[Genma Saotome]

Jeff,

I understand you've been given a copy of the Route Riter source code, w/ permission to enhance it... is that right?

I can think of one thing right off that would be good to add: .Engs and .wags now have a command word "Include()" that is a pointer to another text file (type .inc). Any test for file completeness that doesn't deal with these Include files will fail. I can provide some files for testing if needed.

Something old: RR never tested the logical integrity of water layer levels across a range of tiles -- are all four corners equal where 4 tiles touch?
Something new: Test for untextured polys.

[Jeffrey Kraus-Yao]

Erased.

[Genma Saotome]

The include file is unique to OR and yeah, fixing the SU exporter is proper; I've been asking for help in that regard for years but nobody takes up the task. FWIW Ruby is pretty interesting... it was inspired by Smalltalk so it's quite OO.

A list of shapes that are not used would also be useful but only at the very end when you're cleaning things up. That may already be there.

Something else (I hesitate to mention it): it seems that RE allocates memory from record 0 to the largest value UiD() of the 4 tiles currently displayed -- in full, w/o regard to the fact that many of those UiD() values are not actually in use. This can lead to RE aborting while trying to save to a world file. I've found that renumbering the UiD() values to a tight range -- no gaps in the list of numbers used) both fixes that problem and leaves RE more well behaved / faster/smoother than before. Naturally renumbering UiD's is fraught with risk... leaving to this suggestion: AFAIK the UiD() values for static objects can be safely changed. Taking those objects are renumbering them from the known gaps can drop the max UiD() value by several thousand. It's something you might consider addessing.

A second suggestion, also w/ some hesitation: I assign a StaticDetailLevel() value to every object I use. This leads to the sorting of entries in the world file. Anything from the \global\shapes directory gets zero. Track bridges, etc. get 1, Berms 2, Buildings 3, vegetation and all other items are 4-9 with 10 reserved for tools. Aside from the utility of making it a bit easier to work with having all of the world file entries located in one continuous block and everything else in another means only 2 system calls are needed to change the read directory whereas w/o it leads to hundreds, perhaps thousands of such calls per tile. I would expect similar benefit were identical static shapes sorted into sequential order. It's a small thing when done... a sure amount of complexity while doing, but it might make a noticeable difference in the tile loading phase of train.exe.

[R H Steele]

This is an extremely interesting prospect. I've been following this thread over at Train Sim. I have found RouteRiter to be an exremely valuable tool and extending it's usefulness to ORTS and also adding functions is a very good idea.
:rotfl: to Mike for releasing the code to Jeffrey and much thanks to Jeffrey for taking on the task.

[Jeffrey Kraus-Yao]

Erased.

[conductorchris]

Route-Riter is a great tool and I am grateful to you Jeffrey for taking it on to keep it integrated with Open Rails. And grateful that Mike Simpson has given that possibility life through his donation of his code.

I wonder if it is worth some rethinking of it's format for the future? For example, many of the functions to help route-builders might profitably be integrated into a future route builder. The consist builder might go well with a future activity builder.

Likewise, some functions might be better tucked away a bit for beginners to find later. Or appropriate warnings and explanations added.

I also wonder if there is a possibility of working with Mike Warren and merging the consist builder with YardMaster making them part of open rails.

Christopher

[Jeffrey Kraus-Yao]

Erased.

[copperpen]

Jeffrey Kraus-Yao, on 18 January 2015 - 01:15 PM, said:

One integration I have in mind for Open Rails is if an .eng or .wag file has an Include statement but is missing elements required for Train Simulator that these required elements will be copied from the included file.

Not sure what you actually mean here. An include statement in an OR style eng or wag file is there so that many can utilise the include, with only one file to edit to make a change to the include. So brakes and couplings to mention just two items, will be absent from an OR eng/wag file. To then copy these from the include into the eng/wag will defeat the object of using includes.

[Jeffrey Kraus-Yao]

Erased.

[conductorchris]

Jeffrey, I am really happy to hear what you are writing.

An advantage of folding this into open rails is that it doesn't depend so much on one person. When one person burns out or moves on, it is there for others (like yourself) to pick up. A similar advantage is that different people have different talents. For example, many open source projects have usability groups that improve what the main coders have done.

Christopher