Policy question: What happens when projects merge?


Nick Young
 

Hi TOC, and everyone else,


With the successful launch of Envoy Gateway, there's a question about what to do about Contour's lifecycle as an incubated project in the CNCF.


I firmly believe that Envoy Gateway's launch is a huge win for the cloud native landscape, and the deduplication of effort possible by having us all working together will prove worthwhile once we get something built.


Normally, there's a (reasonable) requirement for incubating projects to show motion in the direction of becoming graduated, and a timeline to become graduated. But it seems to me that graduation implies a level of "this will be supported for the foreseeable future" that I don't know is viable for Contour.


VMware has committed to ensuring our maintenance of the project will be ongoing until users are ready to move away, see [our blog](https://blogs.vmware.com/opensource/2022/05/16/contour-and-community-build-new-envoy-gateway/) for more details. To summarize that blog, Contour's VMware maintainers will be helping to ensure that current users of Contour are looked after with features and support as long as possible, Wearing my Contour maintainer hat, we're an incubating project that will, within the next year or two, be mostly obsoleted by a functional, production-ready Envoy Gateway.  This implies that I need to figure out what we're going to do about Contour's incubation status.


The scenario I had assumed is that Contour may not graduate, and personally this makes me a little sad, but I recognize the larger opportunity that Envoy Gateway represents for the cloud native community. I know this situation hasn't come up before, but I suspect that this won't be the last time that CNCF projects interact in this fashion. I’d definitely like the TOC’s guidance here rather than making assumptions though.


Should Contour stay in incubating until Envoy Gateway is a viable alternative and users have moved away, then be archived? The timeline for this I would see as at least two years.


I personally see this as a successful exit for Contour, and have heard it referred to as "the open-source version of being acquired", but it would be nice for the TOC to give both us and other projects who may run into this situation in the future some guidance here.


Thanks for your time, everyone!


Nick Young

Wearing two of my maintainer hats for this email (Contour and Envoy Gateway).



Richard Hartmann
 

Something similar did happen in the past: As part of OpenTelemetry's
move from sandbox to incubation, the OpenTracing project agreed to
archive itself.

*Without having looked at the details, going just from what you described*:
While the project is maintained, even without substantial feature
work, it seems that no immediate action on project status is needed.
Personally, I would put deprecation notices and pointers to migration
guides in all the right places to avoid onboarding new users and
making offboarding existing ones more effective and efficient. Once
you see that a sunset is feasible send a timeline to your last users
and share it with TOC.

NB: It seems unlikely based on what you described, but the approach
above allows interested third parties to step up to become Contour
maintainers if they want to carry the project forward long-term.


Davanum Srinivas
 

Nick,

Thanks for reaching out. One of the main objectives of the process is to give assurance to the end users of CNCF projects that we as a community stand behind the projects they are using (esp graduated ones). So we cannot do this in good faith for Contour. So my feeling is that we should leave contour where it is now, put the effort on the replacement, figure out docs (esp migration etc) and slowly wean the community off contour onto the new replacement which is Envoy Gateway.

thanks,
Dims


On Tue, Jun 14, 2022 at 1:15 AM Nick Young <inocuo@...> wrote:

Hi TOC, and everyone else,


With the successful launch of Envoy Gateway, there's a question about what to do about Contour's lifecycle as an incubated project in the CNCF.


I firmly believe that Envoy Gateway's launch is a huge win for the cloud native landscape, and the deduplication of effort possible by having us all working together will prove worthwhile once we get something built.


Normally, there's a (reasonable) requirement for incubating projects to show motion in the direction of becoming graduated, and a timeline to become graduated. But it seems to me that graduation implies a level of "this will be supported for the foreseeable future" that I don't know is viable for Contour.


VMware has committed to ensuring our maintenance of the project will be ongoing until users are ready to move away, see [our blog](https://blogs.vmware.com/opensource/2022/05/16/contour-and-community-build-new-envoy-gateway/) for more details. To summarize that blog, Contour's VMware maintainers will be helping to ensure that current users of Contour are looked after with features and support as long as possible, Wearing my Contour maintainer hat, we're an incubating project that will, within the next year or two, be mostly obsoleted by a functional, production-ready Envoy Gateway.  This implies that I need to figure out what we're going to do about Contour's incubation status.


The scenario I had assumed is that Contour may not graduate, and personally this makes me a little sad, but I recognize the larger opportunity that Envoy Gateway represents for the cloud native community. I know this situation hasn't come up before, but I suspect that this won't be the last time that CNCF projects interact in this fashion. I’d definitely like the TOC’s guidance here rather than making assumptions though.


Should Contour stay in incubating until Envoy Gateway is a viable alternative and users have moved away, then be archived? The timeline for this I would see as at least two years.


I personally see this as a successful exit for Contour, and have heard it referred to as "the open-source version of being acquired", but it would be nice for the TOC to give both us and other projects who may run into this situation in the future some guidance here.


Thanks for your time, everyone!


Nick Young

Wearing two of my maintainer hats for this email (Contour and Envoy Gateway).




--
Davanum Srinivas :: https://twitter.com/dims


Evan Anderson
 

The OpenTracing and OpenCensus communities seem to have blazed the way here with OpenTelemetry -- I don't know if they have any learnings that they'd want to share about what worked well, or what could have been done better.

From: cncf-toc@... <cncf-toc@...> on behalf of Davanum Srinivas via lists.cncf.io <davanum=gmail.com@...>
Sent: Tuesday, June 14, 2022 3:27 AM
To: Nick Young <inocuo@...>
Cc: CNCF TOC <cncf-toc@...>
Subject: Re: [cncf-toc] Policy question: What happens when projects merge?
 
Nick,

Thanks for reaching out. One of the main objectives of the process is to give assurance to the end users of CNCF projects that we as a community stand behind the projects they are using (esp graduated ones). So we cannot do this in good faith for Contour. So my feeling is that we should leave contour where it is now, put the effort on the replacement, figure out docs (esp migration etc) and slowly wean the community off contour onto the new replacement which is Envoy Gateway.

thanks,
Dims

On Tue, Jun 14, 2022 at 1:15 AM Nick Young <inocuo@...> wrote:

Hi TOC, and everyone else,


With the successful launch of Envoy Gateway, there's a question about what to do about Contour's lifecycle as an incubated project in the CNCF.


I firmly believe that Envoy Gateway's launch is a huge win for the cloud native landscape, and the deduplication of effort possible by having us all working together will prove worthwhile once we get something built.


Normally, there's a (reasonable) requirement for incubating projects to show motion in the direction of becoming graduated, and a timeline to become graduated. But it seems to me that graduation implies a level of "this will be supported for the foreseeable future" that I don't know is viable for Contour.


VMware has committed to ensuring our maintenance of the project will be ongoing until users are ready to move away, see [our blog](https://blogs.vmware.com/opensource/2022/05/16/contour-and-community-build-new-envoy-gateway/) for more details. To summarize that blog, Contour's VMware maintainers will be helping to ensure that current users of Contour are looked after with features and support as long as possible, Wearing my Contour maintainer hat, we're an incubating project that will, within the next year or two, be mostly obsoleted by a functional, production-ready Envoy Gateway.  This implies that I need to figure out what we're going to do about Contour's incubation status.


The scenario I had assumed is that Contour may not graduate, and personally this makes me a little sad, but I recognize the larger opportunity that Envoy Gateway represents for the cloud native community. I know this situation hasn't come up before, but I suspect that this won't be the last time that CNCF projects interact in this fashion. I’d definitely like the TOC’s guidance here rather than making assumptions though.


Should Contour stay in incubating until Envoy Gateway is a viable alternative and users have moved away, then be archived? The timeline for this I would see as at least two years.


I personally see this as a successful exit for Contour, and have heard it referred to as "the open-source version of being acquired", but it would be nice for the TOC to give both us and other projects who may run into this situation in the future some guidance here.


Thanks for your time, everyone!


Nick Young

Wearing two of my maintainer hats for this email (Contour and Envoy Gateway).




--
Davanum Srinivas :: https://twitter.com/dims



Nick Young
 

Thanks everyone for the reminders about OpenTracing and OpenTelemetry, I had completely forgotten about that merge! I'd definitely be interested in hearing more from anyone involved in that either here, or via direct email or Slack DM if you want to talk about it.

Thanks for the response as well Dims, I'm glad we agree on the path forward.

Very interested to hear if anyone else has differing opinions still though!

Nick

On Wed, 15 Jun 2022 at 07:06, Evan Anderson <evana@...> wrote:
The OpenTracing and OpenCensus communities seem to have blazed the way here with OpenTelemetry -- I don't know if they have any learnings that they'd want to share about what worked well, or what could have been done better.

From: cncf-toc@... <cncf-toc@...> on behalf of Davanum Srinivas via lists.cncf.io <davanum=gmail.com@...>
Sent: Tuesday, June 14, 2022 3:27 AM
To: Nick Young <inocuo@...>
Cc: CNCF TOC <cncf-toc@...>
Subject: Re: [cncf-toc] Policy question: What happens when projects merge?
 

⚠ External Email

Nick,

Thanks for reaching out. One of the main objectives of the process is to give assurance to the end users of CNCF projects that we as a community stand behind the projects they are using (esp graduated ones). So we cannot do this in good faith for Contour. So my feeling is that we should leave contour where it is now, put the effort on the replacement, figure out docs (esp migration etc) and slowly wean the community off contour onto the new replacement which is Envoy Gateway.

thanks,
Dims

On Tue, Jun 14, 2022 at 1:15 AM Nick Young <inocuo@...> wrote:

Hi TOC, and everyone else,


With the successful launch of Envoy Gateway, there's a question about what to do about Contour's lifecycle as an incubated project in the CNCF.


I firmly believe that Envoy Gateway's launch is a huge win for the cloud native landscape, and the deduplication of effort possible by having us all working together will prove worthwhile once we get something built.


Normally, there's a (reasonable) requirement for incubating projects to show motion in the direction of becoming graduated, and a timeline to become graduated. But it seems to me that graduation implies a level of "this will be supported for the foreseeable future" that I don't know is viable for Contour.


VMware has committed to ensuring our maintenance of the project will be ongoing until users are ready to move away, see [our blog](https://blogs.vmware.com/opensource/2022/05/16/contour-and-community-build-new-envoy-gateway/) for more details. To summarize that blog, Contour's VMware maintainers will be helping to ensure that current users of Contour are looked after with features and support as long as possible, Wearing my Contour maintainer hat, we're an incubating project that will, within the next year or two, be mostly obsoleted by a functional, production-ready Envoy Gateway.  This implies that I need to figure out what we're going to do about Contour's incubation status.


The scenario I had assumed is that Contour may not graduate, and personally this makes me a little sad, but I recognize the larger opportunity that Envoy Gateway represents for the cloud native community. I know this situation hasn't come up before, but I suspect that this won't be the last time that CNCF projects interact in this fashion. I’d definitely like the TOC’s guidance here rather than making assumptions though.


Should Contour stay in incubating until Envoy Gateway is a viable alternative and users have moved away, then be archived? The timeline for this I would see as at least two years.


I personally see this as a successful exit for Contour, and have heard it referred to as "the open-source version of being acquired", but it would be nice for the TOC to give both us and other projects who may run into this situation in the future some guidance here.


Thanks for your time, everyone!


Nick Young

Wearing two of my maintainer hats for this email (Contour and Envoy Gateway).




--
Davanum Srinivas :: https://twitter.com/dims


⚠ External Email: This email originated from outside of the organization. Do not click links or open attachments unless you recognize the sender.


Richard Hartmann
 

Do you still need intros? If yes, I can introduce you to Alolita who was instrumental on OTel side.

Sent by mobile; please excuse my brevity.

On Tue, Jun 14, 2022, 20:02 Nick Young <inocuo@...> wrote:
Thanks everyone for the reminders about OpenTracing and OpenTelemetry, I had completely forgotten about that merge! I'd definitely be interested in hearing more from anyone involved in that either here, or via direct email or Slack DM if you want to talk about it.

Thanks for the response as well Dims, I'm glad we agree on the path forward.

Very interested to hear if anyone else has differing opinions still though!

Nick

On Wed, 15 Jun 2022 at 07:06, Evan Anderson <evana@...> wrote:
The OpenTracing and OpenCensus communities seem to have blazed the way here with OpenTelemetry -- I don't know if they have any learnings that they'd want to share about what worked well, or what could have been done better.

From: cncf-toc@... <cncf-toc@...> on behalf of Davanum Srinivas via lists.cncf.io <davanum=gmail.com@...>
Sent: Tuesday, June 14, 2022 3:27 AM
To: Nick Young <inocuo@...>
Cc: CNCF TOC <cncf-toc@...>
Subject: Re: [cncf-toc] Policy question: What happens when projects merge?
 

⚠ External Email

Nick,

Thanks for reaching out. One of the main objectives of the process is to give assurance to the end users of CNCF projects that we as a community stand behind the projects they are using (esp graduated ones). So we cannot do this in good faith for Contour. So my feeling is that we should leave contour where it is now, put the effort on the replacement, figure out docs (esp migration etc) and slowly wean the community off contour onto the new replacement which is Envoy Gateway.

thanks,
Dims

On Tue, Jun 14, 2022 at 1:15 AM Nick Young <inocuo@...> wrote:

Hi TOC, and everyone else,


With the successful launch of Envoy Gateway, there's a question about what to do about Contour's lifecycle as an incubated project in the CNCF.


I firmly believe that Envoy Gateway's launch is a huge win for the cloud native landscape, and the deduplication of effort possible by having us all working together will prove worthwhile once we get something built.


Normally, there's a (reasonable) requirement for incubating projects to show motion in the direction of becoming graduated, and a timeline to become graduated. But it seems to me that graduation implies a level of "this will be supported for the foreseeable future" that I don't know is viable for Contour.


VMware has committed to ensuring our maintenance of the project will be ongoing until users are ready to move away, see [our blog](https://blogs.vmware.com/opensource/2022/05/16/contour-and-community-build-new-envoy-gateway/) for more details. To summarize that blog, Contour's VMware maintainers will be helping to ensure that current users of Contour are looked after with features and support as long as possible, Wearing my Contour maintainer hat, we're an incubating project that will, within the next year or two, be mostly obsoleted by a functional, production-ready Envoy Gateway.  This implies that I need to figure out what we're going to do about Contour's incubation status.


The scenario I had assumed is that Contour may not graduate, and personally this makes me a little sad, but I recognize the larger opportunity that Envoy Gateway represents for the cloud native community. I know this situation hasn't come up before, but I suspect that this won't be the last time that CNCF projects interact in this fashion. I’d definitely like the TOC’s guidance here rather than making assumptions though.


Should Contour stay in incubating until Envoy Gateway is a viable alternative and users have moved away, then be archived? The timeline for this I would see as at least two years.


I personally see this as a successful exit for Contour, and have heard it referred to as "the open-source version of being acquired", but it would be nice for the TOC to give both us and other projects who may run into this situation in the future some guidance here.


Thanks for your time, everyone!


Nick Young

Wearing two of my maintainer hats for this email (Contour and Envoy Gateway).




--
Davanum Srinivas :: https://twitter.com/dims


⚠ External Email: This email originated from outside of the organization. Do not click links or open attachments unless you recognize the sender.


Nick Young
 

Thanks Richard, I'll catch up with Alolita and talk about OTel some more.

Thanks everyone for the help and advice, very much appreciated.

On Tue, 21 Jun 2022 at 22:26, Richard Hartmann <richih@...> wrote:
Do you still need intros? If yes, I can introduce you to Alolita who was instrumental on OTel side.

Sent by mobile; please excuse my brevity.

On Tue, Jun 14, 2022, 20:02 Nick Young <inocuo@...> wrote:
Thanks everyone for the reminders about OpenTracing and OpenTelemetry, I had completely forgotten about that merge! I'd definitely be interested in hearing more from anyone involved in that either here, or via direct email or Slack DM if you want to talk about it.

Thanks for the response as well Dims, I'm glad we agree on the path forward.

Very interested to hear if anyone else has differing opinions still though!

Nick

On Wed, 15 Jun 2022 at 07:06, Evan Anderson <evana@...> wrote:
The OpenTracing and OpenCensus communities seem to have blazed the way here with OpenTelemetry -- I don't know if they have any learnings that they'd want to share about what worked well, or what could have been done better.

From: cncf-toc@... <cncf-toc@...> on behalf of Davanum Srinivas via lists.cncf.io <davanum=gmail.com@...>
Sent: Tuesday, June 14, 2022 3:27 AM
To: Nick Young <inocuo@...>
Cc: CNCF TOC <cncf-toc@...>
Subject: Re: [cncf-toc] Policy question: What happens when projects merge?
 

⚠ External Email

Nick,

Thanks for reaching out. One of the main objectives of the process is to give assurance to the end users of CNCF projects that we as a community stand behind the projects they are using (esp graduated ones). So we cannot do this in good faith for Contour. So my feeling is that we should leave contour where it is now, put the effort on the replacement, figure out docs (esp migration etc) and slowly wean the community off contour onto the new replacement which is Envoy Gateway.

thanks,
Dims

On Tue, Jun 14, 2022 at 1:15 AM Nick Young <inocuo@...> wrote:

Hi TOC, and everyone else,


With the successful launch of Envoy Gateway, there's a question about what to do about Contour's lifecycle as an incubated project in the CNCF.


I firmly believe that Envoy Gateway's launch is a huge win for the cloud native landscape, and the deduplication of effort possible by having us all working together will prove worthwhile once we get something built.


Normally, there's a (reasonable) requirement for incubating projects to show motion in the direction of becoming graduated, and a timeline to become graduated. But it seems to me that graduation implies a level of "this will be supported for the foreseeable future" that I don't know is viable for Contour.


VMware has committed to ensuring our maintenance of the project will be ongoing until users are ready to move away, see [our blog](https://blogs.vmware.com/opensource/2022/05/16/contour-and-community-build-new-envoy-gateway/) for more details. To summarize that blog, Contour's VMware maintainers will be helping to ensure that current users of Contour are looked after with features and support as long as possible, Wearing my Contour maintainer hat, we're an incubating project that will, within the next year or two, be mostly obsoleted by a functional, production-ready Envoy Gateway.  This implies that I need to figure out what we're going to do about Contour's incubation status.


The scenario I had assumed is that Contour may not graduate, and personally this makes me a little sad, but I recognize the larger opportunity that Envoy Gateway represents for the cloud native community. I know this situation hasn't come up before, but I suspect that this won't be the last time that CNCF projects interact in this fashion. I’d definitely like the TOC’s guidance here rather than making assumptions though.


Should Contour stay in incubating until Envoy Gateway is a viable alternative and users have moved away, then be archived? The timeline for this I would see as at least two years.


I personally see this as a successful exit for Contour, and have heard it referred to as "the open-source version of being acquired", but it would be nice for the TOC to give both us and other projects who may run into this situation in the future some guidance here.


Thanks for your time, everyone!


Nick Young

Wearing two of my maintainer hats for this email (Contour and Envoy Gateway).




--
Davanum Srinivas :: https://twitter.com/dims


⚠ External Email: This email originated from outside of the organization. Do not click links or open attachments unless you recognize the sender.