Locomotive fixes for 1.06

[Gumboots]

I assume that was stopping you getting an export script were the bytes for what WP&P called the "main direction". These come after the UV mapping in the .3dp and have no obvious way of being generated by Blender, since their values in default files often make no sense compared to the face's orientation.

I looked into those and found that, as far as I could tell, changing the values there had no effect of graphics while playing. So I took the easy option and ignored them. The export script just sets those bytes to zero and it seems to be fine.

I suspect they are deprecated code that is left over from alpha. We have found quite a bit of that. For instance, the weights in the .cgo files are not used by the game (it only uses the .car file weights) and the base price for passengers in the .cty file has no effect on prices during a game (the price must be hard-coded into the .exe).

I found Blender's learning curve was fairly easy, once I took a few hours to actually read the relevant parts of the manual and go through some of their basic tutorials. Before that it was impossible. The manual is essential to making any progress, but you don't need to read all of it. A lot of it is to do with Blender's game engine and other stuff we don't need.

But yes, lack of time is a problem. So is having too many ideas and wanting to implement them all at once.

I don't worry about the limit on the number of locos. I figure people should just make whatever they feel like making. If there end up being too many locos you can just remove the .lco files for a few of them when you want to use others. Easy.

[bombardiere]

I have been reading the posts and I have to say you have done a huge work. My warmest thank yous.

I still have a lot to read, but I installed the Blender and tested your script. Two minutes and I did have a 3D Duke Loco in the Blender display. I have not yet tested the skins, as I am bit lost with the texture UV mapping in general, but I will learn if I have will.

I have probably a million questions, but I need to do more reading.

I really like the way you transformed the P8 into a tank engine. Very clever, and I might take the same route. For a learning project, I am thinking to transform 4-4-0 Duke into Highland Railway 4-6-0 Jones Goods. It is similar and I would need to make longer and fatter boiler.

By the way, I think I read in a post that a member "Tomix" said that he ripped Sid Meier's Railroads locos and skins. Are these in any use and have these converted in RRT3? There are some nice locomotives, which I would like to see in RRT3 rooster.

But yes, lack of time is a problem. So is having too many ideas and wanting to implement them all at once.

Tell me about. I am always full of ideas. Often too many to implement all. I see that you have had a lot of patience with the hexfile. A lot more than I did. I did some experiment with X Y Z cordinates, and in this forum should be a hilarious picture of a main wheel doing large circles around the loco, but I never was able to set it right. Perhaps I made a mistake thinking that the cordinates were measure units, such as meters, rather than thinking those as pixels.

[Gumboots]

I had thought a Big Goods would be a nice addition at some stage. I have scale drawings for it and the Duke, along with some clear detail photos of the Big Goods. If you want copies, let me know.

I don't know about the Railroads skins. You'd have to ask Tomix about those.

The XYZ co-ordinates are scaled so that 1 unit = 10" (or 250mm, near enough). That problem with the wheel would have been due to the attachment point* being wrong. I've done the same thing myself several times, but it's easy to fix once you know how the .3dp works.

*The point the wheel spins around.

[bombardiere]

Ok, getting the blender and scripts etc set up was easier than getting the RRT3 to run in a modern computer. But thanks to the fixes in this site, I was able to get the game to 1.05 standards. Now I am going to play the campaign first to see if the game still has the charm.

For overall game play purposes the Jones Goods is not very useful engine. I think it would be very similar to P8, or even inferior. But it could be a nice start. Sure, if you have drawings those would be useful, but I think I can do most of the job by estimation. I have some drawings from old Railway Modeller magazine, such as Stanier's 5 and 2-8-0 8F and Terrier and back then I dreamt for a full British Railway Tycoon.

Which Hexeditor would you suggest? I haven't used a hexedit since I left RRT3.

[Gumboots]

I just love having drawings because it saves me having to guess. If I use the actual dimensions it will look right, and it's faster than trying stuff to see if it works, then having to change it when it looks wrong. The more drawings I can get, and the more accurate they are, the happier I am.

The drawing I have for the Duke is pretty basic, but is probably good for proportions. The one for the Jones has actual dimensions added in several places, which makes it a lot more useful. TBH the two locos are not all that similar. The cab and steam dome and handrail are almost identical, but that's about all. You'd have to trash most of the framing for the Duke to get a Jones, and the cylinder height and angle needs changing too.

I use a paid version of HexEditorNeo, purely because before the Blender script was available I wanted an editor that had awesome bookmarking ability. AFAICT the best free one seemed to be HexEdit.

[Gumboots]

Hey about this:

So this morning I was thinking about possibly modelling some of the very early Australian locomotives. It turns out that the first one ever still exists, which I didn't know (https://en.wikipedia.org/wiki/Locomotive_No._1). Dimensions and good quality, modern, detail shots are readily available, and I can even go and see the real thing next time I'm in Sydney. This is all relevant because it was built in the same period, and by the same engineers, as the original DX Goods. It's almost identical in many ways, since it's really just a 0-4-2 variant of the DX, and will be useful for modelling both locomotives. Once this one is done, slight changes will give a very accurate DX Goods with very little extra effort.

It's amazing what I get up to when I should be doing other things.

So anyway, and don't ask me why, I ended up playing around with the model for NSWGR Number 1. Then I thought well hey, why not do those tricks to get an easy DX Goods out of it too? I have enough basic info (driver diameters, wheelbase, etc) and can wing it on the trivia as long as I don't get too pedantic (stopping myself getting too pedantic is the tricky bit).

Messed around with it some, and here's the result.

Second shot is an accurate model of an original, done by some bloke in the UK, just for comparison.

[Gumboots]

Have done a bit more work on the DX Goods model. I decided to model it in the later configuration, with basic cab, etc and a plain black livery, since this is how they spent most of their life. It still has some errors in it and needs a bit more detail, and obviously needs skinning/graphics, but it's getting close to something workable. Verts+polys count should end up being pretty close to the default H10 Mikado, so it should run without causing frame rate problems.

[Gumboots]

Bomber's idea of a Jones Big Goods got me curious, since I had thought of doing that one myself at one stage (but have done nothing about it of course ).

Anyway, it turns out to be very similar in size and shape to the NSWGR C32 class (another 4-6-0). Which I suppose is not surprising since they were both designed and built within a couple of years of each other.

The Big Goods has about 10% more nominal grunt than the original non-superheated version of the C32 (classed as P6 back then) and around 10% less than the upgraded superheated version (the C32 per se). But with superheated just being plain more efficient, actual grunt for the C32 would be more than +10% over the Big Goods. I'm inclined to think that the C32 would be a more useful addition to the game.

OTOH, the Big Goods is incredibly clean and simple (hardly any thingummys at all), and should be easier to model and skin. Just for a starting point in case anyone ever wants to take it further, I quickly changed a copy of my C32 base model into a Big Goods. It's accurate as far as it goes, but obviously doesn't go very far at the moment. Nevertheless, it's a start if anyone wants it.

[Gumboots]

Had an idea. The 1.06 Ten Wheeler is another one that's a bit of a mess and could do with fixing. It's a bit of an odd one too. 4-6-0's were around before Moguls, and were used for freight because they were an obvious extension (more adhesion) of the standard 4-4-0 in widespread use in the 1850's. So although in the late 19th century 4-6-0's of a different style were used as pure express locos, before 1870 they were less extreme and more of an all-rounder. Freight, mixed traffic, express on heavy grades. Stuff like that.

Anyway, the US loco roster has the Mogul and Connie available for freight starting in 1865 and 1875, but there's no upgrade for US express until the Ten Wheeler and Eight Wheeler both come in at the same time in the early 1890's. Which seems daft. Lirio thought it was daft too, which is why her stats revision pack puts the Ten Wheeler's start date back to 1875.

That may still be a bit too late though, since the American 4-4-0 is pretty weak and not that good for express. However (this is the good bit) the American's model can, with a bit of tweaking (like I did to the Connie to make the new Mogul) be fairly easily made into a pretty close approximation of B&O's Class B Thatcher Perkins. This locomotive was in mainline service from 1863 to 1892.

Which looks a lot better than the American anyway. Even before it gets re-skinned. So perhaps introducing a beastie like this, along with the Mogul in 1865, would be a good use for 1.06's Ten Wheeler. It'd give a good, but not over the top, express upgrade to take things through to the Eight Wheeler in the early 1890's.

[RulerofRails]

I like this idea. Our loco selection is generally more limited in the 19th century so fixing these little choofers helps gameplay.

[Gumboots]

Yup, that's what I was thinking. Fills a hole in the roster nicely.

I was surprised how few tweaks the default American model needed to stop it looking putrid. Just a few adjustments to proportion and balance vastly improve it, and a re-skin will obviously help even more.

That then got me looking around at early 4-4-0's as well. Some of them were a lot more visually appealing than the beast that PopTop foisted upon us. The William Mason (good pix here and here) is something I'd personally find a lot easier to live with than the default. Making a replacement model and skin is certainly possible.

The catch here is that all the default beast's files are hidden in 3D\rt3_DDSF.PK4 and 3D\rt3_3DPF.PK4 where they can't be got at, and I don't like the idea of using hundreds of loose override files for a stack of entire locomotives. What I think I'll experiment with is breaking down those huge PK4's and seeing what we can get away with. It may be that we can leave the 3D directory empty and just use individual PK4's in PopTopExtraContent, or alternatively just put individual PK4's in 3D, or some compromise like a placeholder PK4 in 3D, etc.

It'll depend on how brutally and evilly the devs locked things down. Hopefully they overlooked this bit, like Sauron missing a couple of stray hobbits, and the forces of good will prevail.

[Gumboots]

Had a bit more of a go at it, after checking dimensions for the original. This is getting pretty close. The cab and steam dome still need work, along with general graphics. I haven't touched the tender yet, but the default tender is roughly right for scale and style, which is handy. I think this project is a goer.

[Gumboots]

Mongrels! Scum! Bounders, cads, and misbegotten descendants of sewer rats!

So I was just flipping through a new addiction, namely Internet Archive online issues of "Die Lokomotive". This (obviously German) magazine was apparently published from sometime in the early 1900's (1905 is the earliest I've found so far) up until the start of the The Great Debacle and General Silliness in 1939, which appears to have permanently knackered said magazine.

Which is a pity because it was absolutely fascinating. In any issue you might find articles on anything from narrow gauge mini-choofers from the Congo, to an Erie triplex or a prototype Pennsylvania duplex (forerunner of the famous T1). In many cases the drawings and photographs provided are so good that you could pretty much build the entire locomotive for real.

So why is I grumpy now? Because in their 1911 issue the rotten sods included an article about the 1.06 loco we all love to hate: the Vittorio Emanuel II 4-6-0 aka "Bog Awful Dunny Shed on Wheels". With drawings, and specifications. This means there is no longer any excuse to not fix the thing.

On the bright side, the same issue has an article on the Württemberg Class C, which was a totally cool little Pacific that I've been wanting information on.

[Gumboots]

I had another look at the "Thatcher Perkins" version of the 1.06 4-6-0. I had done some more work on it, months ago, after posting the last screenshot. That left it looking like this...

...which is a lot better than it did before. The new mesh for the domes was nicked from my version of the Mogul, and there have been a few other tweaks as well. In fact, come to think of it, it's not far off being a usable finished product. Which means I should probably finish it.

[Gumboots]

Had a look at the loco roster, and split it down into categories. This is just to get my head around what's available at the moment, and what gaps (if any) could usefully be filled. Some categories are covered as far as "stuff they actually need", even if they will never be fully covered in terms of "stuff that anybody might want for Christmas to go with their new pony and can we have fries with that kthnx".

Broke it down into Euro, US and World and each of those into Express, Mixed and Freight. The World category is going to need some stand-ins. For instance Mikados were common worldwide but they weren't like the H10 as such. And in the early years Samson can stand in for freighters everywhere, because they were all much the same. And some locos really fit in two categories. Etc. But still...

1/ Euro express is pretty well covered. Could add more for giggles, but don't really need them.
2/ Euro mixed and freight are a bit patchy, especially since many of the current "assets" are so rough that a lot of people refuse to use them. But even if that gets fixed those categories could still use a few more locos here and there.

3/ US express is pretty well covered.
4/ US mixed is a bit patchy.
4/ US freight is pretty well covered.

5/ The World category, as such, doesn't really exist yet and will have to use stand-ins for a while.

This is all for steamers. I haven't checked the situation with diesels and electrics yet. Spreadsheet attached, in OO and Excel formats.

[Gumboots]

So it looks like I might be able to stand fixing the Vittorio "Bog-ugly Dunny Shed on Wheels" Emanuele beastie. Last night I wasn't up for any serious work, and just wanted something to mess around with until I felt like sleep, so hauled out a basic Blender file I had started yonks ago for this thing. Started throwing mesh at it, and it started looking more presentable. It's still one of the ugliest things ever, but at least it now looks like a locomotive (of sorts).
.

.
Throwing mesh at a model like this is pretty quick and easy. I'm working from two original works drawings, which are slightly different of course , and the one and only available photo of the actual locomotive, which is slightly different to both of the works drawings of course , so that means deciding what to take from which source and how to meld it all together into something useful and coherent. It's now reached the point where taking it further means getting serious about optimisation, mapping, detailing, and all the rest of it. That's the stuff that takes most of the time, so I'll put this on the back burner while I sort out Samson and the DX Goods.

But eventually we should end up with a VE mongrel of a thing that we can actually enjoy using. Which would certainly be a novel experience for anyone who has played 1.06.

[Gumboots]

RoR was saying via PM that most of the 1.06 electric and diesel locomotives have the glow in the dark bug, and asked if it was easy to fix.

For anyone who uses those locomotives, and hates having them glow in the dark: yes, it is very easy to fix. All it requires is using selections and the threshold tool on the alpha channel. This only takes a minute for each image, in either Photoshop or GIMP. I've taken a quick look at the Paint.net site and it looks like that app will do the same job, although I have never used it myself. So that's three apps, two of them free, that will handle it without drama.

However I never use electric and diesel locomotives, so I don't care. This means I aint gonna do them. I have enough on my plate already. If anyone wants them fixed, you'll either have to do them yourself or talk someone else into it.

[RulerofRails]

Thanks Gumboots! I have GIMP, as time permits I will try to figure out what to do.

Also,
Congrats on 3000 posts!

[Gumboots]

Short version:

Select alpha channel (Turn off RGB channels temporarily just to save confusion.)

Select all to get the whole image.

Use the "remove from selection" thingy to take out any chunks where you want to keep non-binary alpha (In GIMP: hold down Ctrl, while drawing a second selection, to cut bits out of the first selection.)

Anyway, you should now have a selection on your alpha channel that includes all the bits you don't want glowing in the dark. In PS go Image > Adjustments > Threshold. In GIMP it's Layer > Transparency > Threshold Alpha. It's set to 128 by default and that will be fine.

Hit Enter.

Save as DXT3 DDS. Note: If there's an option to export mip-maps, make sure that is not selected. RT3 doesn't like mips.

Go play trains.

[RulerofRails]

Thanks a lot! Instructions were spot on. Makes it painless.