LÖVE

Hi! Today i'm publishing my Flump runtime for LÖVE2D.
For those who don't know it, Flump allows to export animations made in Flash into a texture + json format.
This allows for smooth animation, low memory footprint, and generally makes it possible to follow a Flash-based workflow when working with LÖVE.

You can find the library here: https://github.com/sixFingers/lump
You find Flump tools here: https://github.com/tconkling/flump

I'm available for support and suggestions, just open an issue / PR on Github.

Your lib is dependent on globals. I did not try it out yet but looking at the code you use globals and the lib depends on globals. Both of which is generally not good design for a library. The library will not work as is in a new project.

Also, not really familiar with what kind of transforms are in flash, but löve does support shearing, and i believe you can "combine" transforms to get what you want, within one lg.push level; granted it needs a bit more math than just manually giving all values of a 4x4 matrix, but should still be possible.

bobbyjones wrote:Your lib is dependent on globals. I did not try it out yet but looking at the code you use globals and the lib depends on globals. Both of which is generally not good design for a library. The library will not work as is in a new project.

Hi bobbyjones, thanks for your comments. The library has been extracted from a game of mine, and still needs some work. Take it as a first, very rough iteration. Some things i'm already fixing are:

- collapse the library in a single file
- try to fix that "globals" problem. As it is now, the two globals being utilized are JSON and Class. I'm still not sure if it's worth it to incorporate both; i may let the developer take care of decoding JSON, so he/she can choose which JSON library to use; while i don't really know how to remove the dependency on Class. Any suggestion is really welcomed.

zorg wrote:Also, not really familiar with what kind of transforms are in flash, but löve does support shearing, and i believe you can "combine" transforms to get what you want, within one lg.push level; granted it needs a bit more math than just manually giving all values of a 4x4 matrix, but should still be possible.

Yep, i know, but after a long while, i couldn't find a way to calculate it. The strangeness begins with Flash, which exports rotation _and_ skew in just two values (instead of three). I really think it's possible, but was unable to come up with a solution. Again, any hints are really appreciated

Look great. Thanks.

intAligned wrote:while i don't really know how to remove the dependency on Class. Any suggestion is really welcomed.

You could just use Class Commons; it's a set of common features/API supported by a variety of libraries, so the user can just use any one of those and your lib should just work.

intAligned wrote:Yep, i know, but after a long while, i couldn't find a way to calculate it. The strangeness begins with Flash, which exports rotation _and_ skew in just two values (instead of three). I really think it's possible, but was unable to come up with a solution. Again, any hints are really appreciated

Do you have any pointer to the file format where these values are specified?

pgimeno wrote:

intAligned wrote:Yep, i know, but after a long while, i couldn't find a way to calculate it. The strangeness begins with Flash, which exports rotation _and_ skew in just two values (instead of three). I really think it's possible, but was unable to come up with a solution. Again, any hints are really appreciated

Do you have any pointer to the file format where these values are specified?

pgimeno, Flump exports a plain JSON. You can find a non-modified, standard Flump json in the demo directory at the link I provided. That's the very same file read by all other runtimes available for other languages. What Flump does is reading transforms from the SWF you publish from Flash. So the problem is inherent to Flash itself. More info on the issue can be deduced from https://github.com/tconkling/flump/issues/6. If you want to play and see if you get it working, again the demo is enough, as it has some skews, which my runtime can't handle. As an hint, you can detect which frames are skewing by looking at the values for the "rotation" field. When they're different, one of them (always the same, still can't figure it out) is the rotation, the other the skew. If they're equal, there's no skew but just rotation.

I don't have any Flash testcases to experiment with, sorry. I'd like to know what the two values are for:

• a figure without any rotation
• a figure with a 30° rotation
• a figure where the vertical axis remains vertical, but the horizontal axis is rotated by 30°

My current hypothesis is that the two values may be the angles of the axes of the transformation basis, with respect to the original axes. If that's the case, then my prediction is that the values will be, respectively:

• 0 and 0 for a figure with a 0° rotation
• 0.523 and 0.523 for a figure with a 30° rotation
• 0.523 and 0 for the rhomboidal figure described above

It will be interesting to figure out the skew transform if that's the case.