Robie Basak: Platform for DMB Election (June 2024)

I've been on the DMB a while now, so you can judge me on my past performance. This text is largely the same as my platform from 2022.

While the DMB's powers might be to consider and act on individual applications, I think a wider goal for everyone on the DMB should be seek to get as many applicants qualified to our standards and then approved. The more contributors we can add, the better Ubuntu can be. If we're declining applicants regularly, that's a problem. We need to be asking ourselves what we can change to ensure that the majority of applicants are approved first time.

I noticed that there was an increasing number of failing applications from within Canonical. I've been working to address that internally. We don't yet have any general materials to share as a result of this effort that would be applicable outside Canonical, but I expect to publicly share anything that we do produce that isn't Canonical-specific. (I know that others are also working on improving the packaging guide which is where such materials should probably end up, but I cannot take credit for that.)

I am pro-active and do what I can to improve things beyond just deciding on applications. I usually have a promising contributor I'm focusing on to help get upload privileges next. This includes non-Canonicalers from the wider Ubuntu community, not just colleagues at Canonical. I've made various proposals for process improvements over the years. Some reached approval and were delivered; alas some were unable to reach consensus. I notice that I'm the most prolific contributor to our internal process documentation.

How I make application decisions

I think the key thing is that new uploaders do not upset existing uploaders were they to review their uploads. If I don't think this is likely to happen then I vote +1. If I'm concerned that it may happen for reasons that can be headed off by asking an applicant to continue getting sponsorship for now, then I vote -1.

My bar is highest for core dev applications because suboptimal uploads to core packages have the potential to cause the most widespread damage and challenges in reverting changes. My bar remains high for MOTU because I don't consider the universe component to have any lower quality requirement—especially from the perspective of our many MOTU volunteer contributors. My lowest bar is for "leaf" package PPU applications where the consequences of suboptimal uploads would be the lowest.

However, I don't decide purely on a social assessement. I look at technical understanding of Ubuntu processes too. This is because I don't believe it's possible for a prospective uploader to avoid upsetting other uploaders if they don't understand the technical basics of our processes. It doesn't need to be deeply technically detailed, but I think an understanding of what things are is essential to be able to know when to seek help and whom to ask.

I try to avoid asking a question unless the answer will make a difference to my vote.

Apart from basic technical questioning at an application meeting, I rely heavily on endorsements to judge the social and deeper technical aspects. Conversely, missing endorsements from sponsors who have sponsored many uploads for an applicant suggest to me that they don't think the applicant is ready.

I have my own documentation on what I personally expect from applicants. I'd like to see this expanded to be a general DMB stated opinion, but of course documenting this type of specifics is more difficult since this is only possible with consensus across the board on the details.

How I communicate application decisions

Unless abstaining due to a declared conflict of interest, I provide a definite +1 or -1 to every applicant I consider. I consider "+0" to be a cop-out when it's my job as a DMB member to form an opinion, because "+0" risks a hung vote when not all members are present (this is the norm) even though we are "quorate".

If I feel I have to decline an applicant, I will vote "-1" but make sure I provide a detailed, documented path forward and invite re-application when that path is concluded. If dealing with a re-application I will try not to add surprise new requirements. I should need only to check that the previous path I detailed is concluded to my satisfaction and my vote is automatically a "+1".