Arno Gerretsen

Time for a user manual

Tonight I have been working on MDL Tweaker II again. I have finished the vertex list dialog, that allows you to alter vertices of your object and also provides an option to "normalize" the normal vectors of the vertices. This last option is useful if you want to make objects that do not change color with the direction of the sun, like the autogen trees do.

Each time I finish one of the dialogs, the tool is getting closer to loosing the alpha status. The more recent versions also seem to be rather stable already. Before the first beta version will be released, I still want to implement a dialog to view the BGL command list of the object and also a statistics page. But first I am going to add the functionality to the texture list dialog, that allows you to correct the texture size. This would make my little CorrectTexScale tool obsolete, but integrating it into MDL Tweaker II would be a logical choice of course.

The more closer I am getting to a beta version, the closer I am also getting to writing the user manual for this new tool. Until now I just dropped the alpha versions on the forum and everybody played with it. But a user manual is really needed of course. It would help the novice user to learn more about the capabilities of the tool and also for the experienced user it is useful, as he can then discover all hidden features I have put in. So I really think I should start with creating the user manual in the Wiki soon.

CAT is "fixed"

The CAT headache is solved for now, as I have released a new version. As I already wrote about before, this new version does also support the ASM source files created with FSDS3. Although both GMax and FSDS3 use MakeMDL to export and create the ASM file, there was a little structure difference. I have learned CAT these differences now, so that both formats are supported.

But the other problem Nick and I found while debugging a few days ago was more difficult to solve. Actually I could not really solve it, I could only add an option that gives the user a choice between two ways of doing it. And both have there ups and downs.

The first way is to apply the animation condition on the animation trigger. This is how I had been doing it since CAT v1.01. The advantage of this method is that, when the animation condition fails, the reverse animation will nicely be shown to return to the static state. But when multiple copies of the same object (sharing the same GUID thus) were placed this did not work. After some thought, this behavior was very logical of course. When you call a library objects you are expecting the object to share the same memory space (that is one of the advantages of libraries). So that means they also share their local variables to store the animation state. If the animation condition then fails for one of the objects you placed, that means the animation state is set back to static state for all of them. So this means that you animation will never ever start, unless the animation condition is true for all copies you placed at the same time.

One alternative for this would be to create a new GUID for each object you place, but when you need to place 100 gates at an airport or so, this is no real option. It also destroys most of the advantages of using a library. The second alternative is to apply the animation condition, like I did before CAT v1.01. So in that case the animation condition works on the animation itself and not on the trigger. So the trigger can remain shared between all objects and when the animation condition fails a switch is made to the static condition right away, so if the animation condition fails before the trigger fails, this means you get a sudden jump.

In the new version of CAT you can use both methods, so depending on the condition under which you want to use your object, you can choose the best way to apply the animation condition.

CAT headache

Version 1.02 of CAT is get closer to release again. This evening I had a good debugging session with Nick on a bug he had since CAT 1.01. I finally was able to reproduce his animation condition problem. Basically the problem means that the local variables used to store the animation state are shared between all copies of the object you place. So now that the animation condition works on the animation trigger (to make sure you get a nice reverse animation when the condition is no longer true), this means that multiple placements of the same object will block each others trigger, unless the animation condition is true for all of them at the same time. This is really a nasty feature, or should I call it a bug?

One easy option is to change the way the animation condition works back to the status of CAT 1.00, but that means that the animation will suddenly jump back to the static condition when the animation condition is no longer valid. This isn't really nice either. So I think I will go for the second option, think a bit harder on a way to combine both ways of doing it. It just must be possible, must be...

The CAT has grown a bit

Until now CAT was only a tool for users of GMax, but now that FSDS3 is also using MakeMDL it is time to change that. In theory the ASM sources file created by MakeMDL when exporting from FSDS3 should work right away with CAT. But of course the reality is that there are a few minor differences in the structure of the source code. I have now figured out what these differences are and I am patching CAT to understand both "formats" of the ASM file.

So expect an updated version of CAT soon, that will also allow FSDS3 user to use the conditional animations. And besides the changes needed for FSDS3, I also want to implement the IFMSK condition, as this one has been on the to-do list for quite some time already. With the IFMSK condition it will be easier to use CAT together with the ActiGate module.

Seasons

I had a very productive weekend, as I was visiting my parents and could therefore not be disturbed by forum questions or people chatting with me. I had brought my laptop, so I made some nice progress on MDL Tweaker II.

I finished implementing the code that allows you to add seasonal conditions to your textures (and it should also be able to read that information back from the MDL file the next time you open it). You can find the latest alpha version of MDL Tweaker II on the forum. You will have noticed that I wrote alpha version, so be aware of the bugs that might be hanging around in the code.

In the end it proved to be quick a challenge to make a user interface that was both rather easy to use and flexible enough to allow all kind of conditions on the textures. For example some people might only want a HW and a SU version, others want all seasons or maybe only a night variant. In the end I hope the GUI can suit everybodies wishes.

There is only one thing I am still wondering about. Why can't we just assign the seasonal textures as a subtexture type, like we can do with the night textures? That would save us a lot of condition code to switch the textures correctly. The fact that the season variable does not match the season of the mesh scenery makes things even more complex. We now have to use the day of the year variable to get the correct textures that match the mesh scenery transitions as well. I really hope that MS will make seasonal textures easier in FsX or else at least provide us with a season variable that is in sync with the mesh scenery.