Afterburner:
Although my recent posts to this site may make me look like an intersection-obsessed nutjob, I am actually a big fan of housing blocks. So, I was following developments in your thread, here, with considerable interest. Then you posted reply 5, and I read it with horrified fascination, and realized shortly thereafter that I needed a change of shorts.
About halfway through reply 5 you show another variant of the flexiblock with a temple to Ptah in the upper left and you direct our attention to a column of three gardens nearby to the right. In the next figure you replace the gardens with a road-through-roadblock structure and reported that the addition of the road "shorted-out" (which is a term I like, too) the temple, so its priest neglected some housing. So far, my shorts are just fine, thank you.
Then I hit the bulleted list. ("Some experimentation reveals the following:")i. "Removing the road and roadblock and leaving those tiles blank (i.e. no gardens) short-circuits the path and destabilizes the block." (Sure, why not.)ii. "Removing the road and roadblock and replacing it with gardens re-stabilizes the housing." (Aiyaa! I'm a dead man!)The only difference I see between the two layouts is plus-or-minus gardens. Inescapably, you seemed to have observed a significant difference in walker behavior that varied with the presence or absence of gardens very near to a temple.
I just spent more hours than I care to admit recording walker paths on four-way intersections of vast, well-isolated straight roads which were marked off using gardens (!!), because my preliminary testing showed (I thought) that gardens were utterly inert, i.e., they had no discernible effect on walker behavior, and now I learn that gardens matter!?
Even worse, by my count (It's tough to be sure when glyphy gets mangled) it looks like the gardens would be seven squares away from the north square of your temple to Ptah, and I just knackered myself deducing that even a four-way intersection should not be seen by that temple at such a great distance.
So before I drag myself off to the smallest room in my house for a learning-experience-reinforcing round of self-administered "swirlies", I wonder if I could ask you to take pity on me enough to answer a couple of questions:1. In reply 5, are you still following your earlier convention of keeping north to the upper left when copying your game geometry to glyphy?
2. Is there any chance that you might have changed locations, lengths, or positions or either roads or road blocks within six squares of the north square of your temple to Ptah at about the same time you were recording the observations (that stop my heart) regarding the gardens' effects on walker behavior?
Many thanks for any additional light you could shed on these matters. Like User Maat says, "Inquiring minds want to know!"
StephAmon