Posted by pepe
on May 13, 2007 at 3:18 AM PDT
Just a follow-up of what is going on with Swash. This post has absolutly nothing to say about javaone.
Javaone is everywhere, looks like i will have to get there one time, just to know how it feels. Too bad i have to cross the ocean for this ( /me dreams of an european javaone... ).
Here is a totally non J1 related post, for your minds to rest from that avalanche of posts, news, information of all kinds.
I have been very silent for latest weeks, due to huge lack of free time.
Nevertheless, there are things being prepared within the rare bits of free time i can gather and when my day job gives me an occasion to work on domains that crosses my night job ones.
I am investigating an XML file type to save timelines. This will permit the creation of an editor within netbeans to edit/save timelines in a codeless mean, all graphically.
This XML will be read by a Timeline factory and give you the possibility to load different timelines with same root code. There are still things to solve with the links to swing (how do i define which JComponent of my swing program will be linked to my DisplayableActor? how will i know there are DisplayableActors to feed? What about listeners, ...)
I am also thinking about making the XML read and compilation streamable, that is, like flash, some parts of the timeline could play without all media being loaded. For this feature, an asynchrounous feeder will be created, which will add actors and media to the timeline. I would like the timeline to be detached from the feeder concept, so maybe reference to the feeder will be a timeline-level variable. There are issues to resolve, as to i would not be able to limit count of feeders to one for the timeline, which could result in inconsistencies if multiple feeders are present and need to be accessed. But, maybe it would be interesting to have multiple feeders. Analysing the timeline flow and creating tree of feeders could optimize pre-loading to a great level.
I am preparing the source code for release. It is not complete, needs documentation and cleaning, but it is now working enough and i am confident enough in it for people to read it.
Debugging time related programs is very difficult. I did investigate in stoppable timers within jpda/jvmti, but it seems that there are not much people in that field. Once i have enough time, i will get in that deeper.
I am preparing a serie of blogs with practical examples on uses of swash, which cannot easily be solved by other solutions around. This serie will be preceded by an entry where the spirit and inners of swash will be described.
An article is being prepared too, but as you can imagine, it will still take some time before it is ready.
Oh, and swash can be accessed here . Be sure to check its wiki too .
If you have been reading down there, i love you.
I mean it.