https://doodle.com/poll/5bx4c7krsb3xctxu Pick your times if you want to be part of the discussion… - Martin
Ok, there seems to be some serious lack of interest. Only got response from David and “Robot” (supposedly Donald?). (and Jafar saying that he doesn’t have time) Not sure how to proceed, so here is my rant: I’m thinking about just closing the Redhat Package PR (not merging, just close) and someone else can address it again in maybe 6 months or whenever there is any interest at all. I was really hoping to get something in - maybe not perfect, but at least a start and everyone would be welcome to improve it afterwards with PRs on top of the base. The PR doesn’t break anything existing (current packages fail on all OS for me, so I’m not building any RH packages at this time). So I think it’s better than what we have now and it doesn’t break anything existing. Based on this, my view would be to get it in. But it seems the discussion goes down the rathole on how to have all perfect back to 2.0 release and rather have nothing instead of this. There is very little technical feedback (Donald provided some and I thank him for this - even if he has different views, I highly appreciate them. The issue here is NOT the fault of Donald, but the lack of feedback from just about everyone else. It just caused a “standstill” as Donald and I have different personal views on some of the technical points. But I think a 3rd/4th opinion could have solved this very easily) Now from me pushing the long overdue decision, the whole things seems to derail on how to get packages into distro’s. This is secondary to me as they always lag far behind and I think there is no way around providing up-to-date packages directly like most other projects do. </rant> Steps forward: I formally ask this now to be moved to the TSC for a vote on how to move forward (as our charters outline). I want an agreement to either - Close (abandon) the PR without merging - Accept the PR as it is - Clear indication of what is missing/broken and required to be fixed for it to be accepted. Let’s see how our TSC process works… - Martin On 23 May 2017, at 8:56, Martin Winter wrote:
https://doodle.com/poll/5bx4c7krsb3xctxu
Pick your times if you want to be part of the discussion…
- Martin
So do you want to have the meeting to discuss the issue or not? Robot On Wed, May 24, 2017 at 8:21 PM, Martin Winter <mwinter@opensourcerouting.org> wrote:
Ok, there seems to be some serious lack of interest.
Only got response from David and “Robot” (supposedly Donald?). (and Jafar saying that he doesn’t have time)
Not sure how to proceed, so here is my rant:
I’m thinking about just closing the Redhat Package PR (not merging, just close) and someone else can address it again in maybe 6 months or whenever there is any interest at all.
I was really hoping to get something in - maybe not perfect, but at least a start and everyone would be welcome to improve it afterwards with PRs on top of the base. The PR doesn’t break anything existing (current packages fail on all OS for me, so I’m not building any RH packages at this time). So I think it’s better than what we have now and it doesn’t break anything existing. Based on this, my view would be to get it in.
But it seems the discussion goes down the rathole on how to have all perfect back to 2.0 release and rather have nothing instead of this. There is very little technical feedback (Donald provided some and I thank him for this - even if he has different views, I highly appreciate them. The issue here is NOT the fault of Donald, but the lack of feedback from just about everyone else. It just caused a “standstill” as Donald and I have different personal views on some of the technical points. But I think a 3rd/4th opinion could have solved this very easily)
Now from me pushing the long overdue decision, the whole things seems to derail on how to get packages into distro’s. This is secondary to me as they always lag far behind and I think there is no way around providing up-to-date packages directly like most other projects do.
</rant>
Steps forward:
I formally ask this now to be moved to the TSC for a vote on how to move forward (as our charters outline). I want an agreement to either - Close (abandon) the PR without merging - Accept the PR as it is - Clear indication of what is missing/broken and required to be fixed for it to be accepted.
Let’s see how our TSC process works…
- Martin
On 23 May 2017, at 8:56, Martin Winter wrote:
https://doodle.com/poll/5bx4c7krsb3xctxu
Pick your times if you want to be part of the discussion…
- Martin
_______________________________________________ TSC mailing list TSC@lists.frrouting.org https://lists.frrouting.org/listinfo/tsc
On 25 May 2017, at 4:25, Donald Sharp wrote:
So do you want to have the meeting to discuss the issue or not?
Personally, I don’t see this to get us anywhere moving forward (Moving forward, as defined to come to an agreement of one of the 3 choices listed below). I do believe that you raised some technical concerns and you are pretty solid in your believe of them and they are different to my views (and I’m pretty solid in believing in them). Just to be clear: I don’t blame the holdout on you - I highly appreciate your views, but I believe the only way forward is by getting more people to chime in (preferably after they spent time investigating it) So unless we either get either more people involved, I do believe forcing this to the TSC will get us a decision and I would love to see if the TSC process actuality holds up in practice as defined. If you see any better way to actually get this resolved, then do please explain. My preference is to get this resolved one way or another ASAP. Until this is resolved I see little chance to get any other packages built for FRR (I agree that they should have at least the same look and feel) and I would prefer to get packages for 3.0 and Debian /FreeBSD in there VERY soon. Or did I completely miss the reason why we would have a meeting? - Martin
On Wed, May 24, 2017 at 8:21 PM, Martin Winter <mwinter@opensourcerouting.org> wrote:
Ok, there seems to be some serious lack of interest.
Only got response from David and “Robot” (supposedly Donald?). (and Jafar saying that he doesn’t have time)
Not sure how to proceed, so here is my rant:
I’m thinking about just closing the Redhat Package PR (not merging, just close) and someone else can address it again in maybe 6 months or whenever there is any interest at all.
I was really hoping to get something in - maybe not perfect, but at least a start and everyone would be welcome to improve it afterwards with PRs on top of the base. The PR doesn’t break anything existing (current packages fail on all OS for me, so I’m not building any RH packages at this time). So I think it’s better than what we have now and it doesn’t break anything existing. Based on this, my view would be to get it in.
But it seems the discussion goes down the rathole on how to have all perfect back to 2.0 release and rather have nothing instead of this. There is very little technical feedback (Donald provided some and I thank him for this - even if he has different views, I highly appreciate them. The issue here is NOT the fault of Donald, but the lack of feedback from just about everyone else. It just caused a “standstill” as Donald and I have different personal views on some of the technical points. But I think a 3rd/4th opinion could have solved this very easily)
Now from me pushing the long overdue decision, the whole things seems to derail on how to get packages into distro’s. This is secondary to me as they always lag far behind and I think there is no way around providing up-to-date packages directly like most other projects do.
</rant>
Steps forward:
I formally ask this now to be moved to the TSC for a vote on how to move forward (as our charters outline). I want an agreement to either - Close (abandon) the PR without merging - Accept the PR as it is - Clear indication of what is missing/broken and required to be fixed for it to be accepted.
Let’s see how our TSC process works…
- Martin
On 23 May 2017, at 8:56, Martin Winter wrote:
https://doodle.com/poll/5bx4c7krsb3xctxu
Pick your times if you want to be part of the discussion…
- Martin
_______________________________________________ TSC mailing list TSC@lists.frrouting.org https://lists.frrouting.org/listinfo/tsc
So I specifically asked for the meeting last week so that the interested parties could hammer out an initial agreement( or to refine their arguments ) for the TSC to help minimize the time in the meeting.. Mainly because I wanted to avoid getting into the meeting and spending most of it on this issue. donald On Fri, May 26, 2017 at 5:40 PM, Martin Winter <mwinter@opensourcerouting.org> wrote:
On 25 May 2017, at 4:25, Donald Sharp wrote:
So do you want to have the meeting to discuss the issue or not?
Personally, I don’t see this to get us anywhere moving forward (Moving forward, as defined to come to an agreement of one of the 3 choices listed below).
I do believe that you raised some technical concerns and you are pretty solid in your believe of them and they are different to my views (and I’m pretty solid in believing in them). Just to be clear: I don’t blame the holdout on you - I highly appreciate your views, but I believe the only way forward is by getting more people to chime in (preferably after they spent time investigating it)
So unless we either get either more people involved, I do believe forcing this to the TSC will get us a decision and I would love to see if the TSC process actuality holds up in practice as defined.
If you see any better way to actually get this resolved, then do please explain. My preference is to get this resolved one way or another ASAP. Until this is resolved I see little chance to get any other packages built for FRR (I agree that they should have at least the same look and feel) and I would prefer to get packages for 3.0 and Debian /FreeBSD in there VERY soon.
Or did I completely miss the reason why we would have a meeting?
- Martin
On Wed, May 24, 2017 at 8:21 PM, Martin Winter <mwinter@opensourcerouting.org> wrote:
Ok, there seems to be some serious lack of interest.
Only got response from David and “Robot” (supposedly Donald?). (and Jafar saying that he doesn’t have time)
Not sure how to proceed, so here is my rant:
I’m thinking about just closing the Redhat Package PR (not merging, just close) and someone else can address it again in maybe 6 months or whenever there is any interest at all.
I was really hoping to get something in - maybe not perfect, but at least a start and everyone would be welcome to improve it afterwards with PRs on top of the base. The PR doesn’t break anything existing (current packages fail on all OS for me, so I’m not building any RH packages at this time). So I think it’s better than what we have now and it doesn’t break anything existing. Based on this, my view would be to get it in.
But it seems the discussion goes down the rathole on how to have all perfect back to 2.0 release and rather have nothing instead of this. There is very little technical feedback (Donald provided some and I thank him for this - even if he has different views, I highly appreciate them. The issue here is NOT the fault of Donald, but the lack of feedback from just about everyone else. It just caused a “standstill” as Donald and I have different personal views on some of the technical points. But I think a 3rd/4th opinion could have solved this very easily)
Now from me pushing the long overdue decision, the whole things seems to derail on how to get packages into distro’s. This is secondary to me as they always lag far behind and I think there is no way around providing up-to-date packages directly like most other projects do.
</rant>
Steps forward:
I formally ask this now to be moved to the TSC for a vote on how to move forward (as our charters outline). I want an agreement to either - Close (abandon) the PR without merging - Accept the PR as it is - Clear indication of what is missing/broken and required to be fixed for it to be accepted.
Let’s see how our TSC process works…
- Martin
On 23 May 2017, at 8:56, Martin Winter wrote:
https://doodle.com/poll/5bx4c7krsb3xctxu
Pick your times if you want to be part of the discussion…
- Martin
_______________________________________________ TSC mailing list TSC@lists.frrouting.org https://lists.frrouting.org/listinfo/tsc
Hi Donald, Martin, The discussion is about https://github.com/FRRouting/frr/pull/378 ? The packaging process may differ for each distribution, and I think contrary to the core application of frr, the packaging should differ between distributions. I would tend to accept the PR as is. Then, if any issues on specific Linux distributions, the rpm folks will have to show up. Martin did a nice job to bootstrap and to try to aggregate packaging, but I would tend to focus on the Debian/Ubuntu ones (startup and .deb). Then, I would ask the Linux vendors to manage the over-mess of their startup logics+packaging logics. Cheers, Philippe On Mon, May 29, 2017 at 8:29 PM, Donald Sharp <sharpd@cumulusnetworks.com> wrote:
So I specifically asked for the meeting last week so that the interested parties could hammer out an initial agreement( or to refine their arguments ) for the TSC to help minimize the time in the meeting.. Mainly because I wanted to avoid getting into the meeting and spending most of it on this issue.
donald
On Fri, May 26, 2017 at 5:40 PM, Martin Winter <mwinter@opensourcerouting.org> wrote:
On 25 May 2017, at 4:25, Donald Sharp wrote:
So do you want to have the meeting to discuss the issue or not?
Personally, I don’t see this to get us anywhere moving forward (Moving forward, as defined to come to an agreement of one of the 3 choices listed below).
I do believe that you raised some technical concerns and you are pretty solid in your believe of them and they are different to my views (and I’m pretty solid in believing in them). Just to be clear: I don’t blame the holdout on you - I highly appreciate your views, but I believe the only way forward is by getting more people to chime in (preferably after they spent time investigating it)
So unless we either get either more people involved, I do believe forcing this to the TSC will get us a decision and I would love to see if the TSC process actuality holds up in practice as defined.
If you see any better way to actually get this resolved, then do please explain. My preference is to get this resolved one way or another ASAP. Until this is resolved I see little chance to get any other packages built for FRR (I agree that they should have at least the same look and feel) and I would prefer to get packages for 3.0 and Debian /FreeBSD in there VERY soon.
Or did I completely miss the reason why we would have a meeting?
- Martin
On Wed, May 24, 2017 at 8:21 PM, Martin Winter <mwinter@opensourcerouting.org> wrote:
Ok, there seems to be some serious lack of interest.
Only got response from David and “Robot” (supposedly Donald?). (and Jafar saying that he doesn’t have time)
Not sure how to proceed, so here is my rant:
I’m thinking about just closing the Redhat Package PR (not merging, just close) and someone else can address it again in maybe 6 months or whenever there is any interest at all.
I was really hoping to get something in - maybe not perfect, but at least a start and everyone would be welcome to improve it afterwards with PRs on top of the base. The PR doesn’t break anything existing (current packages fail on all OS for me, so I’m not building any RH packages at this time). So I
think
it’s better than what we have now and it doesn’t break anything existing. Based on this, my view would be to get it in.
But it seems the discussion goes down the rathole on how to have all perfect back to 2.0 release and rather have nothing instead of this. There is very little technical feedback (Donald provided some and I thank him for this - even if he has different views, I highly appreciate them. The issue here is NOT the fault of Donald, but the lack of feedback from just about everyone else. It just caused a “standstill” as Donald and I have different personal views on some of the technical points. But I think a 3rd/4th opinion could have solved this very easily)
Now from me pushing the long overdue decision, the whole things seems to derail on how to get packages into distro’s. This is secondary to me as they always lag far behind and I think there is no way around providing up-to-date packages directly like most other projects do.
</rant>
Steps forward:
I formally ask this now to be moved to the TSC for a vote on how to move forward (as our charters outline). I want an agreement to either - Close (abandon) the PR without merging - Accept the PR as it is - Clear indication of what is missing/broken and required to be fixed for it to be accepted.
Let’s see how our TSC process works…
- Martin
On 23 May 2017, at 8:56, Martin Winter wrote:
https://doodle.com/poll/5bx4c7krsb3xctxu
Pick your times if you want to be part of the discussion…
- Martin
_______________________________________________ TSC mailing list TSC@lists.frrouting.org https://lists.frrouting.org/listinfo/tsc
_______________________________________________ TSC mailing list TSC@lists.frrouting.org https://lists.frrouting.org/listinfo/tsc
The discussion is about https://github.com/FRRouting/frr/pull/378 ?
Yes.
The packaging process may differ for each distribution, and I think contrary to the core application of frr, the packaging should differ between distributions.
I think what Donald is after is to have the look and feel of the packaging process the same across all distributions -- to not confuse users. Is this a reasonable goal? If so, then we can work out _how_ to do this (one file or many)...
Martin did a nice job to bootstrap and to try to aggregate packaging, but I would tend to focus on the Debian/Ubuntu ones (startup and .deb).
Maybe we should focus only on Debian & Ubuntu, keeping these two a single look & feel, while leaving the rest to do what they want to do? 😊 /r
I think both goals are reasonable. I'm inclined to take the "contentious" patch and then discuss alternative / consolidate approaches separately. Lou On May 30, 2017 7:10:03 AM "Russ White" <russ@riw.us> wrote:
The discussion is about https://github.com/FRRouting/frr/pull/378 ?
Yes.
The packaging process may differ for each distribution, and I think contrary to the core application of frr, the packaging should differ between distributions.
I think what Donald is after is to have the look and feel of the packaging process the same across all distributions -- to not confuse users. Is this a reasonable goal? If so, then we can work out _how_ to do this (one file or many)...
Martin did a nice job to bootstrap and to try to aggregate packaging, but I would tend to focus on the Debian/Ubuntu ones (startup and .deb).
Maybe we should focus only on Debian & Ubuntu, keeping these two a single look & feel, while leaving the rest to do what they want to do?
😊 /r
_______________________________________________ dev mailing list dev@lists.frrouting.org https://lists.frrouting.org/listinfo/dev
participants (5)
-
Donald Sharp -
Lou Berger -
Martin Winter -
Philippe Guibert -
Russ White