Wednesday, April 18, 2012

Proposal: Network Latency

Timed out and passed, 4-3. Josh

Adminned at 20 Apr 2012 03:02:23 UTC

To the rule “Cycle Resolution”, add:-

For each Cycle’s resolution, the Net should make all required announcements, ruleset changes and GNDT updates, including those effects triggered by the start of a new Cycle, within a single one-hour window.

It’s been a little confusing that the Net has sometimes waited a couple of days before updating the GNDT, but this Cycle he’s also skipped over updating the Powered/Unpowered status of Institutions, updating the Bonuses and (ten hours into a new 72-hour cycle) is currently holding back on sharing last cycle’s bidding details until “things look solid”.

I appreciate the ruleset doesn’t require him to do any of this at a particular time, and that the GNDT, Powering and Bonuses can be updated by any player, but do a quorum of players agree that this would be clearer if it was all done at the same time, and by the same person?


Kevan: HE/HIM

04-18-2012 08:57:51 UTC

I’ve gone ahead and Powered the relevant buildings. (The Reactor and Public are Unpowered because they were amended by enactments that occurred during the current cycle.)


04-18-2012 12:23:58 UTC

against God knows I make mistakes - lets at least give me some time to correct them before I publish something I’m not supposed to.


04-18-2012 12:29:25 UTC

for We can check for mistakes once you have published them. The Watchtower’s report will cover every issue when no one influenced it, and almost every issue when someone influenced it.
Besides, this does let you some time, as you are only required to do everything you “within a singl one-hour window”, but not necessarily between 0001 and 0101 UTC.


04-18-2012 12:31:59 UTC

What if I mistakenly believe that no one has influenced the Watchtower, but someone has?


04-18-2012 12:45:29 UTC

imperial Then that is problematic. It doesn’t really matter to me.

Josh: HE/HIM

04-18-2012 12:48:11 UTC

Then we will all be forgiving, but if that’s the only thing that you have to be certain of before publishing then it’s still a quicker process that the status quo.

The task is much easier when you’re spreading the load, rather than putting pressure on yourself to check that everything is correct. The players have a vested interest in ensuring that the gamestate is correct; letting them help you should make the process much easier.



04-18-2012 12:54:02 UTC


Kevan: HE/HIM

04-18-2012 12:59:50 UTC

[south] But how would we the players know that someone had influenced the Watchtower? If you post a Cycle Resolution that says it wasn’t influenced, we can’t confirm or deny that until we see the bidding results. (And if there was a mistake, it’s fine, we’d talk about it and vote on a solution. We’ve had worse.)

I can see that a gradual reveal of gamestate makes it easier to catch some errors at each step before publishing the next one, and if you wanted to run the dynasty that way, we could do that. But the ruleset should ideally be explicit about it, so that the players know what to expect - it’s currently quite confusing to look at the gamestate and not know which parts are accurate, and which parts you’re intending to update later. (“Is the Ansible Unpowered because the Net hasn’t processed this Cycle’s repowering yet, or is it Unpowered because the Reactor shut it down and the Net didn’t need to update the wiki, or is it actually Powered?”)


04-18-2012 13:06:46 UTC

@Kevan, your point is fair - I am cooperative, I’m just trying to avoid damage of my own stupid mistakes as best as possible.

Kevan: HE/HIM

04-18-2012 13:18:15 UTC

[south] Oh, I know the feeling. I’d entirely support a proposal of “Cycle Results should be published within 24 hours, Watchtower output and GNDT/wiki updates made within 36 hours” if you were happier with that. The important thing is really just knowing how far the Net has gotten through processing the last Cycle, and whether we’re looking at an accurate or lagging gamestate.


04-18-2012 14:17:08 UTC

@Kevan, that seems like a reasonable option. The hour thing is doable - I just am trying to keep mess to a minimum.


04-18-2012 20:00:55 UTC