Is the TSC the right size? Should it be expanded?
Proposal: Extend the size of the TSC to 15 (pending Board approval). To get started, add the next 4 candidates from the last election results to the current term.
I believe I read the discussion(s) from the previous TSC meetings on this topic, but I'm not really sure what we are trying to accomplish with expanding the size of the TSC.
I think the basic arguments are the following:
I think Hart got it right. There are many ways and certainly more sophisticated ones to go at this that have been mentioned and that could maybe lead to a more diverse representation but I understand from the staff that if we simply took the next 4 people from the election results this would already improve diversity so I'm happy to go with this as a first step. It has the advantage of being simple and implementable very quickly.
I'm a strong believer in iterative processes and open to refining this further down the line but I don't want investigating other possible ways to come at the cost of doing nothing for months.
OK ... got it. I just worry that people will interpret "diversity" in different ways .... if the real outcome here is just that we think that we need a larger number of members given the current size of the community, then simply adding 4 makes sense.
I would hate to make decisions about the size of the TSC in combination with understanding the increase in diversity. If we did not know this information would we make the same decision? Will this hold true next year? Is the diversity that we are adding just project diversity? Does the additional members hurt our increased gender diversity (2/11 vs. 2/15)?
Thanks for chiming in Tracy. I have to admit my goal was indeed to increase project diversity because that's what had primarily been brought up by Chris when he first talked about this before the election and then after the election by the critics of the results. I had not thought that this might result in a decrease in gender diversity which sadly is actually very likely given the lack of diversity among the candidates in the first place.
So, what should we do about this? Any suggestion? It seems like we can't win on all fronts.
No good suggestions. Based on what we are trying to accomplish (increase in project diversity), adding the next four people may help this year, but is this really a long-term solution. What’s to say that we don’t end up next year with a TSC that is still not representative of the broader community? Do we again increase by 4 to get that? Or do we need to increase by 8 next year? We have a systemic problem with diversity of all sorts and I don’t know that the proposals that are on the table solve this long term. The DCI WG was created to help provide suggestions surrounding diversity, but it will most likely take a while before we see results in these efforts.
I totally second Arnaud's and Hart's inputs. The ecosystem has evolved and grown quite a lot last 2 years. I think there should also be a right spread of the TSC members across geographies. Would also be good to think of how can someone still contribute to the TSC without going thru the election process.
The TSC calls and mailing list are open to anyone and several people have been contributing for years without being official members.
One idea that came out of some discussions about increasing diversity on the TSC - geographic, project representation, gender or other demographic variables - is for a newly-elected TSC to appoint a set of additional TSC seats to increase its own diversity intentionally. This is different from simply changing "top 11" to "top 15 or 17". This would be a delicate process, of course, but could be a powerful way to get additional diversity quickly. I believe this was done in the CNCF community.
This is exactly the kind of other ways I was referring to as "more sophisticated ones [...] that have been mentioned and that could maybe lead to a more diverse representation". There are others. For instance, we could also allocate one seat per project/group we want represented and leave it to each group to figure out who will represent them.
I just fear that figuring out which method to use and all of its details can take quite a while to sort out.
am glad you said that Brian. I was coming from the same school fo thought. I see below that Arnaud also concurs to this idea. Would be glad to contribute further on this.
A small set of the community (vs. the entire community) choosing additional folks seems like the wrong approach.
Having given this quite some thoughts I'm back to my original proposal because I think every way has its own downside.
I thought adding a seat per project might make sense but that actually makes having a project or not much more sensitive. One might want to have a project because it's a way to get a seat, and one might not want to be a subproject or get archived not to lose their seat.
Having 11 people handpicking four more seems equally doomed. I think we'd end up having to hold some internal election again. But then, why not use the result of the first one?
So, I'm back to my proposal: Extend the size of the TSC to 15 (pending Board approval). To get started, add the next 4 candidates from the last election results to the current term.
"I thought adding a seat per project might make sense but that actually makes having a project or not much more sensitive. One might want to have a project because it's a way to get a seat, and one might not want to be a subproject or get archived not to lose their seat."
I've thought about this kind of federated approach a little bit. I think if you wanted to do it, you would need to allocate seats to active projects based on the number of active contributors (with zero seats being a possibility for small projects) according to some (probably sublinear) formula.
This would partially solve the subproject issue–presumably if a group was large enough to get a seat, they could increase an existing project's size by joining as a subproject and get a seat that way as well. If the formula was sublinear, this might not exactly be the case (and the group might not get the number of seats they would otherwise) but it would partially alleviate the problem.
In considering drawing the next 4 names...Let's say the next 6 names are as follows:
In considering expanding the TSC at all...Let's say it took 11 people 3 months to resolve lifecycle questions. How long would it take with 15? (hint: ~4/3 * 3 months =? )
Indeed, with more people decisions don't necessarily get easier, especially with strongly opinionated people like we typically have on the TSC.
I don't think there is a perfect solution. I think increasing the number of seats increases the chance of diversity, even if it doesn't guarantee it, and is trivial to implement. For the reason aforementioned I don't think we should increase too much but 15 seems like a middle ground to me.
If someone is no longer interested we can keep going down the list or leave the seat empty until the next election which would be no different than what happens today if one of the elected members decided to withdraw.
WRT remarks by Tracy Kuhrt elsewhere in this forum: we would need to go to 20th place to reach the next woman in voting; she is tied with a man.
I understand why you brought this up, but I don't think it's a good practice to have the TSC voting on expanding the TSC with the knowledge of who will be taking the seat. Otherwise we may as well just let the TSC pick directly.
Somebody (I think it was Tracy?) made the good argument today that adding additional TSC members now might create "unequal" status for TSC members since everyone would know who the last four people (out of the hypothetical 15 total) would be. This is something we should definitely consider, and not just in the case of adding new TSC members after the election. While I am in favor of expanding the TSC, one of the things I worry about is the TSC getting big enough that it is essentially dominated by a small clique with the result that most of the people on the TSC don't really have a voice.
As an example, if we look at who dominated the conversation in the TSC meeting today (both TSC members and non-TSC members), most of the people who talked the most were what I call "Hyperledger dinosaurs." In other words, these are folks who have been around Hyperledger, generally in leadership positions in some form, for most of the existence of Hyperledger. I think this set of "dinosaurs" plus LF staff (Ry and Silona) probably were responsible for 80-90% of the conversation in the meeting–thanks to Angelo for being one of the newer folks to speak up! While this may have been an atypical meeting–we discussed elections for most of the meeting, and having knowledge of past elections was a huge advantage in terms of knowledge of the discussion–it's something we probably want to avoid for more general discussions.
So, if we do expand the TSC, what can we do to make sure that everyone is heard and considered? A diverse (in terms of people and projects) TSC helps us the most if everyone contributes to the discussion. If we can concretely answer this question, then I think expanding the TSC becomes a no-brainer.
Thanks Hart, I think you're bringing up a very interesting point. As a matter of fact I was thinking after the call that we had not heard at all from some of the members.
As chair I will try to make an effort to reach out to the less vocal ones, prompting them to get their voice heard. This is actually a technique I'm used to use when chairing over difficult issues for which no consensus seems to emerge. Typically it's the most vocal ones who dominate the discussion and pull the group in opposite directions. Having the less vocal members speak up can both be refreshing and enlightening. They may end up showing the way forward.
But please don't wait for me to do it, go ahead and ask for their opinion if I don't.
By the way, on the question as to whether people would mind being added to the TSC now with the consequence of having the world find out that they were ranked lower than everyone else who's already on the committee I think rather than speculate about it we should simply have the staff confidentially reach out to them and ask how they feel about it. They may value being on the committee more than the fact that people will know they were part of the second batch.
I am totally against adding the next n winners until next election.
It does not matter whether we pick the next four, hold another election made up of those who did not make it, hold another election, appoint the next four .... in each of those cases the four additional members were not elected the first time around.
To Hart's last paragraph, maybe we need to decide this in conjunction with the Membership diversity issue. TSC Membership Diversity
Perhaps we are going about this the wrong way ... people continue to use the term diversity ... but in multiple ways: project, gender, geography, etc. Are we really trying to force some type of mix here? For example, taking project, gender and geography, are we really going to say one male, one female, one non-binary per project per geography? Because that's where this type of thinking leads. I'm not saying that's wrong, I'm just saying that's where it ultimately leads.
Maybe we need to step back and look at how/why the current and past board members were elected and why others were not. Of course I don't think we really want to get into every election theory ever created, but it's worth looking at. People tend to vote for those they know or those that appeal to their beliefs. Also, not everyone will vote ... only those that actually feel they have something at stake tend to vote. And of course people have to know that there is an election to vote.
Let's look at past elections and ask two questions:
In my opinion, until you at least answer these questions, I'm not sure how or why we would actually make any immediate changes to the current board.
I will also state that I do not believe it should be the responsibility of the current TSC to come up with the "solution" to whatever the underlying issues are (yes I understand it will be a proposal which goes to the Governing Board).
Powered by a free Atlassian Confluence Community License granted to The Linux Foundation. Evaluate Confluence today.