Yet another cargodist question

[Lollipop]

The wiki says about asymmetric distribution:

Asymmetric distribution will distribute the cargo or passengers in the network without such restrictions.

So how does asymmetric distribution choose it's destinations then? My current working theory is: nothing apart from distance matters. Maybe saturation, but assuming unsaturated routes, station A would distribute 100 food to B and C -both 20 tiles away- at a 1:1 ratio. It doesn't matter, if A is a 50 ppl hamlet and B a metropolis or if I use (enough) horse cargo trams for A and 747's for B.

Is that correct?

The reason I'm asking: I'd love to get cargodist working for asymmetric cargo, but on a simple iron to steel mill route I don't need it, on a more complex distribution scheme it fails. Because distance is the least relevant criterion. In the context of a GS script, where settlements need food to grow, the poor hamlet would be drowned in food while the metropolis would starve to death.

I guess I could influence the ratios by station placement, i.e. 2 tiles away from the metropolis and 18 tiles away from the hamlet, but that leads to kinda artificial layouts and a inflated number of stations.

Do I miss something, it wouldn't be the first time

Thanks

[Alberth]

Don't know how cargodist works either, but influence of distance is a separate setting. By default it's at 100%, which means it's an important notion. Reduce it to lessen its impact.

Symmetry of cargo is not about distance, it's about volume, as far as I know. It tries to give you an equal amount of cargo in both directions. I believe it's a silly restriction, and always turn it off.

cargodist working for asymmetric cargo, but on a simple iron to steel mill route I don't need it, on a more complex distribution scheme it fails

I don't see why you don't need it for one way cargos, but maybe cargodist has an exception for it.

What does "fail" mean here? Stuff isn't balanced? Of course it's not. You give cargo-dist freedom in picking destinations, and it does so without consulting your complicated distribution scheme. The only fail I see here is that you want control over where stuff goes, but the whole point of cargo-dist is that you don't have it. The program decides where cargo goes, you adapt the network to it.


It's the same with breakdowns. You can argue breakdowns are broken, and disable them so you can have heaps of unreliable engines packed onto a few tracks. You can also see it as a challenge, pick your engines with care (reliability becomes the primary driver for selection/replacement) and make your network robust against the occasional breakdown.

[Lollipop]

Stuff isn't balanced? Of course it's not. You give cargo-dist freedom in picking destinations, and it does so without consulting your complicated distribution scheme.

You're right. I want the cookie and the cake. Please auto-distribute, but obey my rules, I'm aware of that. I'm just hoping, that someone smart -smarter than me- comes up with a solution, that gives me both cookie and cake.

You can argue breakdowns are broken, ....

I'm not saying anything is broken. In fact I can't imagine playing Pax without cargodist.

[andythenorth]

Yes, for asymmetric, the only effect on demand is distance. If distance effect is set to 0%, demand is level for all linked destinations).

I had a long irc discussion with fonso (cargodist author) about it one day, and I should update the cargodist wiki page, but eh

Asymmetric cargodist mostly works well for me with FIRS, including distributing supplies cargos BUT I use multiple pickup stations at industries, each serving only one or two destinations. TL;DR that works better for the way I play. Cargodist is basically 'automated transfers' using this method, and I control the distribution. I also have distance effect set to 0%.

[HGus]

but on a simple iron to steel mill route I don't need it, on a more complex distribution scheme it fails.

This video shows an example for industries, hope it is usefull for you.

https://youtu.be/s0Jo94BtRyo