C2tP

Thanks for the reply! Looking forward to future updates!

Hey guys I tried searching the forum but couldn't quite find an answer. Sorry if this has been asked before. When editing Z-Order is there any way to do multiple frames at the same time? Not the "apply to all frames" but to specific frames. For example a hand may by default always be in front of the body but based on the animation the hand may go behind the main body during a short period of time. Currently it seems like I have to select every key frame individually and change the z-order of the image. I want to select multiple key frames and change them all at once. Is this possible? Thank you

Cool! Glad you got it working! Yep gotta be careful when adding the other outside information like screen position. This is the game programming idea of spaces. Need to do all the animation work in "local space" and then shift to "world space." Congrats on getting working implementation. :D -C2tP

No I am using my own personal implementation of scml in my own personal engine. -200 and -90 are the results you want. Remember -90 and 270 are the same thing. The problem you're having lies in the idea of pivot points. The x and y positions in scml shouldn't be thought of as screen coordinates. They are offsets from the origin. So if some image is floating to the right 50 pixels, when you flip It should float 50 pixels to the left, aka -50. This means that the 0,0 point in scml is the *whole* image's pivot point. So you should make your images centered in scml if you can. If you didn't make your animations centered in the editor, its not a huge deal. Basically you'd be drawing something like you would if you were using sprites. When you want to draw a sprite centered at positionX, you'd do something like draw(positionX - (width*flip /2)) This would draw the sprite centered at positionX. This is just an example if you are not using centered pivot points. Basically long story short, I assume you made your animation to the right of the origin (all objects have positive x values) thus are getting big movements when you flip. Just offset the draw when flipped by the animation-as-a-whole's width. Remember, the origin in scml editor can be considered the animation's pivot point. I'm at work now (using phone) but if ya need ill draw pictures and examples when I get home.

Ok I didn't really check the forums for a few days and seems this thread is still going on here so I'll comment again. I have working solution to flipping in my engine. First off, dont flip the bones. keep all the calculations and work flow the same. You're building the 'object' as a whole, facing the default direction. (not saying you cant workout the transformations with the bones but this way works too...and i think less error prone if anyone is having trouble doing it the bone way) In order to flip you only need to flip the actual image object right before you draw. So you build the object up, do all the animations/tweening and when everything is good to go, you flip. Here is some pseudo code: SCMLObject* obj = blahblahblah; processObject(SCMLObject* obj, ...){ //get current key and next key info for this object //tween //add parent's information if(flip){ obj->x *= -1.0f; obj->scale_x *= -1.0f; obj->angle *= -1.0f; } Renderer->drawObject(obj); } I used the example knight that grimfang has in his sample folder. and flipped: (image link: http://imgur.com/2pa1K,RsAVe) As you can see they are flipped nicely, and yes the animations are working nicely too. I hope this works out for you guys too. -C2tP

This is probably outside information you'd pass into your SCML draw calls. (aka character information like position and facing direction and stuff) Drawing the animation flipped can be implemented different ways. This can be a use of different texture coordinates, using a negative scaleX value, etc etc.

I think these are probably the most important. I definitely want to add generic triggers. These can be used for numerous things and should be considered required for 1.0 scripting sounds cool but can definitely come later. A first party c++ lib would definitely be icing on the cake for selling purposes, but ya some people have made some public APIs(good job grimfang :D ). Also there are so many engines and different languages out there it'd be more overhead for Brashmonkey and I think their dev time might be better spent on just the editor. Leaving API to community might be the best decision. Honestly though most people I think will want to change and adapt it to their own engine or whatever so a generic API can only go so far. (great if you want to use it as a base though!) Also implementation in c++ took me one afternoon starting from scratch with reading in the XML to drawing and animating so even if you do your own implementation you should be able to punch it out quick I think.

Ok I'm kind of spamming now but I got one more. I would like to see a way to propagate a newly inserted item (bone or object) to all future keyframes (from the point you inserted). say you have a big long animation...its long and awesome and works great. then you copy that animation and basically use it to create another set of animations for a different character...only this one has a hat. When inserting an object, you should have an option to propagate the hat through all frames. if i have 8 animations with 50 keyframes each and i want to add a hat.... I think in its current state I'd have to go through all 400 frames and add the hat manually. Granted, I'd have to go through the animations and animate the hat, but seems like a big time saver and headache-reducer if I can throw in the hat object, click apply to all frames and bam! got a hat on every frame. Seems like a pretty simple and straight forward addition that I think a lot of people would use. Thanks again, C2tP

The current monster example looks to be an example build from august. Is there a new one coming soon? I have finished implementing my bone system in my engine and all tests seem to be working so I want to test with like a proper fully animated model, basically exactly like the old monster but with using bones. ALSO please please please, in your future example SCML, make all textures power-of-two. Not all hardware can load the images you released for the monster (example. foot_0.png is 139x159 ). I can't just add padding to these in photoshop because of pivot points. Thank You, C2tP

Ok here's a few other things I would like to see. 1. Grid (super easy) draw some more lines on the canvas (main view). this helps give perspective of how big you're making something. Custom grid spacing would be nice too. And it would be an awesome bonus if there was a photoshop style thing where you can create guide lines by dragging from the ruler. For instance say my game is in a specific resolution...say 800x600, and i want my artist/animator to make this end game score animation that takes up the whole screen and involves a lot of scaling and such. It'd be nice to set some lines to represent the actual game's canvas(800x600) so he can animate to fit. (instead of mousing all over the place to see the cursor position at the bottom) 2. Quick hotkey (or just do it automatically) to open of the image information in the "Objects in Frame" viewer. If i click a bone it pops up all the bone information (x,y,scale,angle,etc) but if i click a texture, first i gotta switch windows on the left side, then scroll up/down to find the selected image, then click the little expand arrow. It's a pretty big pain when im going back and forth between images/bones to look at their info 3. Display width/height somewhere in the image data area (next to all the x/y/scaleX/scaleY/angle/etc stuff) If you do a lot of scaling...i want to know EXACTLY how many pixels (width/height) my image is using. Not only for general ease of mind, but if you expand and then contract I want to *make sure* that it's back to it's original width/height. This seems straight forward by making sure scaleX/scaleY is back to 1.0f but when you add bones into the mix, its not as easy to read. Like a bone could have a 0.56344 scale X which might offset the image scale X to 2.54567 or some random number...which obviously isn't very readable. Just displaying the current tweened height/width would be enough for me...but hey could also be cool to auto adjust scaling by entering width/height into that info box. So if i just type in 64x64 it will adjust the scaling to that it will draw at that size (be the scaling 1.0 or 2.54567356). Thanks again! C2tP

Please change stretching in the viewer to only change scaleX and scaleY. Right now it changes both the scale and the actual x/y values. This doesn't really make sense to me why it would do this. So for now I can only do scaling on the left in the "Objects In Frame" window by changing the numbers there. Thank you C2tP