Sifting through issues

Week 31 is over. This week was, as announced, dedicated to some “minor” issues, which actually comprise the rest of work remaining before the new round. Let’s start from beginning.

Follow me clusterfuck and Stubbs

An issue was raised in the first weeks of Gamma 1 as to who moves when, and why, after all avatars and their followers have moved, some avatars fell behind.

This was due to non-deterministic handling of avatars – or to explain as simple as possible – wrong avatars moved first. For example, if A follows B follows C, and B decides to move first, B will move, pulling A onto C’s location, and then C decides to move and ends up hexes away from B and A. This is clearly wrong, but what is the correct procedure?

  1. Deal with everything that has fixed destinations. Avatars traveling to specific hexes. Avatars targeting thrones, or settlements.
  2. Move those avatars that are following avatars who are NOT following avatars. This means their targets have already moved in step 1, and their destinations are also considered fixed.
  3. What remains?

Only avatars following other avatars who themselves are following avatars. And here the potential circular references, internally called clusterfucks, appear:

This has been the cause of those moving inconsistencies. Originally we moved avatars on a first come (from the database) first serve basis, but this is obviously horribly wrong. See the red Stubb in the upper right that has a follower of his own? There is no technical difference between him and the clusterfuckers in the center – it follows an avatar and is followed by an avatar, but move him first and a mess happens.

Why? Because he will pull his chain of Stubbs towards his position, decide he’s done and stop prematurely. Then some CF moves and that particular Stubb + his chain won’t move at all.

Anyway, this is now solved. We determine the ringleaders and then poke one of them into moving. As a result they pull the entire cascade and the movement is more or less deterministic. Yes, it is possible that poking different clusterfuckers might result is slightly different end positions but the error is 1 hex at most and we’re not leaving any Stubbs behind, which is the most important benefit.

Abandon Elder

This is now a thing. There is no functional difference between abandon and banish – you get booted either way and your belongings get deleted. Your setts will remain but for a short time, and once you return it’s Trial all over again.


Artifacts are now smeltable. You can recover a small amount of forge mana when destroying them, but this functionality is necessary due to artifact caps that are now in place. So, get rid of the wooden pendants and focus on actual artifacts.


Nobles, Neayaless ones at least, are low-level thugs and enforcers for their Emperors. They have a “special” ability to ride to an enemy settlement, enter the Council Hall, put a knife to the Council of Elders’ throats and ensure their cooperation, at least for as long as they are in the room.

Only neayaless nobles (not emperors!) can do this, actual nobility has better things to do, like creating armies and waging war. As long as they sit on the settlement they deny the rightful owner of the settlement the omnipresence due and redirect it into their own coffers. But, as soon as the settlement loses the majority influence and the village council is disbanded, their methods become useless.

Avatar respec

Avatars can now be reset to base, and their points redistributed. Cost for such an action is level^2 * 200 mana, so it gets expensive quickly, and after level 10-12 it becomes really uncomfortable. It’s not that 25k or 50k mana is a large sum, but you need to have it on your throne and are risking a fissure in the process. However, the option is there, for those that really need it.

Settlement rebellions and draft costs

There is now a chance, a small but quite persistent chance that a settlement you stole from its rightful owner and have neglected to maintain (regular drafts) will suddenly rebel and return ownership to its original owner. The chance is low, don’t worry, but it’s possible.

Vagrant settlements – naturally spawning and those left behind when their owner GEd or was banished will not rebel. Hopefully this will prevent Sim City type of play – goal is to win, not amass 15 settlements which you have no chance to draft from due to mana issues.

Draft costs have also been altered. Each estate of distance will add 1 to draft costs. In practice there is rarely more than 5-6 estates between any two elders but the rising cost is especially painful if you steal settlements from the other side of the map via WCG or other means. Keep in mind that costs can be (drastically) brought down via military settlement specializations.

New sett images

I’ve been teasing new settlements over the last few Monolith Mondays. Here’ s a picture with a few of them – from left to right you can see the hamlet (trade lvl 1), camp (military lvl 2), fort (military lvl 3), and two villages (trade lvl 2) – one own and one without owner.

Very keen eyed will notice minor differences between the two villages (left one is haad, right one is erol). These are small differences that distinguish specific races but will be more readily apparent on the new settlement forms (once you click on the settlement) – but this hasn’t been completed yet since I’m waiting on the other settlement images.

Server migrations

A bit of back and forth – I decided to move the game server onto another machine, one more suited to running this type of software (3.5GHz Xeons), have moved and then moved everything back due to some quite foreseen technical issues. I have to finish these migrations this week and then we’ll be fully stable.

Anyway, this is all for week 31, next week it’s research crossbleed and other minor issues still remaining. Avatar ETA anyone? Geez that’s a tough one.