Talk:Wiimms SZS Tools

@wimm Can you make the Yaz0 compression page?

I can't find out how it works...

kHacker35000vr 17:04, 6 April 2011 (CEST)

Yes, it's already planned, At the moment I make the U8. Perhaps my sources may help:

• yaz0 header
• decompression function
• compression funtion

Do you need it now?

Wiimm 22:17, 6 April 2011 (CEST)

I have noticed that you can't shift kmp points without encoding or decoding files, this can be very annoying. Maybe you can add a function for this.
kHacker35000vr 21:26, 1 January 2012 (CET)

Do you mean the new options --shift -scale .... ?

For them there is no decoding needed.

Wiimm 22:18, 1 January 2012 (CET)

I have tested alot of things but didn't manage to do without decoding, I'm not used to barch scripts at all but at least it worked.

kHacker35000vr 00:45, 2 January 2012 (CET)

Again: Do you mean the new options --shift -scale .... ?

If not: What do you mean exactly?

Perhaps this may help. If I work with a track, I have extracted it with "wszst extract --decode". After editing the KMP text file I create a new szs with "wszst create track.d --overwrite --fast" and a second later the file track.szs is ready. A second script put the track in an already extracted ISO image, patch the track names in the BMG files, and create and copy a new image to my WBFS drive. This takes about 40 seconds. And then I'm ready to test & play directly on my wii.

Wiimm 01:08, 2 January 2012 (CET)

Forgot to say, I use the --shift command and I didn't have to decode it to txt file. I used wkmpt DECODE course.kmp --shift x,y,z but it would be handier if there was a way just to edit the kmp without decoding to .txt.

kHacker35000vr 11:23, 2 January 2012 (CET)

Then read this and use "wkmpt encode ..." as explained here. And don't worry about encode/decode: Both commands read both file types (binary+text) and only the output is encoded (binary) or decoded (text).

Wiimm 12:36, 2 January 2012 (CET)

For some reason my kmp froze after using your tool, could you look at my kmp? here, I can't find any error

kHacker35000vr 14:46, 2 January 2012 (CET)

Please give the the track before modifying too. Wiimm 16:02, 2 January 2012 (CET)

Here's the whole track: [1]

kHacker35000vr 16:09, 2 January 2012 (CET)

Was the freeze there before shifting? If yes: Please give me the KMP from before.

Wiimm 17:40, 2 January 2012 (CET)

Got it!

I havechanged 2 things in 1 turn:

• It was either that some enemy points are to close together (I know this problem for item points).
• Or it was (more possible) the mandatory checkpoints:
  • Sections 2+3 are parallel, but only section 3 has a point with mode==4
  • Section 6 has out of order point with mode==2

Here is my KMP.

Wiimm 18:36, 2 January 2012 (CET)

The course still freezes before loading, and because I used the kmp of the orginal model it should work fine. The only thing outside the szs modifier I did was shifting the points.

kHacker35000vr 20:35, 2 January 2012 (CET)

Did you see my second message: I got it! Wiimm 21:17, 2 January 2012 (CET)

I have downloaded the one you gave in the link, I assumed that it was the fixed one. --kHacker35000vr 21:46, 2 January 2012 (CET)

I have tested it again without freeze. Here is the complete track. The zip files contains 2 versions: In one I have removed all invalid objects with ID 0. No problems with both tracks (tested on a Wii).

Wiimm 23:22, 2 January 2012 (CET)

It's still not loading, not on the wii and not on dolphin. It couldn't be the slot because I have replaced ridgehighway_course as it should. Which mode do you use when testing?

kHacker35000vr 10:55, 3 January 2012 (CET)

I'm used slots 3.1 and 3.2 (star cup) for the tests. I haven't tried he special slot 4.2 (moonview), because many tacks freeze here.

Wiimm 10:59, 3 January 2012 (CET)

Did you create your console with Visual C# 2010 Express using C Sharp? And could you send me the .cs file? Thanks. --HackSizer 08:48, 7 April 2012 (CEST)

I use POSIX/GNU C, and fpr Windows Cygwin. The code is GPL2, you can find it here: http://opensvn.wiimm.de/viewvc/wii/trunk/wiimms-szs-tools/

There is no special code for the console.

Wiimm 09:45, 7 April 2012 (CEST)

KCL Tool Bugs

Recently I've been experiencing some issues with your KCL Tools. Using the current version of your Tools to lower the walls in my KCL, it added a fall boundary that had not existed before in the KCL Model I created (I have checked several times). See this: [2] (Skip to around 2:50). It caused my course to have a lap count failure because that fall boundary is by a check point that re-spawns you ahead. It is normally impossible to perform this, because there was never anyway to respawn at that point. I don't know what could have caused this issue, but I thought that I should let you know since this only occur ed after I ran it through the wkclt.--SuperMario64DS 16:48, 18 June 2012 (UTC)

If you lower the walls it is possible, that a hidden fall boundary is now "visible". Please send me the KCL before lowering the walls for analysis.

btw: The track is now much better.

Wiimm 18:49, 18 June 2012 (UTC)

Thanks! You can just use the KCL that already exists on the track, as that is the same KCL but without lowered walls (And I did use the lower wall script for the bad KCL). Hopefully this problem will be resolved soon, it's such a useful feature. --SuperMario64DS 21:20, 18 June 2012 (UTC)

But I need the KCL before lowering the walls to compare it and to found the bug, if it is a bug.

Wiimm 15:27, 19 June 2012 (UTC)

I had quickly updated the download with a KCL Fix before anybody could download it (Although it shows that about 4-5 got it before then). That KCL is just like the one I had before I fixed the walls, so it should work (Although it has one extra face). I remember awhile back I had the same issue when simply running the KCL I made through your tools, on the first bridge there was a random fall boundary to the right of the bridge, even if one did not exist in the model. If you'd like the tools that I made the KCL with, then ask Chadderz. It's a debugged version of the CTools KCL Editor. Oddly enough, it also had a similar bug thag adds an invisible wall that I cannot delete, so I tried to use your tools to fix it which instead added the fall boundary in a different area. I no longer have the original KCL I made before then in KCL Form, but I do have it in OBJ form, and if you'd like, I can generate a KCL using the tools I used to re-create the KCL prior to the fall boundary. Somebody I know was also one of th original few to end up with the bad KCL, so I can get that for you as well. --SuperMario64DS 15:43, 19 June 2012 (UTC)

And here are the results of my analysis:

My tools aren't buggy, but your KCL is it. Let me explain it:

• Per definition a floating point variable can hold more the the usual values. NAN, -NAN (not a number), INF and -INF (infinitive) and also -NULL.
• In your KCL there are 10 triangles with length NAN, 5 are "fall boundary". The internal handling didn't expect such invalid data and the fall boundary becomes active.

Solution: Use my next release, which which is more sensible on this, and drop this invalid data. Or you can use this new KCL.

Wiimm 14:00, 22 June 2012 (UTC)

V1.17 BUG

The Tools can scale Model,KCL,KMP,VRCORN and the minimap but it can´t scale the Objects´s KCL. They stay where they were in the original.(Mushroom Gorge & Bowser Castle) --Rukasudo90 18:51, 24 April 2013 (GMT)