I don't know nearly as much as I would like about using forced walkers, but I have seen unexpected behavior when using long loops that utilized gardens for part of the path. In a contest a while back, I used a loop that utilized both roads and gardens in one end of the loop. In the test phase and through the first half of the scenario it worked perfectly. But at some point after developing other parts of the map that were completely unrelated to the initial loop, it suddenly stopped working properly. I tried several solutions but was unable to fix it or to isolate exactly where the problem started. I finally just wrote it off to the highly scientific and precise concept known as "one of them things".

I know this doesn't really help you with your problem; I guess it's just my attempt to empathize. Good luck!

Thanks, Vectorgod. Your comment does help, in that it leads me to think that what happened may not be a random glitch of my computer. Of course, that's worrisome for the rest of the city's construction, but I'd rather know.

About weird spawning: something similar happened to me too. A school built in the middle of 32 tiles loop was sending out 3-1 children (NW-SE). After putting gardens for desirability for the GI, it started sending 0-4, but deleting and rebuilding solved the problem.

I saw another unexpected spawning. In every case I examined before now, when a farm (like most buildings) touched multiple unconnected roads, the citizen (and cart pusher) spawned on the longest of the roads. However, while trying to build a city of 320 luxury palaces, spawning occurred on the short road for 4 grape farms (in essentially identical layouts). There are also 8 wheat farms (2 places in the same 4 essentially identical layouts) which touch multiple unconnected roads, but each of those farms spawned on the longer road (as expected).

Once again, it was fairly easy to fix the problem, but that isn't very reassuring.

I'm having similar problems. but luckily I was able to solve them.

This is the food block I'm using:



I have a few of those blocks. They have no road connection to the outside. The forced markets travel through gardens.

The weird thing is, granaries and roads that I build on the other side of the map cause these markets to stop working. Markets start getting access from the wrong side and not from the granary side.

If the finish tile of a forced market is a corner tile this wouldn't happen. But I had to squeeze in a barber, doctor and bathhouse to the food block and therefore had to finish tiles on the sides of the markets instead a corner.

The solution to these problems is to have more road tiles on the next block. For example if a food block has 58 road tiles, the next one should have 59 and the next 60 tiles and so on.

But I think it would be much better to avoid these problems by using corner tiles as the finishing tile for forced markets.

However, as far as weirdness goes, nothing tops the list like this colosseum that needs his side covered with aqueducts to work.



Normally you would use a gatehouse if you want an unstaffed coloseum with shows. I was going to do just that when accidentally I discovered I can do it without the gatehouse if there is an access ramp and a road tile like in the image.

This worked fine and I build a palace block with that but as I continued building things elsewhere on the map the coloseum stopped working.

After lots of experiments I found that building a low bridge on one of the fishing lakes made the coloseum non-functional. I need the low bridge so I continued experimenting with adding and deleting road tiles to see if that makes it work again but it didn't.

I remembered that walls can get rid of farmfield. So I thought maybe they have an effect on the terrain and started building walls around the coloseum. To my surprise the coloseum started working again.

I thought it is weird. If I build 5 tiles walls next to the coloseum, the lion house sends lions. If I delete the walls it stops sending the lions.

I made some more test and found out that walls have no unique quality here. It can be aqueducts or farms. The coloseum just needs its side to be covered with something.

[This message has been edited by Philon (edited 07-17-2012 @ 10:15 AM).]

Hi Philon,
Yes. Perhaps someday we'll find out whether a few of these rare things are bugs or actually have a purpose.

Philon,

I don't recall seeing anything like you describe before. I'm wondering if the objects you place against the colosseum (or indeed elsewhere on the map) would interfere with the shortest path to the map entry point? Also, are there other colossea disconnected from the lion house that supplies the one illustrated? Just curious ...

Here is a sav file if you found that interesting.

When you open the file the lion house is sending lions to the coloseum. But if you delete the aqueducts it stops sending.

In my city I have two coloseums and they both need their sides covered to work. The location of the lion house doesn't have any effect. The roads in the city don't have any effect either.

I don't think it has something to do with the entry point. In an earlier version the entry point was on the other side of the map.

In this version the low bridge effect seems to have gone because the distance to the lake is now different.

Hi,

Thinking again, it might indeed have something to do with the entry point. When the entry point was on the left side of the map I had to cover the left side of the coloseum to make it work.

Deleting a low bridge on the left side would also make it work without the side covered. The low bridge didn't have any road connections. It had gardens on both sides.

[This message has been edited by Philon (edited 07-18-2012 @ 05:16 AM).]

I have no explanation at present, but I find that I can delete the tile of aqueduct against the colosseum's northern tile without 'breaking' it. I also find that the blockage does not have to be right against the colosseum - I can move it several tiles further away (not determined just how far) provided I continue to block the same five 'x' coordinates, with the further proviso that if I move the structure beyond the steps immediately to the north-east I must also block any path up the steps. Very strange.

Inspection of the building table does not yield any clues.

It is a strange thing and I have no explanation either but I don't mind it. At the moment the coloeseums work exactly the way I want them. They have no labour, no road connection, no gatehouse but they have shows.

I don't think we can find some useful feature from here. Maybe there are no great features to discover anymore. Brugle has already found them all.

[This message has been edited by Philon (edited 07-18-2012 @ 02:44 PM).]

My guess (based on a few observations) is that the colosseum uses the connection based on the tile in offset 26. If the connection is on the ramp (stairs), then lions can get there through the ramp. If the connection is not on the ramp, then lions cannot get there through the ramp.

Hi Brugle,

I have no idea what offset 26 means. Is that a tile 26 tiles away in a certain direction from the coloseums north tile?

After Trium's comment I tried moving some of the 5 required aqueducts. I can move them if the bottom part remains covered like in the middle graphic below. The top part can be completely open. But as soon as one tile is open on the bottom part (like in the third graphic) it doesn't work anymore.

[This message has been edited by Philon (edited 07-18-2012 @ 03:44 PM).]

It is like there is a diagonal flow towards the coloseum from the entry point (from bottom right) and you need to block the flow. The top tile of the coloseum's side doesn't need coverage because diagonally you can't reach the corner.

Philon,

Offset 26 is the offset into the building's table, as described in Cartpushers, traders, etc. - which building is nearest?. (Trium strikes again!)

There appears to be a bug in reply #13's diagram, since the third picture is just the second picture with some more aqueducts.

I fixed the third graphic. The difference between two and three is that, two has its top part open and it works, three has its bottom part open and it doesn't.

I was experimenting more and when you cover the access ramp near the entry like this, you don't need any other aqueducts:

[This message has been edited by Philon (edited 07-18-2012 @ 04:14 PM).]

Quoted from Philon:

and when you cover the access ramp near the entry like this, you don't need any other aqueducts:

As expected, if my explanation is correct. Please read the thread I referenced.

Quoted from Brugle:

My guess (based on a few observations) is that the colosseum uses the connection based on the tile in offset 26.

Naturally I was looking for something to do with that. The colosseum in Philon's save has its 'access tile' set to the northernmost tile of the flight of steps against which the building sits, i.e. two tiles to the north-east of the plaza tile which serves as the destination for lions. As you say, those tiles are connected.

When I delete the aqueduct the 'access tile' becomes the tile adjacent to the colosseum's eastern tile on the north east side. This is not connected to the plaza tile. What remains a mystery to me is why that tile is selected.

Quoted from Trium:

is why that tile is selected

Why is that tile selected to be the access point? Because it is closer to the entry point.

Why is that tile selected to be the one which gives the colosseum its road number (and therefore determines whether an entertainer school can send its entertainers there)? Because that is how a building gets its road number. (I'm assuming that each tile in the map has a road number, or 0 if it isn't something that behaves like a road.)

Quoted from Brugle:

Why is that tile selected to be the access point? Because it is closer to the entry point.

I considered that possibility but didn't have time to test. As far as I know, that has not been hypothesised before - my description in the thread referenced in reply #15 was that the 'access tile' was the first road tile encountered when circling the building clockwise from north or, in the absence of a road tile, the first empty tile encountered similarly.

I did note in my description of the Building Table that I had seen anomalous cases which I could not explain, but it never occurred to me that it was the direction of the shortest orthogonal (non-diagonal) route to the entry point that was affecting the tile selection.

I've now tested and found that Brugle is right (I've edited my earlier description accordingly) so that particular discovery counts as one of his. Brugle strikes again!

Note - where a road exists the access is still the first tile clockwise from north regardless of where the entry point is)

[This message has been edited by Trium (edited 07-22-2012 @ 06:37 PM).]