Hi Gunnar.
Thanks for your reply.
What I am trying to figure out is if Emily's message (We let you off the hook) was on behalf of the steering committee (or who/whatever else needed to approve the release) and was directed at me or the graphql workgroup.
If she speaks in her own capacity, then my question is why does she have the power to let anyone / any spec group of any hook ??
If she speaks on behalf of the body that is allowing the release to happen, I must assume the body had meeting / discussion, and decide that we (MicroProfile GraphQL Workgroup) or me (Phillip) are totally wrong in that this is a service release, but, they will be so kind as to let this one slide. (But don't let this happen again !)
So I am less concerned about the process. Although I think there are flaws in the process, like the fact that we can never ever do any fixing to what is currently the majority of versions of spec in current users projects and vendor implementation footprints in the market, and that makes it impossible to do long contracted support on a certain version of a runtime that implements these specs. That is beside the point. I assume that changing the process/policy will take much longer than just releasing a minor version (even if it's technically a service release). That is basically what myself and Keven agreed after he pointed this out in the process. And for GraphQL, that is fine with me as we do not yet have any supported products for this.
My concern is more with the attitude and culture busy forming in MicroProfile around this new process. So again, my question was to clarify under what capacity Emily is letting me or us (also need clarification) of the hook. My concern is that if we are moving to a culture (irrelevant of the process) where we have police that will let the citizens off the hook for their stupid mistakes, that this is not a comunitee that I want to be part of. I also believe this is not what we want, but this is a warning to everyone that we need to be careful on what culture is being busy forming around this new process. My suggestion above can still work within the bounds of the process. Assuming that Emily spoke on behalf of the approval body, and addressed me as a representation of the workgroup, she could have rather checked with Andy (IBM representation on the workgroup) on why this is a service release. In fact, they could have discussed it much earlier in the process if she wanted to be more involved. she could also have asked for a clarification. But she let me off the hook.
There has been no official request (or if there was I missed it) from the body, that we should explain to them why this is a service release. Just induvidual comments on how we differ from what a service release is. This GraphQL release contains zero changes that impact the user. If Emily or Keven, in personal or Steering committee capacity, cared to check in with their representation on the workgroup, he could have explained this to them. (Basically my suggestion above). The process can still continue as is.
But, again, my concern is with the attitude / culture. And your last comment confirms my concern.
> I sense that you are refusing to work with the community. I'd like to say that this is not something we will accept as the PMC.
Let's start with OpenAPI. There is NO Community. This spec currently is dead. There is no involvement from any vendor or community to take this spec forward. I have been joining the bi-weeky meeting for more than a year. Mostly it was me, then after 5 or 10 mins I left the call. Sometimes I was joined by other Red Hatters.
W.r.t GraphQL, I think (I might be wrong) that we are currently the most active community of any of the specs in MicroProfile. We have regular, well attended meetings where we discuss how this spec will go forward. I personally have done a lot of work with the MicroProfile GraphQL community, represented by other Vendors and interensted members in their personal capacity. See
https://github.com/eclipse/microprofile-graphql/graphs/contributors. I have never refused to do any work and have been putting my hand up more often than not, to move this spec forward. This release in question I am not doing for me or Red Hat. I am helping out Oracle.
Most of the work I have done for this Workgroup has been done in my free time, as during my work time (i.e what I get paid to do) I mostly work on the implementations and runtimes. The two releases currently in question were done on a Saturday morning, and again (when I had to redo them because of the process explained above) on Thuesday night. So when I say that I (Phillip) will not be doing any future releases, I am saying I am putting my hand down.
So I would like to say that the PMC is welcome to not accept my work that I do for free during my personal time. Again I would warn that this is not the attitude that will build a vibrant community. We need to be careful how we enforce this new policy.
I hope that this makes it more clear. Happy to discuss this further.
