Kujos to avoid

LAST EDITED ON Jun-14-04 AT 01:21 AM (EDT)

I do route development moreso than simming... and a with a couple decades in software analysis and design I feel qualified to comment on some project errors committed by the original development team. In the route development area there are a few steaming heaps of kujo that could be avoided with a bit of up front planning so a few suggestions with that in mind:* Put the all the texture files in the same directory... IOW don't use \terrtex and \textures. A texture is a texture is a texture.* Let the top texture directory be the default and use 12
subdirectories -- 4 seasons x 3 weather types. This way the developer has full flexibility for creating ground and/or shape textures for each season as well as each weather type -- or ignore it all and just use the defaults. Or some combination.* Get the water texture filename out of the .env file and put it into the texture directories, just like any other texture. Let's the designer alter water appearance by season and/or weather. Ditto for sky. * Let the designer specify water textures exactly the same as any terrtex. This is a simple way to offer multiple water types, like lakes vs. rivers.* Let the designer enter numerals via the keyboard if desired. Just having the roller buttons isn't enough... editing qdirection in the .w files isn't enough. RE is essentially a simple cad program so let the user have access to the numbers as a simple way to gain precision.* Keep the .w files data open. There are times when you simply need to get in there and mess with the data. It's normal for me to edit the .w files whenever I run RE. If you move them from text into a database (something that should greatly improve stability and quality once you do "Begin Transaction" and "Commit"), please use something that allows the data to remain accessible via import and export functions.* Use a visible 3d tool for the land deformation (instead of drawing a rectangle), such as a circle, oval, triangle, and/or wedge and allow the user to size and rotate them numerically in a side window -- the actual deformation logic on the tiles should remain the same.* There really shouldn't be a requirement to order development in the way it is now...first all track then... logically development should be done completely as you move along the route. The problem is, I beleive, in the poor way Kujo manages adds and deletes across the .w, .tdb, and .rdb files. Again, moving that into a database with Begin Txn and Commit functions should provide the controls that are needed to eliminate the kinds of corruptions what have ruined many routes.* Talk somebody into doing a node viewer and integrate it into RE. Allow copy and paste (with full edit capability for position and qdirection values, absolute or relative).I'm sure by now eyes are rolling. Fine. But read the list again... everything there is meant to improve quality and reduce development time. What good are new features if the route takes 3 years to create and always crashes?Dave NelsonWP mainline, Oakland-Stockton: 100% complete;
SP mainline, Oakland-Stockton: 20% complete;
Aug 1950
<<new .sig under construction>>

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

RE: Kudus to avoid, OTTODAD, Jun-14-04, 08:29 AM, (1) RE: Kudus to avoid, muskokaandtahoe, Jun-15-04, 12:21 PM, (2) RE: Kudus to avoid, Vern, Jun-16-04, 11:13 AM, (3) RE: Kudus to avoid, RailDriver, Jun-16-04, 07:30 PM, (4) RE: Kudus to avoid, 7andY, Jun-17-04, 07:40 AM, (5) RE: Kudus to avoid, landnrailroader, Jun-18-04, 10:59 AM, (6) RE: Kudus to avoid, Vern, Jun-18-04, 12:13 PM, (7) RE: Kudus to avoid, brucek, Jun-18-04, 05:22 PM, (8) RE: Kudus to avoid, muskokaandtahoe, Jun-18-04, 06:05 PM, (9) RE: Kudus to avoid, brucek, Jun-19-04, 01:21 AM, (10) RE: Kudus to avoid, muskokaandtahoe, Jun-19-04, 03:47 PM, (13) RE: Kudus to avoid, landnrailroader, Jun-19-04, 05:23 AM, (11) RE: Kudus to avoid, OTTODAD, Jun-19-04, 06:52 AM, (12)

Conferences | Topics | Previous Topic | Next Topic

Hi Dave !No, you have not said TOO MUCH, NOT ENOUGH is more likely ! There is need for the entire Train Sim Files to be being structured and organised into Folders in such a way as to make their use super-efficient.There should be a SHARED GLOBAL Data Registry for items needed by ALL Routes and not like at present each route having to have a copy of them !This can be done easily by using Index Key based B-TREE Databases or SQLs like the one from PERVASIVE, the follow-on from BTRIEVE and not Microsoft's.The option of displaying Objects in SHAPE FILE VIEWER fashion should be standard prior to selecting them in the RE as well as when assembling Consists in the AE, working like the ConBuilder ! Complex repetitive Tracks assemblies like Double-Slips forinstance and multiple objects should be able to be COPYied / PASTEed elsewhere by simply dragging a box around them and repeating them elsewhere.Forest and other areas to be filled should give a choice of various types and densities, to be placed into non-square areas defined by dragging a line from anchor point to anchor points, curving it as required.Having been programming Accounting and Business Software for 20 years now and have used various Cad-Cam and other Graphics Design programs in a limited way, I am aware of the need for DATABASE efficiency in Business Software and all the various design options available for creating shapes in Graphics programming.What is needed is an efficient basic infrastructure for TMTS, not bits and pieces bolted together and an equally efficient way of running a single Route without having to load a lot of other stuff it does not need ! Let's see what PIE-TMTS will bring !I have no doubt that the Performance and Quality of the Graphics will be compatible with the best in the Gaming and Simming Software Sectors. Got a feeling there might even be a few "Capricorns" or other Efficiency Experts in their team too ! Take Care, O t t o.

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

>There should be a SHARED GLOBAL Data Registry for items
>needed by ALL Routes and not like at present each route
>having to have a copy of them ! Very helpful for freeware, not really helpful for payware.
>The option of displaying Objects in SHAPE FILE VIEWER
>fashion should be standard prior to selecting them in the
>RE as well as when assembling Consists in the
>AE, working like the ConBuilder ! I dunno. Could be handy. Monitor size is a factor there. What would bring a smile to me tho is changing the size of that small placement window... the list of items should be more than 6 and the window should be wider so lengthy names are displayed in full. A stretchable window would be best. Drives me nuts when I'm using about 7 or 8 track shapes and am always going back to the master list to get the one I need.>Complex repetitive Tracks assemblies like Double-Slips
>forinstance and multiple objects should be able to be
>COPYied / PASTEed elsewhere by simply dragging a box around
>them and repeating them elsewhere. Yeah... RE is a butchered cad program. It could/should be extended as I desribed above to let the designer specify all position and qdirection values via the keyboard as well as just plunk 'em down with the mouse as is done now. Once you start thinking CAD you think of location functions like snap-to nearest point, rotation equal other object you click on, rotation perpendicular to other object you click on, etc. which are really very simple functions of reading the position and qdirection of the other object and using those values on the new one. An ideal state of affairs would be copy and paste across routes!>Forest and other areas to be filled should give a choice of
>various types and densities, to be placed into non-square
>areas defined by dragging a line from anchor point to anchor
>points, curving it as required. Yeah, I forgot about that. Non-Square... same issue as land deformation: a 4 sided mouse drag isn't enough... back to Cad again, what's being asked for is some combination of a basic polygon function (circle, rectangle, etc.) or a freeform polygon formed by multiple points (essentially a complex line forming a closed polygon). The UI for both land deformation and forest propogation would be similar, as would the use of the polygons and specified points, leaving the final creation of the data unique for land vs. forest. Just think: a few mouse clicks along a track curve, move away a bit, and raise the land 2m (or do a forest) vs. the present way of pull a rectangle, raise the land, lower it again where the track is, and edit a seemingly infinite numer of individual vertexes where it's still wrong... and wonder about the worth of spending so much time at this stuff.Dave NelsonWP mainline, Oakland-Stockton: 100% complete;
SP mainline, Oakland-Stockton: 20% complete;
Aug 1950
<<new .sig under construction>>

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

Equally you don't want to over engineer the plumbing...What's needed is a Route Editor that is simple and easy to use - more people creating routes means more choice for end users and more copies of the programme sold. Quite a few basic functions in the MSTS RE were incomprehensible until you learned them, like the use of angles for gradients instead of perecentages or 1 in xx and radians for degrees of curvature on dynamic track. The people who understand the structure of a real railway line - track layout, gradient profile etc. aren't necessarily design artists or PhD's and they need a programme which allows them to interpret that information into a viable route as painlessly as possible. Don't think I'm suggesting that it shouldn't require some thought or time and care to get good results, but ease of use for a non scientific (for want of a better word) end user must be paramount. I'm not the world's biggest Trainz fan but something akin to Surveyor (apart from the difficulty of getting real world data in there) is what's needed.

Chief Track Welder

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

Vern,I agree, I DO have a PhD and all my track charts are still in % grade. Kuju was never contracted to build a route editor. The tool was simply a tool for internal use and they released it for general public use. As with all internal tools you make the best you can to get by. They them selves are not revenue producing so there is no need to expend the tremendous effort needed to produce a consumer ready product. TrainZ makes a good product, the Surveyor is pretty easy and fun to use. Our editor will probably be a careful comprise between ease of use and complete control for real route creation. TrainMaster is dedicated to accurate simulation including route creation. Our tools will be aimed at the production of accurate, real life routes in a quick and efficient manner. Where possible we will try to offer several methods to input data. We are fully aware of the need to support the developers. We are these developers for our professional work. But for TMTS we must also remember that the actual sim and consumer out of the box experience is paramount to bringing in new people into the hobby and continuing its long term stability. We are dedicated to providing a top quality, easy to use sim. The same is true for the editors but the final user friendly versions of those may have to take a second place to the sim itself. Unlike Microsoft, we do intend to support the tools because we do recognize that it is the 3rd party developers who keep this hobby alive. Michael

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

Please, please, please don't make the Route Editor too easy to use. If you do, we'll end up with lots of awful routes, with sub 100 metre radius curves, points (switches) that turn on a sixpence, grade changes that would be shameful even on Britains' roads etc - just look at a whole range of offerings from that other Sim.Just my two pence worth...Cheers, 7&Y

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

7andY,I would agree with you, except for this. Freeware is controlled by many of the forces that control Payware. If a handful of users see an exciting route (they think), download it, and find it is substandard, we will all hear about it, and Nels will find very little traffic for it's downloads, such that soon, the author may even ask to have it removed from the library. Small radius curves might work fine, if the sim. was to be of a Model Railroad, but the real world has it's limits, except for trolley & subway routes. Anything greater than 15 degrees, or 3% would be rare on a real US railroad. Raton is, I believe 3.5% and when you ride the SWC over it, especially westbound, you know it is steep. I would guess the curves go to 10 degrees on it.
Point I make is, only reasonable routes would make the grade as far as long term downloading.
I am looking forward to the new PI simulator, and I would love to be doing a route for it, even now. Hopefully some of the best routes that have been done, will be tapped for use as out-of-the-box sims for it.J. H. Sullivan, "landnrailroader"

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

Easy does not necessarily equal poor quality routes. On the contrary it means people with a knowledge of the prototype but not much in the way of computing skills can get something reasonable down in a short space of time.Producing routes for a train sim is almost unique in the computer gaming/sim world, with the need for complex detail either side of the track for scores of miles. In a flight sim, apart from the airports which probably get away with identikit buildings, there is no need for such finite detail unless the scenery author chooses. The terrain is viewed from 30,000 feet or whatever so no worries about someone moaning you haven't included their favourite crag and of course the latest generation of MSFS etc. work wonders with autogen.Or in the case of car rally etc. games, each track is probably only 15 - 20 miles long (perhaps even less) which have been pre-drawn for the particular application and unlikely you are going to get a track editor included. Most of the scenery is optimised for viewing at 60 MPH plus while concentrating on getting your turns etc. right. Ditto golf games, where the action takes place in a relatively small world.As a veteran of the MS Route Editor, producing over 500 miles I have to say that what I did for that simulation was at the cost of just about anything else leisure/computer time wise over two and a half years. That is something I am unlikely to repeat in future in any programme, it was a unique period of time. So again, the production of route based content for TMTS, the TA programme even good old Rail 3D absolutely hinges on the tools being far quicker and easier to use than MSTS. No one expects to go back to Railsim or Train Driver 3 where a 60 mile route could be knocked together in a weekend, but the punishing regime of 3 - 4 months to produce even a passable 50 mile route is no longer tenable (IMHO) either.

Chief Track Welder

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

Quote:- Please, please, please don't make the Route Editor too easy to use. If you do, we'll end up with lots of awful routes,....So what?
It also increases the likelihood of the number of quality routes being created.It is very easy to sort the wheat from the chaff.Bruce.

AustRail

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

>We are dedicated to providing a top quality, easy to use sim. The >same is true for the editors but the final user friendly versions >of those may have to take a second place to the sim itself.I'm sure that's exactly what Kujo executives said too. Exactly.You know when it takes 1-2 years to build a comprehensive route from scratch, whether it's to give away for free or to sell for a nominal sum of money, all in support of some other parties software product thereby making it more, not less viable in the market, it's a little hard to feel okay about being told your needs in doing just that are clearly secondary. You know the old saying, fool me once, shame on you, fool me twice, shame on me.... Well route designers have been kujo'd once already.It is your time and money so I don't begrudge you making a decision to prioritize as you see best. Seriously. I don't. It's how software projects always go. I'm sure you'll understand others are faced with identical decisions, be they route designer, 3d modeler, or end consumer.I will wait and see. A departing word: good luck to you, it'll be a big, big job.Dave NelsonWP mainline, Oakland-Stockton: 100% complete;
SP mainline, Oakland-Stockton: 20% complete;
Aug 1950
<<new .sig under construction>>

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

Quote:- A departing word: good luck to you, it'll be a big, big job.That's several words, Dave.
Incidentally, is that "goodbye" or "au revoir"?Bruce.

AustRail

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

> Incidentally, is that "goodbye" or "au revoir"?Well mine was a strongly worded post and I wanted to make clear it's just a difference of opinions, strong opinions, and not meant as anything personal. Is is just a software game we're discussing afterall, nothing *really* important, and I do genuinely wish them well.As to goodbye, yeah, probably. Having busted my self imposed quota of less than 1-2 T-S posts/month (a protest of sorts), I will depart for other places.Dave NelsonWP mainline, Oakland-Stockton: 100% complete;
SP mainline, Oakland-Stockton: 20% complete;
Aug 1950
<<new .sig under construction>>

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

I don't have a PhD, but I do have a MSEE and PE license (Tennessee) and I have done programming in Fortran, C, C++, and Machine Language off and on for over 30 years, and I tend to agree with Michaels comments. I hope the editor, produced for TMTS is more stable than the KUJU editor, and I would hope that a more comprehensive rebuilder is also available for those inevitable times when glitches occur, but the editor is secondary to the simulation.Now this might change, if PI decides to use some of our better route builders to produce routes (which are essentially databases) for their professional simulators, that can also be used on their hobby version. If the editor was more stable, and had precision placement methods, similar to Microstation and other profession CAD packages, routes could be produced faster. On average, the ones that I have produced so far, have taken 6 months, but it was not constant work, it was an hour or two here or there. However, if PI decided to pay the developer a nominal sum for his work, and then resell it, a better editor would be indirectly profitable, and many routes could be made available quickly. It is just a thought.J. H. Sullivan, P.E.

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

Hi Jerry !I am sure that similar thoughts are on the mind of the PIE-TMTS development management !I think that KUJU too used a certain amount of subcontracted stuff for MSTS ?It would make sense not to increase your design team temporarily, if suitable Graphics Designers are available, at a price, KUJU themselves had problems with.I am sure that many of our best ADD-ON creators, may that be Routes or Rolling Stock, would only be too willing to contribute to TMTS and would not refuse some sort of remuneration for doing so. It would take a long time otherwise for the PIE-TMTS team having to do it themselves getting an attractive package together !O t t o

Alert | IP

[ Home | MSTS Forum | MSTS FAQ | How To... | File Library ] Copyright © 2004 by Nels Anderson. All Rights Reserved.