NWM (Network Widening Mod) - Development and Support

[GDO29Anagram]

@ michi_cc: WHOA!!! iDid not expect that*...

A few things: I want to see that with the grid set up. (Nah, I can just copy the pic and overlay a grid on top...)

What did you use for the RA? Did you Photoshop a piece of the MAVE-4?

Here's what I'm thinking: Automata impossibilities include double-decker networks of the same network; Cars'll jump between the two decks... iThink your request is possible, but iWanna hear it from a NAM associate as well... The problem I personally see is all the textures and coding/pathing needed to make the RA compatible the way you illustrated. I'm not sure myself, since I know nothing of how the NAM's put together; iJust know how textures are assigned...

But as I said before, how would AVE-6 be connected (It's a three-tile network) and if we're gonna want an RA to connect to an AVE-6, we're gonna need a 5X5 RA...

I just remembered: There are plans for another asymmetrical network that Tarkus dubs as AVE-5. (Tark, in my opinion, that's a misnomer; Its more accurate name should be MAVE-5 or ARD-5). It's a combined MAVE-6 and MAVE-4; How would you connect that to an RA? (That's a stupid question; it's a half-and-half fix for that...)

Maybe I should try making something...

* - I meant your picture of the RA with the connected MAVE-6, MAVE-4, AVE-4, and OWR-1. Your recent picture, however, may need more smoothening out. I wanna try out making a smoother one...

Wait,... Can't we just use the MAVE-4 to AVE-4 transition piece and then slam an RA on top of the AVE-4 bit...?

[jdenm8]

In my mind, MAVE-6 will have to lose a lane for the purposes of the roundabout. It's not too unusual here for a six-lane avenue to lose a lane on both sides on approach to a four-lane roundabout. You just can't have three lanes coming off of a two-lane roundabout.

AVE-6 will be a problem though. Here anyway, most six-lane roads (if losing a lane isn't an option) are built with an overpass (usually four-lane) following the primary route over any roundabouts. You usually have to take an offramp (exiting lane and new lane) down to the roundabout.

[marsh]

     Is it just me or is the OWR nwm starting to turn into a secondary RHW? 

[Shadow Assassin]

In my mind, MAVE-6 will have to lose a lane for the purposes of the roundabout. It's not too unusual here for a six-lane avenue to lose a lane on both sides on approach to a four-lane roundabout. You just can't have three lanes coming off of a two-lane roundabout.

Alternatively, you can have the extra lane become a slip lane. Pretty common here, too.

[Tarkus]

I just remembered: There are plans for another asymmetrical network that Tarkus dubs as AVE-5. (Tark, in my opinion, that's a misnomer; Its more accurate name should be MAVE-5 or ARD-5). It's a combined MAVE-6 and MAVE-4; How would you connect that to an RA? (That's a stupid question; it's a half-and-half fix for that...)

MAVE-5 is actually its official name, and the one I used back in June.  It's also referred to that way in the little bit of RUL 0x10000002 code that's been in place for it since before the first release.

The "ARD" designation is kind of an odd one-off right now . . . there is one more ARD network planned, however--a dual-tile ARD-4, which is basically a 3+1 setup, along the lines of Southeast Morrison Street in Portland, OR.  It is being looked at in terms of long-term, post-Version 2 asymmetrical network efforts.

Wait,... Can't we just use the MAVE-4 to AVE-4 transition piece and then slam an RA on top of the AVE-4 bit...?

Yes, that is indeed possible--this is what it ends up looking like:


I'm intrigued by the other ideas proposed here as well, and admiring the nice texture work.   The implementation aspect with further roundabout functionality is ultimately key.  The existing Avenue Roundabouts are set up along the lines of FLEXFly--a "paradraggable" puzzle piece (paradraggable = not actually draggable, but mimics prime aspects of draggable functionality after plopped).  Thus, within reasonable limits, it should be possible to create a larger variety in this manner for use with wider networks, though it may end up being a fair bit of work.  The other idea that's been thrown around is a Roundabout-equivalent to TuLEPs, which I've coined "RABETs" (Roundabout Extention Tiles).

-Alex

[Shadow Assassin]

Here's what I'm thinking: Automata impossibilities include double-decker networks of the same network; Cars'll jump between the two decks...

Technically, it isn't an impossibility.

It can be worked around. But it's kind of finicky to do.

[io_bg]

As we're speaking of roundabouts... What about some bigger ones, say 3 or 4 lanes wide and 6x6 tiles big? They would look more realistic with MAVE6 and RHW6 I think.

[samerton]

@ michi_cc: Wow! More great work!
@ jdenm8: nice! 

[michi_cc]

What did you use for the RA? Did you Photoshop a piece of the MAVE-4?

Yeah, the base are straight MAVE-4 and MAVE-6 tiles which I bended with Photoshop. The lanes are again parts of the straight pieces brought into form with the freeform grid. Some selective erasing and stamping for the finish. It's quite fast that way.

But as I said before, how would AVE-6 be connected (It's a three-tile network) and if we're gonna want an RA to connect to an AVE-6, we're gonna need a 5X5 RA...

I just remembered: There are plans for another asymmetrical network that Tarkus dubs as AVE-5. (Tark, in my opinion, that's a misnomer; Its more accurate name should be MAVE-5 or ARD-5). It's a combined MAVE-6 and MAVE-4; How would you connect that to an RA? (That's a stupid question; it's a half-and-half fix for that...)

Are there any graphics for AVE-6? Connecting a three-tile network would indeed mean a 5x5 RA, which brings some nice complications with two-tile networks as they can't be connected in the middle any more. (Drawing that isn't the big problem, but you'd have to mod a lot more combinations.)
AVE-5 (or whatever you want to call it) should be no problem indeed, just combine the MAVE-6 and MAVE-4 pieces (only the currently transparent median of the MAVE-4 pieces would need to be filled for a nice look.)

The other idea that's been thrown around is a Roundabout-equivalent to TuLEPs, which I've coined "RABETs" (Roundabout Extention Tiles).

As I can't mod that stuff, I couldn't care less how it would be modded As far as I can see, you have your fingers in most of the projects anyway

EDIT: Transitions from ARD-3 to MAVE-4 and AVE-4:

[Shadow Assassin]

As for making the textures with Photoshop, I tend to let the BAT do all the bending work for me. Much faster that way.

Anyway, for making 5x5 roundabouts... it wouldn't really matter if the 2-tile networks entered off-centre. I've had some ideas for making a 4x3 roundabout for avenues to interface with normal 2-lane roads, as well as "micro-roundabouts", which would be 1x1 and plopped on road or street intersections.

[samerton]

@ michi_cc: more great textures! Nice work! 

[ivo_su]

As we're speaking of roundabouts... What about some bigger ones, say 3 or 4 lanes wide and 6x6 tiles big? They would look more realistic with MAVE6 and RHW6 I think.

I also think it would be super
if someone hire to do it will carry the game into a new dimension
and I think we all are very satisfied

[Tarkus]

michi_cc, you've outdone yourself yet again!  Those look fantastic.

As we're speaking of roundabouts... What about some bigger ones, say 3 or 4 lanes wide and 6x6 tiles big? They would look more realistic with MAVE6 and RHW6 I think.

The long-term design plan for the RHW for some time now has had the RHW-4 being the largest network allowed to have at-grade surface crossings.  RHW networks with 6 or more lanes are intended to be strictly access-controlled networks, so they won't be getting any sort of roundabouts or intersections (aside from overpasses/underpasses).  Besides, that's what the OWR-4, OWR-5, MAVE-6, AVE-6, TLA-7, AVE-8 and TLA-9 are for.

Are there any graphics for AVE-6?

There are--and I've attached some below.

And since we're on the subject of AVE-6s and transitions . . . here's what's in place so far:



This one's actually got two modular transition pieces . . . a center tile one that allows for switching between triple-tile AVE networks (AVE-6/8) and triple-tile TLA networks (TLA-7/9) and one allowing for switching between 3 and 4 travel lanes in one direction.  It also shows all six possible triple-tile networks (TLA-7, 8 and 9; AVE-6, 7 and 8).



-Alex

[ivo_su]

I'm so disappointed to hear that there will be bigger and wider roundabouts
make me cry 
it is not fair
 
- Ivaylo

[GDO29Anagram]

TLA-9?!! AVE-8?!! iThought the absolute limit was TLA-7 and AVE-6!!!

So with the setup of TLA-7 and TLA-9 similar to that of RHW6c and 8c, that means it's possible to make... EVEN-NUMBERED TLAs!!! (I didn't pay attention to what you wrote, just the pictures... Heh heh...)

What's more is that I've seen an actual real-life setup for a TLA-4 in the city in which I live in.

I wanna organise stuff: There are three types of non-OWR networks: AVE, MAVE, and TLA. At this point now, I'd say that ARD would be a redundant name; It's still a type of MAVE. (Could we replace that name...?)

Let's see what network combinations are possible and theoretical (The boldfaced ones are non-existent ones):

AVE-2, 3, 4, 5, 6, 7, and 8.
MAVE-2, 3, 4, 5, 6, 7, and 8.
TLA-3, 4, 5, 6, 7, 8, and 9.

For AVE and MAVE, odd-number notation denotes asymmetry, whereas in TLA, even-number notation denotes asymmetry. How feasible is it to make these kinds of things?!! (If we run out of SPs, that's where problems are gonna arise... How many SPs can there be?!! I know there's 16 in the SAM, but how many are there for Road...? I know double and triple tile starters will greatly multiply the number of possible starter pieces, but I'm rather curious to know if we have enough starter pieces...)

TRAM integration would be another story...

The nightmare has ballooned out into a full-out hell: How in the world are we gonna organise EVERYTHING IN THE NWM?!!

Whoever said that the NWM is a second RHW was right... iPredict that by Version 3 or 4, its setup is exactly like that of RHW (In terms of how many menu buttons there are), but we're gonna need more breakdown in starter pieces... And transition pieces... (I'd propose organising starter pieces by AVE, MAVE, TLA, and OWR; The current method is a mess...) I'm confident that the NWM team will achieve this; This is still version 1.1.1 (or something), and there's still room for improvement...

I honestly didn't expect things to accelerate so fast, but maybe it's because I never witnessed the progress of RHW, so I don't have any first-hand experience... I feel I should contribute in any way I can... But if so, let it be by organisation at this point.

Here's the part where I comment on your pictures: For the AVE-4 to 6 transition, the median should be set so that it's like an actual lane (Imagine if it were TLA-5 to 7 instead), since it has inconsistent width throughout the actual transition.

Another thing: With how T21 exemplars are set, how would a median be set for AVE-6, AVE-7, and AVE-8? (That's also another story, I guess...)

[Ryan B.]

Speaking from experience, GDO - when things get rolling, they really get rolling!

[michi_cc]

TLA-9?!! AVE-8?!! iThought the absolute limit was TLA-7 and AVE-6!!!

Three tiles are enough for a MAVE-10 (Blue lines denote tile borders; not prettified)

[Tarkus]

I'm so disappointed to hear that there will be bigger and wider roundabouts
make me cry 
it is not fair

- Ivaylo

Actually, if you look over my post more closely, you'll notice I did not exclude the possibility of bigger and wider roundabouts.  Quite the contrary, in fact.  I merely excluded the possibility of the RHW-6S/6C connecting into them.  Interfacing would be accomplished by NWM networks and Maxis Roads, OWRs and AVEs.

TLA-9?!! AVE-8?!! iThought the absolute limit was TLA-7 and AVE-6!!!

Nope.  In fact, the TLA-9 has been on the drawing board since the original NWM private development discussions almost 4 years ago.  We've just never really mentioned it until now.  The TuLEP designs on the docket for the OWR-5 will also allow it to function similar to an quadruple-tile AVE-10.

I wanna organise stuff: There are three types of non-OWR networks: AVE, MAVE, and TLA. At this point now, I'd say that ARD would be a redundant name; It's still a type of MAVE. (Could we replace that name...?)

I'd agree the ARD-3 is basically a single-tile cousin of the MAVE, and classifying it as one does have a certain degree of logic to it.  However, how would one define this network (which is planned)?



We've already got a MAVE-4 . . . ARD-4 seems to make the most sense.

So with the setup of TLA-7 and TLA-9 similar to that of RHW6c and 8c, that means it's possible to make... EVEN-NUMBERED TLAs!!! (I didn't pay attention to what you wrote, just the pictures... Heh heh...)

What's more is that I've seen an actual real-life setup for a TLA-4 in the city in which I live in.

I'm aware of quite a few TLA-4s and AVE-3s here in Oregon, too.  In fact, those two networks are planned for inclusion after Version 2.  The preliminary design I have in mind involves a dual-tile setup along the lines of the as-of-yet unreleased Type C Road TuLEPs, where one of the lanes is off on its own tile, resulting in a true, functional asymmetrical capacity.  I still have to figure out how AVE-5s and TLA-6s will work, though.

For AVE and MAVE, odd-number notation denotes asymmetry, whereas in TLA, even-number notation denotes asymmetry. How feasible is it to make these kinds of things?!! (If we run out of SPs, that's where problems are gonna arise... How many SPs can there be?!! I know there's 16 in the SAM, but how many are there for Road...? I know double and triple tile starters will greatly multiply the number of possible starter pieces, but I'm rather curious to know if we have enough starter pieces...)

Actually, we won't really need much more in the way of starters, believe it or not.  Why?  Well, prepare for a mind-blowing revelation . . .  when you're placing a TLA-5 starter, the stubs you're actually placing are a TLA-3 starter next to something called a "NWM Road +2-Starter", which is essentially "dummy" network tile which adds 2 lanes to a Road-based single-tile network.  

A side-by-side override is then coded to transform the combination of TLA-3 next to NWM Road +2-Starter into a TLA-5 upon placement of the puzzle piece.  Similarly, the TLA-7 starter is actually a TLA-3 starter sandwiched between two NWM Road +2-Starters, and the MAVE-4 is an ARD-3 (in a specific rotation) next to an NWM Road +1-Starter.  

The RHW has actually operated on this same paradigm since the developmental builds of 2007, when it was being switched from the "side-by-side" method of Versions 1.2 and 1.3 over to a starter piece setup.

There are no dedicated dual-tile network starters in either mod--it's all modularization, re-use and encapsulation.

The nightmare has ballooned out into a full-out hell: How in the world are we gonna organise EVERYTHING IN THE NWM?!!

Whoever said that the NWM is a second RHW was right... iPredict that by Version 3 or 4, its setup is exactly like that of RHW (In terms of how many menu buttons there are), but we're gonna need more breakdown in starter pieces... And transition pieces... (I'd propose organising starter pieces by AVE, MAVE, TLA, and OWR; The current method is a mess...) I'm confident that the NWM team will achieve this; This is still version 1.1.1 (or something), and there's still room for improvement...

The RHW's kind of been building itself into an organizational crisis as well over time, which we think we've found (an at least a partial) solution for by going from a "by-type" to a "by-network" setup in one of the next releases.  I think the breakdown of AVE/MAVE/TLA/OWR you've suggested fits in with that concept quite nicely, and I think that'd be a front-runner for a new organizational paradigm once we get to that point.

Another thing: With how T21 exemplars are set, how would a median be set for AVE-6, AVE-7, and AVE-8? (That's also another story, I guess...)

The center tile is shared between all three textures . . . and since zone-activated T21s won't be an option (as the center tiles are "roadlocked") they'll have to be on all the time.

Three tiles are enough for a MAVE-10 (Blue lines denote tile borders; not prettified)

Very interesting . . . I hadn't really considered expanding the MAVEs beyond the MAVE-6 really, but I kinda like that.

-Alex

[GDO29Anagram]

AVE-10, Tark?!! Now you're REALLY pushing it... 

iDidn't think of radically asymmetrical roads like the one you've shown... In fact, I've never seen that kind or asymmetry... Ever.

Maybe ARD would be used to describe a radically asymmetrical MAVE...? After all, you called the hybrid MAVE-4/MAVE-6 a MAVE-5... (And I wanted to criticise you for not calling it an ARD-5 at first... MAVE-5 also makes sense...)

No, better: If the term "ARD" were to be abolished from all of the NWM, a proposal of classifying them as MAVEs would be as follows: For your example, A MAVE 1+3 (One lane on one side, and three lanes on the other, and the number of lanes would be in order of least to greatest.) No, that'd be too complicated... In fact, my first option of using ARD to describe radically asymmetrical MAVEs would make better sense... But we still gotta know how many lanes are on each side... (So it'd be ARD 1+3...?)

There are no dedicated dual-tile network starters in either mod--it's all modularization, re-use and encapsulation.

I've always known that there were rail-based constituents (IE, the starters for AVE-2, TLA-3 and ARD-3) for the two-tile starters, but an explanation of the OWR "pseudo-starters" also was in order...

So one tile is a true starter, the other's a "dummy". So how many "dummies" are there? Multiply that by the number of... Wait,... If the nMAVE and ARD-4 are gonna be released in the next version, and if they're both one-tile networks (iKnow the nMAVE is...), where are you getting the SPs for those?

That means,... There'd be FIVE single-tile starters (Four if the ARD 1+3 isn't single-tile), and "x" amount of dummies. iJust wanna know the number of dummies there is, so iCan figure out the number of... Permutations of starter pieces...

...and since zone-activated T21s won't be an option (as the center tiles are "roadlocked") they'll have to be on all the time.

iGuess that'd make sense, since it'd be uncommon (And rather stupid) to see something a massive AVE-6 or AVE-10 dumped in the middle of nowhere... Especially with a median like that of the AVE-4... Now this makes diagonalisation and intersecting even more complicated...

Someone mentioned MAVE-10. Here's my previous list of possible networks, now revised so that everything goes up to 10:

AVE-2, 3, 4, 5, 6, 7, 8, 9, and 10.
MAVE-2, 3, 4, 5, 6, 7, 8, 9, and 10.
ARD-4 (Or 1+3). I don't know what others would be proposed... Maybe a 2+4...?
TLA-3, 4, 5, 6, 7, 8, 9, 10, and 11. TLA-11 would REALLY be pushing things to the limit... But TLA numbering is offset by 1, on account of the "median"...
OWR-1, 2, 3, 4, and 5. No-one propose OWR-6, please.

There are twenty-seven types of possible puzzle-piece derived networks, assuming the asymmetrical ones only have one extra lane on one side, plus five for the OWR networks, plus one more for the ARD-4 you showed, and one more for the nMAVE. There are 38 different surface street networks, that includes all the default networks and Street. We've gone from Street, Road, Avenue, and One-Way Road to 14 different types of networks, that could later become 18 later on...

This is definitely another RHW (But for non-controlled access highways...), with the OWR compared to the MIS, and all the other OWRs compared to RHW-4, 6s, 8s, and 10s, and all the others being crazy variations of c-type RHWs (such as 6c).

EDIT: Crossed out some stuff...

[Shadow Assassin]

I know double and triple tile starters will greatly multiply the number of possible starter pieces, but I'm rather curious to know if we have enough starter pieces...)

We can combine series of single-tile starter pieces to get more use out of them. It's kind of like how the RHW works. So even though we might only have 16 starter pieces to work with, we still can have plenty more.


Oh, and...

...makes me think a centre-tile MAVE-6 would be not a bad thing at all. We already have a centre-tile MAVE-4 - that's the NMAVE-4.

It'd be handy for transitions from 6 to 8, anyway.