and community. Apache is the obvious choice for a project that already embodies Apache values and is driven by developers who are committers on many other Apache projects. Stratos has appeal, and should attract community members,