RHW (RealHighway) - Development and Support

[z]

Hope that helps!

It does very much.  Thanks! 

So as of my Beta 1 version of Simulator Z, I will have the first intersection capacity number be 100%.  If RHW and/or NWM are released before my Beta 1 is ready, I will release a patched Alpha 4 version that has that 100% number in it within a day.

From your answers to my questions, it sounds like 100% is actually the ideal number for the first intersection capacity number.  With 100%, crossover traffic will cause some congestion (though minimal, and only if the road is at or near capacity).  But that's how it works in real life, isn't it?  If you have a highway at or near capacity and people start changing lanes, that starts to slow things down.

Which brings up the question of the 150% figure used in Simulator A.  Doesn't that essentially have the effect of increasing the capacity of all the wider RHW's by 50% over their nominal capacity?  And it would seem to have the opposite problem with the capacity of the wider RHW's (and the TLA's) that the 70% figure does.  If I read your capacity table correctly, you have included this effect in there, and with Simulator A, an RHW-8 has three times the capacity of an RHW-4 (as just one example), which certainly doesn't seem right.  Shouldn't that 150% number in Simulator A be 100%?  That would also make Simulator A behave more realistically at intersections; I can see no downside to such a change.

[choco]

bridge functionality....check.
100b functionality....check.

can't wait till this gets out so i can fill in some holes....

looking great alex

[elliey-sama]

Oh i remember those walls :3 sadly i cant find them anywhere can anyone please tell me where i can download those freeway walls? i used to have them but now i cant find them anywhere on Simtropolis or here because i dont remember the name to well, so if anyone knows the name and where to download them let me know ok? Thank you!

[wes.janson]

elliey-sama:

Try looking here -- > http://www.simtropolis.com/stex/index.cfm?id=12982

[berubium]

Hello, long time lurker, 1st time poster...

It's been great watching all the progress thus far for RHW 3.0...  I am just as excited as anybody to start implementing all the new changes in my cities.

I have one question regarding textures.  This has probably already been dealt with, but I haven't had any luck finding out for sure.  When we develop land adjacent to a RHW puzzle piece (zone or plop something), will odd textures keep showing up in place of grass, or has this been addressed?  Either way, I want to say a huge thank you for making SimCity that much better.  I always hated the narrow Maxis 6-lane highways with unrealistically short on & offramps.  RHW is fantastic!!!

[Tarkus]

Shouldn't that 150% number in Simulator A be 100%?  That would also make Simulator A behave more realistically at intersections; I can see no downside to such a change.

A very good question to ask.    As per the Understanding The Traffic Simulator thread, this was Jason's initial justification for the 150% value in Simulator A:

This causes the Simulation to act as if the 2 Imediate Tiles are "slow downs" for the intersection when coupled with a properly designed Congestion vs Speed Property.  This setup makes is so that the intersection has a proportionally higher capacity since there are two separate routes.  This increase in Capacity at the intersection tile allows for both routes to be used to almost fully and not being hampered by the intersection.

In other words, it is because the intersection is the point at which two networks cross and their volumes combine, they would be at 100% capacity when the networks entering the intersection are at 50% capacity (at least for a 4-way intersection)--assuming a 100% value for the first figure in the Intersection and Turn Capacity Property.  With the default Maxis 70% value, 100% capacity is reached at 35% capacity for surrounding networks, and with the 150% value, it's still 75%.

And of course, if you look at a lot of real-world intersections, of various types (signalized, 4-way-stop-controlled, roundabouts), the point at where the congestion occurs is not the actual point where the two roads intersect, generally speaking.  The congestion results from the queuing created by the traffic control device--people waiting at a stop sign, at a signal, or to enter a yield-controlled roundabout.  The only times when I've seen the actual intersection itself become congested is when two intersections are too close together (i.e. two signalized intersections less than 500 feet apart) and the queuing for one of them actually backs up enough to block the other.

When Jason went with the 150% value, we had not yet discovered that the use of crossover paths on the Wider RHWs and the TLAs in the NWM caused the game to treat them as intersections and apply the Intersection and Turn Capacity to them, either--the reasoning for this value was entirely derived from the actual Intersection capacity. 

And while it does end up meaning that the RHW-8 has 300% of the capacity of an RHW-4, it does have the beneficial effect of making the TLA-3 network in the NWM have 150% of the capacity of a road.  And provided I could find some way to add "dummy crosspaths" to the RHW-6S and the NWM's ARD-3 (Asymmetrical Road--1 lane on one side, 2 on the other), it would actually allow for realistic capacities on those networks.  This technique could also be used to mitigate the "shared tile" congestion effect on Diagonal Avenues.

There are definitely some trade-offs that have to be taken into consideration no matter what.  100% allows for realistic Wider RHW capacities, 150% allows for realistic TLA-3 capacities (and potentially, RHW-6S and ARD-3 if I figure out the crosspath trick).  Each one has its advantages and disadvantages. 

If you do end up using 100%, it will ensure at the very least, though, that the capacities of the Wider RHW and NWM networks will not be less than their base networks (RHW-2, Road, OWR), and it would be RHW 3.0-compliant.

And berubium, welcome out of lurking, and thanks for the kind words!  To answer your question, the so-called "texture artifacting" glitch you are referring to will be fixed with Version 3.0.

Also, just to give everyone a "heads up" on the RHW development front, I managed to knock out a good number of path files last night, so the NAMites and NAM Associates will be getting a little something relatively soon.   Slowly but surely, we will get there.

-Alex (Tarkus)

[Swamper77]

The models I have been creating for Alex's Elevated RHW and MIS are almost done. I just have a large puzzle piece to model and cut up. After that, I'll just be making quick fixes here and there to wrap everything up. I'll be gone from this Friday (11/21) to this Sunday, so I won't be at my computers until Monday at the earliest. With Thanksgiving this next week, I probably won't get any modeling done until after Thanksgiving.

-Jan (Swamper)

[z]

There are definitely some trade-offs that have to be taken into consideration no matter what.  100% allows for realistic Wider RHW capacities, 150% allows for realistic TLA-3 capacities (and potentially, RHW-6S and ARD-3 if I figure out the crosspath trick).  Each one has its advantages and disadvantages. 

I agree; there's no perfect solution here.  But I'd like to add a few points for consideration.  First of all, for the TLA-3, should it really have three times the capacity of a road?  I don't think so; the middle lane is by definition a turning lane, not a travel lane, and has far less volume than the surrounding lanes.  So the volume might be more than 100% of a road, but not by much.

As for the RHW-6S and ARD-3, I'm going to step off the deep end here and ask about an area about which I know very little.  Is there any way to use the unused power line and water pipe network capacities and speeds here?  That would certainly solve a lot of problems, but I have no idea if it's at all possible.

In other words, it is because the intersection is the point at which two networks cross and their volumes combine, they would be at 100% capacity when the networks entering the intersection are at 50% capacity (at least for a 4-way intersection)--assuming a 100% value for the first figure in the Intersection and Turn Capacity Property.  With the default Maxis 70% value, 100% capacity is reached at 35% capacity for surrounding networks, and with the 150% value, it's still 75%.

Yes, this is true.  But it is offset by the recent discovery that no matter how congested a network is, its speed will never drop below 30% of the nominal network speed, regardless of what's specified in the CvS curve.  This was not known when jplumbley made his 150% choice.   Maxis knew this, of course, and I'm sure that's one reason why the intersection value was originally set to .7.

I think it's useful to look at an example of a four-way intersection with stoplights.  In the real world, you may typically have a 30-second cycle for the lights.  Since they're red approximately half the time, that means that the average delay for a car is 15 seconds, plus whatever time is lost by slowing down and speeding back up.

Now take the case where the three values of the Intersection and Turn Capacity Effect are all set to achieve maximum congestion over the five affected squares on one of the roads through the intersection.  In Simulators A, B, and Z, the speed of a car on a road is conveniently about one square per second.  So with no congestion, those five squares would take about five seconds to cross.  With maximum congestion, the time required would be 5 seconds * 1/.3, or about 17 seconds.  This means that with maximum congestion, it is possible to produce a delay of 12 seconds, which is a little bit less than a standard stoplight cycle.

But at 150% setting for the intersection square, speed through the intersection is reduced to 67% of nominal in Simulator A, as opposed to 30%.  This makes the best intersection simulations even farther from real life.  On the other hand, the 100% value will reduce the nominal speed in the intersection square all the way down to 30%, assuming that the two intersecting roads are both at full capacity.  But if either of them are less than full capacity, then the congestion of the five squares centered on the intersection on that road drops, and the lower the capacity, the less the drop in speed.

I think it's also important to ask:  What is the relative usage of intersections vs. TLA-3's, RHW-6S's, and ARD-3's?  Everyone playing the game will use intersections and be subject to the intersection effect.  But the number of people using the above specific roads is going to be far less than 100%.  Therefore, it seems to make sense to favor the intersections over RHW and NWM when choosing values for the Intersection and Turn Capacity Effect, except when doing so causes major problems.

And of course, if you look at a lot of real-world intersections, of various types (signalized, 4-way-stop-controlled, roundabouts), the point at where the congestion occurs is not the actual point where the two roads intersect, generally speaking.  The congestion results from the queuing created by the traffic control device--people waiting at a stop sign, at a signal, or to enter a yield-controlled roundabout.  The only times when I've seen the actual intersection itself become congested is when two intersections are too close together (i.e. two signalized intersections less than 500 feet apart) and the queuing for one of them actually backs up enough to block the other.

Very true, and this occurred to me once you made it clear that only the intersection capacity number needed to be changed.  So in some ways, moving that number up to 100% and moving the others down would make the intersections more realistic, and considering the problems that this solves, I'm quite willing to go along with it.  It makes having a true "stoplight effect" more difficult, although not significantly.  But this is true only using the 100% figure.  For this reason, as well as the other reasons I listed above, I would urge that the simulators standardize on 100% for the first Intersection and Turn Capacity Effect number.

[Thomas Diamond]

The models I have been creating for Alex's Elevated RHW and MIS are almost done. I just have a large puzzle piece to model and cut up. After that, I'll just be making quick fixes here and there to wrap everything up. I'll be gone from this Friday (11/21) to this Sunday, so I won't be at my computers until Monday at the earliest. With Thanksgiving this next week, I probably won't get any modeling done until after Thanksgiving.

-Jan (Swamper)

Your were very close to release a couple of weeks ago so do you think the RHW 3.0 will be out in  mid December?
Also ,do I have to delete the RHW 2.0 in my Plugins folder to install the RHw 3.0?

[Andreas]

I guess this "incident" shows very well why release dates are not practical. You think everything works ok and the release might be soon, but then, you or other people discover a few issues, then RL gets in the way, and soon enough, the project is delayed by a month or even longer.

To answer your last question, yes, RHW 3.0 will replace the current 2.x version completely. If you're using the installers, the replacement should be automatic.

[metarvo]

I hope I'm not nitpicking here, but I have made an observation.  The "What timeframe do you think RHW Version 3.0 will be released?" poll is completely zeroed out.  All I see are a bunch of 0's staring at me.  Is it Alex playing with us again?

[bob56]

looks like the pools been locked...

[un1]

looks like the pools been locked...

If you mean poll then yes it has been locked. And that is very, very, very, very ... good!
That means that he has gave up on keeping us from guessing, because it is going to be released soon. (I hope  )

-un1

[nerdly_dood]

I hope I'm not nitpicking here, but I have made an observation.  The "What timeframe do you think RHW Version 3.0 will be released?" poll is completely zeroed out.  All I see are a bunch of 0's staring at me.  Is it Alex playing with us again?

Whoa! Where did the numbers and graphs and my vote for late November go?

Oh well... Waiting patiently for v.3 release.

I typed that before un1 posted, and now that he's posted I'd rather not think hard enough to take his opinion into account... Brain hurts from computer problems.

[Tarkus]

The poll has been running for quite awhile now and I think it's run its course now.  No significant reasoning behind it, I can assure you.

But I do have a few new screenies here . . .

Want rural roundabouts?  Well, you can now plug RHW-2s into the OWR roundabouts, and get something like this (minus the paths, unless you have DrawPaths turned on).  Before, it just had the Road/OWR Roundabout intersection textures.



And the draggable RHW-4/GLR crossing.



-Alex (Tarkus)

[metarvo]

That's spectacular, Alex.   Now I can recreate that rural roundabout (in my area it was called a traffic circle ) that I saw once.  Of course, the RHW-4 x GLR crossing is also very good.  I'll take any added RHW functionality or crossing that I can get.  Great work, Alex.  I guess you know now that I'm quick to notice any change in this thread.

[nerdly_dood]

Yay! Now I don't need to transition to a road for a roundabout! And great work with the GLR crossing too.

[sim-al2]

I would never drag a GLR across a busy highway, but it's a good addition anyway. The roundabout seems okay, but the transition is a bit harsh and there's arrows sitting on bare ground. It might make the DOT look a bit better if they weren't painting road markings on grass! 

[Monorail Master]

For that RHW/GLR crossing, since I'm studing civil engineering in h.s., I would put some train crossing gates for the MIS-1, RHW-2, and the RHW-4

[io_bg]

The roundabout looks great but for me the GLR crossing isn't a good idea. I don't think you can find something like this in the real world. What would happen if you drive with 100 kph and then suddenly a tram appears in front of you?