[Goku]

Genma Saotome, on 05 December 2016 - 05:29 PM, said:

IOW I'm not at all interested in new subtypes of the UiD definition unless KUJU's RE can be fooled into ignoring whatever it is you've added such that if, for any reason, somebody has to go back to RE in that tile, the route won't be damanged. If you can trick KUJU's RE, go for it.

KUJU's RE will ignore all additional values like MSTS does. If someone will want to go back to MSTS, KUJU's RE will remove all additional features.

Genma Saotome, on 05 December 2016 - 05:29 PM, said:

If you cannot, I recommend you back off on that feature until such time you've got an OR designed replacement for the .w file.

Bad news for you, it won't happen. I think only evolutionary development is reasonable. Development "make everything new from scratch" is very expensive and takes years to make something useful. If you want to wait for new OR data structure, new OR software, new OR editors, new everything - it's your choice.

[Genma Saotome]

It has been evident to me for quite some my preferred course was not going to happen but I did wonder if you were going to brighten up or not. Obviously not. So be it. I know what course I'll be taking now.

[vince]

Goku, on 05 December 2016 - 05:45 PM, said:

KUJU's RE will ignore all additional values like MSTS does. If someone will want to go back to MSTS, KUJU's RE will remove all additional features.,...........snip..........................

Hello Goku,

My 2 cents: As long as the output from TSRE5 is compatible with the existing diagnostic test routines it's ok by me.
If not, then new test procedures must be developed.

Editing a route is a process that has to be kept healthy with proper diagnostics.
It wont do to be informed of a problem by a route failing to load into either TSRE5 or Open Rails . . . like the old days of route editing in MSTS http://www.elvastower.com/forums/public/style_emoticons/default/curiousPC.gif

I've been having various problems using the standard Route Riter test ( Specifically the ICHK test. ) and am wondering what's at fault here; me, Route_Riter, TSRE5 or what . . . !

I cannot edit the Long Island Rail Road with the MSTS RE at all. Crashes on save, many tiles.
TSRE5 is my only hope to edit and continue building the route but the output must pass existing tests or new tests be developed.

regards,
vince

[eric from trainsim]

Goku, on 05 December 2016 - 03:30 PM, said:

RouteRiter, TSUtils, and WFH will work fine, if you not renumber the UiDs of existing items.
There is no possible solution for out of sync if UiDs change using third party software. You have to choose - using new features or very old software.

Reordering UiD's is more or less essential when fixing corrupted databases, and when merging routes. Simply saying "don't do it" isn't an acceptable answer.

[eric from trainsim]

Goku, on 05 December 2016 - 04:07 PM, said:

The old stuff is 99% completely adressed. Remaining "stuff" will be fixed during usage of this editor. It would be stupid to wait for bug reports and do nothing else meanwhile. I think it is better to work on new features.

I really don't buy the statement that you've addressed 99% of the features needed for a route editor. I'd guess that it is probably 80%, but will give you some credit for the new features you've brought in on map display and DEM imports... So let's say perhaps 90% of what MSTS does. Now, you're at the point of dealing with the remaining 10% that isn't quite so easy to do.

As a multi-decade IT person, closing the gap on that 10% will take you as long if not longer than it did for you to get thru the first 90%.

Try to resist ignoring the complex issues in favor of adding more new features, or pursuing an easier approach just because the other approaches that take into consideration some of the issues raised here are too hard...

[Goku]

eolesen, on 06 December 2016 - 08:47 AM, said:

Reordering UiD's is more or less essential when fixing corrupted databases, and when merging routes. Simply saying "don't do it" isn't an acceptable answer.

But really there is no solution. Even if you create new world file format like Genma Saotome wants - all old software will stop working. And not only UiD reordering. Everything!
Only solution is to ask developers to fix their software. I think JKY will make updated version of RR for James Ross second .w file solution.

Genma Saotome, on 05 December 2016 - 09:48 PM, said:

It has been evident to me for quite some my preferred course was not going to happen but I did wonder if you were going to brighten up or not. Obviously not. So be it. I know what course I'll be taking now.

Sorry to disappoint you, but really, make totally new tools for OR is too expensive in time for me.

eolesen, on 06 December 2016 - 08:59 AM, said:

I really don't buy the statement that you've addressed 99% of the features needed for a route editor.

So, tell me what is missing that is in old RE? Only thing I know is support for gantry file, but for me it was most stupid and annoying feature of old RE. I am working on my own, better replacement (first version already included in TSRE).

[eric from trainsim]

Goku, on 06 December 2016 - 09:22 AM, said:

Only solution is to ask developers to fix their software. I think JKY will make updated version of RR for James Ross second .w file solution.

JKY doesn't support TSUtils -- that's a different developer package that RR calls.

Goku, on 06 December 2016 - 09:22 AM, said:

So, tell me what is missing that is in old RE? Only thing I know is support for gantry file, but for me it was most stupid and annoying feature of old RE. I am working on my own, better replacement (first version already included in TSRE).

The last I heard, complex signalling was still resulting in corrupted TDB's.

In addition to gantry (which I agree is a useless function), do you now handle
* deer and hiker hazards
* collision objects
* sound regions
* RGE quadtree and tile generation (that's part of the route editor, even if it is in a separate option from the dashboard)

I'd check myself, but I can't run the executable under 32b Win7. Not all of us run in 64b...

[Goku]

eolesen, on 06 December 2016 - 10:12 AM, said:

The last I heard, complex signalling was still resulting in corrupted TDB's.

I hope it works fine now.

eolesen, on 06 December 2016 - 10:12 AM, said:

* deer and hiker hazards

Yes.

eolesen, on 06 December 2016 - 10:12 AM, said:

collision objects

Missing properties for this.

eolesen, on 06 December 2016 - 10:12 AM, said:

sound regions

Placing sound region on one track works.

eolesen, on 06 December 2016 - 10:12 AM, said:

RGE quadtree and tile generation

Works OK. In editor quadtree and tile generation and terrain from HGT generation.

eolesen, on 06 December 2016 - 10:12 AM, said:

I can't run the executable under 32b Win7. Not all of us run in 64b...

32 bit builds are included in ZIP file. Not always up to date, but one per month or two.

[Genma Saotome]

Goku, on 06 December 2016 - 09:22 AM, said:

But really there is no solution. Even if you create new world file format like Genma Saotome wants

Yes there is, you just don't want to do anything you don't feel like doing.

As I said before you can create any number of static objects that act as a placeholder for enhanced functionality:

Static (
UiD ( 2212 )
FileName ( Enhanced_Car_Spawner.s )
StaticFlags ( 00040000 )
Position ( 316.9 26.01 -31.1381 )
QDirection ( 0 0 0 1 )
VDbId ( 4294967293 )
StaticDetailLevel ( 5 )
)
where the string "Enhanced_Car_Spawner.s" tells the OR software to regard this as an enhanced inter-active.

You can do the same thing for any enhanced item. At such time the OR team chooses to abandon .w file compatibility the change from static entry to true interactive entry will be trivial because the actual enhanced interactive code has already been written.

So the problem here is not technical, not is it an issue of evolutionary vs. complete design revision. The problem is you don't feel like doing this. The reality is that is just the way things go with volunteers... it's your time and your code and everyone who doesn't like your decisions can go suck lemons.

[Goku]

No. This is not a solution. Beside it is extremaly ugly, it will not work.
Where in your solution is connection to existing carspawner and where custom car spawner name is defined?

If you want to extend existing item, you have to connect the objects by UiD. If you don't want existing item extension, James Ross solution is also better. Just start from 99999 UiD and you never make a collision between main .w file and second .w file. Even if you renumber UiDs using third party software.

There is no better solution than James Ross proposal.