RHW (RealHighway) - Development and Support

[Tarkus]

Still alive . . .





-Alex

[vinlabsc3k]

Are A1-B1 Inside Ramps for L1 ERHW 6s?

[Gugu3]

Are A1-B1 Inside Ramps for L1 ERHW 6s?

I can see 8s too

[compdude787]

Yaaaay, now we can annoy all our roadgeek sims with left exits and entrances!!!!

[Tarkus]

vinlabsc3k and Gugu3, you are correct.  All the height levels of the RHW-6S and 8S (and by extension, the 10S) will have A1 and B1 inside ramps, sure to annoy all the roadgeek sims, as compdude787 rightly pointed out.   The MIS and RHW-4 will as well.  And, as of a few minutes ago, the DDRHW-4 received an A1 Inside ramp. 



I also made the DDRHW-4 A1 Outside ramp operate as a FLEXRamp/Draggable Ramp Interface, and will see about the B1 Outside and Inside.  There likely won't be any more DDRHW ramps until the DDRHW-8 happens.

-Alex

[Gugu3]

Very nice Alex!

[APSMS]

vinlabsc3k and Gugu3, you are correct.  All the height levels of the RHW-6S and 8S (and by extension, the 10S) will have A1 and B1 inside ramps, sure to annoy all the roadgeek sims, as compdude787 rightly pointed out.   The MIS and RHW-4 will as well.  And, as of a few minutes ago, the DDRHW-4 received an A1 Inside ramp. 

<image snip>

I also made the DDRHW-4 A1 Outside ramp operate as a FLEXRamp/Draggable Ramp Interface, and will see about the B1 Outside and Inside.  There likely won't be any more DDRHW ramps until the DDRHW-8 happens.

-Alex

So...never? Or have I missed something about the RHW development plans?

[GDO29Anagram]

So...never? Or have I missed something about the RHW development plans?

Current DDRHW plans entail that the next DDRHW to be added be DDRHW-8 (no distinction between S or C), but that entails that it'd be, 1, MHW-based, and 2, possibly draggable. Why? Because it's easy to break apart than, say, DDRHW-6. Something that I wanna revisit, however, is the idea of an MHW-based DDRHW-4, but such an idea would have to render down the current DDRHW-4 useless and be placed down using helper pieces. I'm not sure if it's implementable at this point.

Also, how do you get a Type D1 ramp to work on a DDRHW-4? Do you deal with a DDMIS or do you have asymmetrical DDRHWs, like DDRHW-2+1? Type A1 and B1 are the most logical and the most workable ramp interfaces to add for DDRHW-4 right now.

[roadgeek]

Yaaaay, now we can annoy all our roadgeek sims with left exits and entrances!!!!

Hey! who are you calling a....

[APSMS]

So...never? Or have I missed something about the RHW development plans?

Current DDRHW plans entail that the next DDRHW to be added be DDRHW-8 (no distinction between S or C), but that entails that it'd be, 1, MHW-based, and 2, possibly draggable. Why? Because it's easy to break apart than, say, DDRHW-6. Something that I wanna revisit, however, is the idea of an MHW-based DDRHW-4, but such an idea would have to render down the current DDRHW-4 useless and be placed down using helper pieces. I'm not sure if it's implementable at this point.

Also, how do you get a Type D1 ramp to work on a DDRHW-4? Do you deal with a DDMIS or do you have asymmetrical DDRHWs, like DDRHW-2+1? Type A1 and B1 are the most logical and the most workable ramp interfaces to add for DDRHW-4 right now.

I think you missed my point. Given the public clamor for the RHW-10C and 12S networks, both of which have basic textures already drawn, and neither of which will be added in the foreseeable future due to controller size concerns/crosslinking, and the extremely low focus (already) on Dual height RHWs, I figured that the probability of the DDRHW-8 happening was practically nil.

Interestingly, I never considered the idea of a DDRHW-6. It seems logical, but for some reason my mind always jumped straight to the 8-lane variant anyways. Maybe since the DDRHW's purpose is maximum capacity in minimal space and DDRHW-6 has little functional benefit over the 4-lane variant? I don't know 

[GDO29Anagram]

I think you missed my point. Given the public clamor for the RHW-10C and 12S networks, both of which have basic textures already drawn, and neither of which will be added in the foreseeable future due to controller size concerns/crosslinking, and the extremely low focus (already) on Dual height RHWs, I figured that the probability of the DDRHW-8 happening was practically nil.

Interestingly, I never considered the idea of a DDRHW-6. It seems logical, but for some reason my mind always jumped straight to the 8-lane variant anyways. Maybe since the DDRHW's purpose is maximum capacity in minimal space and DDRHW-6 has little functional benefit over the 4-lane variant? I don't know 

Then you're really missing a lot of the RHW devplan here.

The thing about DDRHW networks is that they're gonna be using a completely different base network, and what's more is that you're really only adding one network to the mix; adding one network to the NAM doesn't do anything to affect controller size, but adding six of them that need to depend on each other does. Plus, we've already had DDRHW-8 planned for ages.

This is why RHW-RHW crosslinkages are so evil to deal with: we're using only one base network. Consider L1 RHW-4 and L0 RHW-4, and then work our way up. If you crossed those two networks with an RHW-2, you'll get an RHW-2×RHW-2 crossing.

First, you would need override code to convert an RHW-2×RHW-2 crossing into an L0 RHW-4×L0 RHW-2 crossing, and then do the same with the L1 RHW-4 to get an L1 RHW-4×L0 RHW-2 crossing.

That's fine if you're crossing the base network with these two override networks. It gets harder if you're crossing the same two networks together.

In that case, one of two things can happen when you cross L0 RHW-4 with L1 RHW-4: you'll either get the L1 RHW-4×L0 RHW-2 crossing or the L0 RHW-4×L0 RHW-2 crossing. Since there is uncertainty as to which one happens, you need to write code for both situations in order to override either case into the end crossing: L1 RHW-4×L0 RHW-4. Add on top of that a lot of stabilise code to keep the whole thing from deconverting when you click something from five tiles away.

So, you're basically overriding an overridden override. That's just half of the problem. The rest of the problem comes from having multiples of these crossings next to each other, and that's what makes crosslinages so code-hungry. Add to that again a lot of stabilise code to keep the whole thing into turning an ugly mess.

There is a way of mitigating that issue, but it only works if you ditch override networks altogether: use a second network and use that as a helper piece.

This is really hard to understand unless you have a comprehensive understanding of RUL1 and INRULs, but basically, you can theoretically define an entirely new network just by using another network's corresponding INRUL file and then define all of the needed crossings using only RUL1, eliminating half of the needed override code.

The one downside is that such a network would have to be rendered down to being placed using helper pieces instead of being drawn like a base network.

If there were a theoretical .dll that allowed the NAM Team to add its own base networks, then you'd eliminate the entire issue with RUL2 entirely, since every single network in the RHW will be basically a new unique network. The downside is that such a plan would require over 30 different network buttons, and that some of those networks are 3 tiles wide, making drawing those network an issue.

With the case of DDRHW-8, the plan is that such a network be based off of MHW. Whether or not it'll be helper-based or an MHW-based override network is uncertain at this point. If it were an MHW-based helper-based network, you'll be basically avoiding all of the issues that using one single base network would entail. This is why DDRHW-8 would be so easy to add. Additionally, DDRHW-8 doesn't really need to be buddied up with multiple networks parallel to it, probably MIS and nothing else, unlike all of the other networks, where you can basically buddy up one network of one width with nearly anything you want.

[Tarkus]

To add onto Ganaram's comments, the RHW-10C and 12S, in addition to creating a lot of complicated crosslinks, don't really add anything new to the table functionally.  The 6C and 8C already occupy the same footprint as the 10C, and the 12S has the same footprint as the 8S and 10S.  I don't think we need three different widths with the same capacity.  The idea of exploiting the higher capacity assigned to the base MHW network has been suggested, but the problem lies with split tile vs. shared tile on the diagonals.  The 10C and 12S are pretty much permanently tabled, until/unless we find some way for them to actually be useful and worth the effort.

The DDRHW-8, however, has been planned since even before the 10C and 12S received prototypes.  There's never been a DDRHW-8 prototype in game, but the plan would be to go for an MHW-based network, which would get a capacity boost by virtue of crossover paths, which act as DIPs (distilled intersection paths).  It would effectively be the highest-capacity highway-type network on a per-tile basis, and thus, add a new capacity/speed niche into the RHW hierarchy.  The shared-tile diagonal issue wouldn't be a problem, and there'd be very few ramp interfaces needed for it. 

The fact that there are no other MHW-based override networks would mean the crosslinkage would actually be quite simple.  The biggest complication is the MHW's AutoPlace tendencies for overpasses and interchanges.

-Alex

[jdenm8]

To add onto Ganaram's comments, the RHW-10C and 12S, in addition to creating a lot of complicated crosslinks, don't really add anything new to the table functionally.  The 6C and 8C already occupy the same footprint as the 10C, and the 12S has the same footprint as the 8S and 10S.  I don't think we need three different widths with the same capacity.

Additionally, proper mainline stretches of of 12S are rare as hell. It's much more common that they're simply acceleration lanes for very wide motorways. The longest continual stretch of 12S I know of are the Gateway and Sir Leo Hielcher bridges.

[APSMS]

Now that answers a lot actually. I could already see the reasons for not adding the two additional ground networks. Handy, but hardly justifiable under the current paradigm.

I forget what exactly INRUL code does (not that I had that great of an idea to begin with), but the whole MHW base for DDRHW-8 makes a lot more sense. Not sure how difficult having to use helper pieces would make construction, esp in the middle of a FLEX-based development cycle, but I can see how it minimizes the code needed. Would the plan be to make DDRHW-4 use the same base, utilizing the split MHW dragging method? Or is it to stay RHW/ANT-based?

All the new elevated stuff is pretty neat, too, since it will let us build more compact interchanges, with a lot more smoothness.

[Tarkus]

I forget what exactly INRUL code does (not that I had that great of an idea to begin with), but the whole MHW base for DDRHW-8 makes a lot more sense.

INRULs handle situations involving just a single network.  The Basic INRUL defines how the base network is dragged, based on the network flags, and the Advanced INRUL allows one to specify what happens in a 5x5 tile radius, based upon flags.  We are using the Advanced INRUL for the RHW network (INRUL14) to define the FLEXRamps and DRIs.

Not sure how difficult having to use helper pieces would make construction, esp in the middle of a FLEX-based development cycle, but I can see how it minimizes the code needed. Would the plan be to make DDRHW-4 use the same base, utilizing the split MHW dragging method? Or is it to stay RHW/ANT-based?

Helper pieces basically are FLEX pieces.  They're "pseudo-draggable items", in that they're placed like puzzle pieces, but they behave as if they're draggable, rather than behaving like a standard puzzle piece.  Essentially, rather than dragging the network, you'd be building it from FLEX pieces.

All the new elevated stuff is pretty neat, too, since it will let us build more compact interchanges, with a lot more smoothness.

Thanks, and there's more where that came from.



The addition of all these elevated ramps will finally make the long-term RHW vision seem clearer.

-Alex

[roadgeek]

... The longest continual stretch of 12S I know of are the Gateway and Sir Leo Hielcher bridges.

You ought to check out the new Grapevine Funnel http://www.dfwconnector.com/

[GDO29Anagram]

You ought to check out the new Grapevine Funnel http://www.dfwconnector.com/

Again, not a very long stretch of 12S (and it's not even shown to be complete on Google Maps). The highways down in San Diego are known to be wider and longer, but that's only one of probably two exceptionally large highways in the world, the other being a 14S equivalent on the other side of the US. There's one more in California that's said to be an RHW-26 equivalent, but that's just multiple carriageways stitched together, and that's for an interchange only, not an entire highway.

No, the über-wide roads in Moscow don't count, either; those would be RD-18 equivalent, and NWM's totally another thing altogether.

[APSMS]

Someone brought up San Diego! Some of the freeways here are of ridiculous size, especially when compared to LA freeways and the fact that many were built around the same time.

Check out Interstate 15 from the 163 Junction up to Escondido. I think most of it can be made with existing RHW 10S, but there are some stretches...

Also, I think the expansion of the I-805 in South San Diego has a few stretches that qualify. Also check out the I-5 near Santa Ana and Anaheim. Again the stretches are relatively few and small (though usually more than the 4km SC4 tile) and the exception to the rule, so the sacrifice/axing of the extra networks is unde

[MandelSoft]

Houston can be pretty insane too. I had a geek-out moment when I passed this stretch of the US-59 there:



Seven lanes next to each other on a single roadway!


In the other direction it's the same:


And how about the I-10 Katy Freeway, just a few miles north-west of the previous stretch?


Seems like everything is bigger in Texas. I think the widest undivided roadway on a motorway here is 6 lanes in one direction, but only for a short stretch. The A2 and A4 have longer stretches of 2x5 lanes.

[GDO29Anagram]

Someone brought up San Diego! Some of the freeways here are of ridiculous size, especially when compared to LA freeways and the fact that many were built around the same time.

I bring up San Diego because my brother and I travelled to the San Diego Zoo last year, and dude, those things get insanely empty at 06:00 on a Sunday, and based off of memories from before 2013, insanely congested at 15:00 on a weekday.

Houston can be pretty insane too. I had a geek-out moment when I passed this stretch of the US-59 there:

Correction: THREE exceptionally large US highways.

And don't get me started on Canada's Highway 401.