This all sounds reasonable to me. The Change owners would be informed that they can ask us for support, if they wish, and if we had a volunteer with free time, he/she would get involved with the Change. Not just help prepare test days, but also subscribe to the tracker bug, look for user feedback in different places and steer them into the right channels (e.g. make use of the tracker bug), ideally also test it a bit themselves.

Of course this all assumes we have people-time resources to give, which we're always low on, but perhaps some community members would be interested as well, and this is a nice way to get more involved. And for really heavy changes, e.g. like DNF5, we would need to get involved anyway, so it makes sense to be part of the process early.

Overall I think this sounds good. Any other thoughts?

I agree with Kamil.

Just to let you know, @amoloney, whenever we could, we have always offered people that they come to us and discuss the possibilities of testing their products on Fedora as early as possible, but a few of them actually have come into touch with us.

So, the question is, how powerful will be the QA position in the process? For instance:

• Will the QA approval be needed for the change to happen?
• Will the QA be able to stop and revert the change if we think it is not ready?
• WIll those who propose need to "get" a QA person on their project and will they be required to get one? Or is putting someone's name on the proposal just a matter of their decision?
• If community would be able to join as QA (as Kamil suggests), how do we know that this a real, reliable and responsible person?
• Will the QA person be responsible for the quality of the change fully, partly, or not at all?

Generally, I am supportive of your proposal, but I think that without certain rights and responsibilities, the thing will become cosmetics (=looks good but does nothing). Yet, on the other hand, if the there is too much to control from our side, we might not have the time slots to do it, as Kamil mentions, too.

So, the question is, how powerful will be the QA position in the process? For instance:

• Will the QA approval be needed for the change to happen?
• Will the QA be able to stop and revert the change if we think it is not ready?

This is not how I read Aoife's proposal and I personally wouldn't even want to go that far. I see ourselves as having a strictly supportive role in this. We have enough responsibilities already. Also, this is FESCo territory, and if we think something's not a good idea or not ready yet, we can always file a FESCo ticket with our concerns.

• WIll those who propose need to "get" a QA person on their project and will they be required to get one? Or is putting someone's name on the proposal just a matter of their decision?

The original description has "optionally" in italics. I believe that's the right approach. Also, provided that we actually have a volunteer and enough free time.

• If community would be able to join as QA (as Kamil suggests), how do we know that this a real, reliable and responsible person?

We already have community members involved in many areas of our work, it's nothing new. Because this would be a project possibly spanning several months, perhaps a good idea is to check the status (whether something is happening) during our QA meetings from time to time?

• Will the QA person be responsible for the quality of the change fully, partly, or not at all?

The original description says "will not own the change". Which is the correct approach, in my eyes.

without certain rights and responsibilities, the thing will become cosmetics

I guess it will depend on how much time the QA person has, how much he/she wants to get involved, how much the Change owners need QA. I see it as a nice improvement, which is optional for interested parties. Basically in the same spirit as Test Days are.

@kparal summed up my replies nicely :)

The only addition I'm going to chime in here with is around the question:

"WIll those who propose need to "get" a QA person on their project and will they be required to get one? Or is putting someone's name on the proposal just a matter of their decision?"

For this, I am going to offer @sumantrom as sacrifice the main driver of coordinating this additional support between QA and the change proposal owners. He and I have been discussing this for the last few weeks and he has expressed an interest in leading the coordination effort, and I would be happy to work with him and support when needed.

To be clear though, overall, this is an optional ... option... for change proposal owners to avail of if there is a volunteer/interested person in QA that has the capacity to help with initial testing of a change before it gets announced.

I want to bring this to the lists & discourse, but I wanted to get QA's buy-in first before we start talking about how best to implement this option (Sumantro and I have ideas) :)

Thanks btw for the positive reception and thoughtful questions so far too!

@adamwill , your thoughts and blessings on this topic will help us (ie Aoife and I) to start this discussion with a larger group of audience.

PS : I am prepared to be the proverbial sacrifice.

I mean, I'm not the godfather, I don't have to sign off on everything :P Kamil's and Lukas' replies seemed to cover everything, so I didn't think it was necessary to add anything.

Here is what @sumantrom and I are planning to send this week. If you dont have access, let me know and I can add you directly to the doc with your email https://docs.google.com/document/d/1twnUfvMj9bgSBw5Vegg6bEsXv4r9wZYt0w2fptLNt6w/edit?usp=sharing