My understanding is that the Recent Outage wasn't because of bandwidth at all, but I do believe bandwidth could become a problem at some future point, and I also believe that having more threads means we use more bandwidth -- see that Natter hasn't slowed down, even with new threads around.

None of which has anything to do with the matter at hand.

Go 6!

I must have said that wrong. If something wins, no one agitates against it for X months. If something loses, no one agitates for it for X months. That's what I recall from bureacracy. Any decision can revisited after the moratorium.

clarifying because this is part of what made me lean toward 6.

So perhaps (it seems, to my mind) the ballot would most fairly look like this:

I actually think that would unfairly penalise 6 months. See, if we go with that option, 6 would need to get over half the primary votes to win. The other two options only need to win a preferential ballot.

It's not, correct me if I'm wrong, ita, that creating more threads causes database problems, but simply that experience shows that having more threads means there's more traffic.

Yes. The table of posts is what Hostrocket says is too large.

I can't figure if you're correcting me because I'm wrong or saying I'm right.

There's one central table for all posts, which are just marked as belonging to a thread, rather than a table for each thread?

But didn't somebody run numbers that suggested that there was a strong causal correlation between numbers of threads and numbers of posts?

That is, you have 3 threads, you have 3000 posts; you have 4 threads, you don't have 3200 posts, you have 4000. (Numbers pulled out of Stephanie's frog-whistle.)

didn't somebody run numbers that suggested that there was a strong causal correlation between numbers of threads and numbers of posts?

I kind of said that, though not in so many words. More threads does not dissipate the same number of posts over a wider area, it increases the overall number, yes.

But if there's a fundamental issue where more posts means more instability, then we have to deal with it sooner or later anyway because the number of posts in that table will always get bigger unless they're actively removed in some way.

They're being actively removed.

Our database is large by HR's standard, and our posts database is most of that.

It was decided to archive many of the Natters, and I'm halfway through the accompanying coding effort to reroute links to those threads.

How will 6 certainly win?

Because, I mean, I've gotten a very strong vibe from this thread that the 6 people comprise a larger group than either the 3 people or the 4 people, if taken separately.

edit: and your ballot totally cuts out people who think 4 is fine, but 3 is too short, and would therefore vote for 4 first, but for 6 in a 3 vs 6 runoff.

... yes. That's a weakness.

I actually think that would unfairly penalise 6 months. See, if we go with that option, 6 would need to get over half the primary votes to win. The other two options only need to win a preferential ballot.

And I was saying that that kind of was fair, because, as I was seeing it, 3 and 4 are much more similar to one another than either of them are to 6.

But, you know, OK, so that quasi-preferential ballot of mine has those weaknesses. I really think we ought to go with an actual preferential thing. I'm with Jon.

I'm with Jon.

Wow! Thanks, Rebecca.