XLua, MMF2 Extension

if it'll let me pump data from mmf to lua without having a lot of

"ok lua here you go"
"kthx send another"
"ok here you go"
"kthx send another"
"ok here you go"
"kthx send another"
"no more lol"
":("

then yeah it'd be fine

On event whatever
> Set Variable t.[1] = "blah"
> Set Variable t.[2] = "blek"
> Set Variable t.[3] = "blarg"
> Set Variable t.strkey = "blug"

On some looping event
> Set Variable "t.[" + Str$(Loopindex("something")) + "]" = "variblah"

On some repeating event
> Set Variable "t.[" + Str$(TableLen("XLua", "t")) + "]" = "variblah"

hot

also is "t.[]" even legal syntax

No, but a.b.[3].c.[4].[5] is much easier to process than a.b[3].c[4][5]

You could also do a.b.3.c.4.5 which also isn't valid lua syntax, but that's equivilent to a.b["3"].c["4"]["5"].

http://hocuspocus.taloncrossing.com/rii/xlua-dist.zip

New release. Name parsing is completely rewritten, and now you can use any valid Lua syntax for naming a variable or function. Contrary to what I said earlier, that means you'd use the mylist[4] syntax. colors['blue'] is also valid syntax, which is the same as colors.blue. Note, mylist.4 is valid in XLua but is actually the same as mylist['4'], which is not the same as mylist[4]. You can also do multi-indexing, e.g. pos[1][2][3].

Added a "Delete Variable" action, which sets a variable to nil.

Added "Table Length" and "Table Max Index" expressions, which are equivilent to the length operator (#) and table.maxn, respectively.

Fixed a small bug in the print function.

Since there was a major rewrite of significant component, there may be a bug buried in there somewhere, so be watchful for that.

I.. I'm not exactly sure what the new changes are, but it sounds like it involves the variable and table system I got going D:

is it possible to explain it in a way so that mentally retarded foreigners can understand it

it means you can set a table's data from mmf

oh okay that sounds brilliant

i will not stand for this

on topic, i'll try fiddling around with this for fun.

Fixed SetAngle/GetAngle in the HWA runtime

Updated the opengl.dll for Looki's update to the OpenGL object.

i'm assuming "Add backdrop" is one of those functions you can't get to, right

did i ask this already

speaking of which, is "set shader parameter" possible to put into the mmf interface

Apparently Push Floating Parameter / Return Value still wasn't working. I though the rSDK floating parameter functions were working, but they aren't anymore, so I had to switch to some Click SDK functions.

Anyway updated build, should have working floating support.

EDIT: I missed the last couple posts

I'll have to look further into add backdrop support.

Shader parameter should be possible in the future, but not until the HWA SDK is released.

This post has been edited by Retriever II on May 14 2009, 10:40 PM

i mentioned this because "add backdrop" being added along with the two backdrop-deleting actions would be orgasmic

especially because i could probably use those with stationary objects like blocks so i can cheat and be lazy with object-object collisions

This post has been edited by Xgoff on May 14 2009, 10:53 PM

I have a general idea of what functions might be used in backdrop pasting, but they are undocumented (they don't even give the argument names in the prototype, just the types). There's also management involved that I don't know enough about, and even if I knew all that, it would only work with active objects (Most of the Object functions, as it stands, are only safe to use with active objects).

Just define a few MMF functions and add them to the mmf namespace. Qualifiers can be amazing help.

I've managed to add two new MMFI functions:

mmf.Object.OverlapsBackground (objid)
mmf.Object.OverlapsObject (objid, objid, layerMode)

(mmf.CollisionMask.TestSprite exists as an alias of mmf.Object.OverlapsBackground)

This gives you rudimentary collision detection support for active objects. Collisions between objects only test between two specific instances. I haven't figured out the more general "type x overlaps type y" yet.

http://hocuspocus.taloncrossing.com/rii/xlua-dist.zip

jesus christ rii MY PANTS

You might want to do something about that

After fighting to invent my own documentation for this crap, I've conquered more of Collision Detection. I've added the following functions:

mmf.Object.OverlapsObjectClass(objid, classid, [layerMode])
This function returns true if the object overlaps any instance of an object "class", which is just the unique number that all instances of a given object share.

mmf.Object.GetOverlapList(objid, classid, [layerMode])
Similar to the above, but this returns a table containing the fixed IDs of every object that's currently overlapped by the object. If classid is given the special constant mmf.ANY_CLASS, then it will return every object currently overlapped. This will return objects regardless of whether or not they're exported.

mmf.Object.GetClass(objid)
Returns the object class of an object, to support the other 2 functions.

So what do you do with fixed IDs? This is a great time to take advantage of that mmf.ExportCallback function, which you can use to automatically build a reverse-lookup table for fixedID -> exportID.

As I continue to dig through this I'll look to add support for simpler bounding-box testing, and maybe at least differentiating obstacle/platform/ladder overlap for backgrounds.

ohhhh shiitttt

this isn't even possible in standard mmf (even Xgoff tried)