The latest version of Flash Player [1]

I haven't updated my blog for a long time. I 've had a lot of things recently and it's hard to say a word. However, the development of Flash Player has not been stopped, and there are a lot of latencies planned in August 13-in my own style, we recommend that you do something later;

The goal of the new Flash Player design is to embed the system in accordance with Flash Player 8. The actual design standards are based on Flash Lite 3.0. The new Flash Player version sets several key points during the design:

1) added video support and video formats supporting h263 and VP6;

2) Smaller codesize to facilitate the transplantation of embedded platforms;

(3) Upgrade of ActionScript (support for Flash Player 8/Flash Lite 3.0)

4) more convenient port interfaces; the new program architecture supports OS systems and nonos systems (but I still suggest at least a simple OS similar to UCOS)

5) better stability and efficiency;

We will describe the actual progress with several screens:

The above is a video demonstration of pucca and a test case sent to me by a user (not very clear, but mainly tested video rotation). The current situation is basically as follows:

1) video decoding h263 and VP6 have been integrated. Currently, they are pure software decoder, however, you can use the manufacturer's Hardware-accelerated Video Decoder (Video Codec may have a license issue) After porting to a specific hardware ), the new program interface supports hardware-accelerated Video Decoder and MP3 decoder-this version is greatly changed compared to Flash Player 6, especially the interface is almost completely rewritten;

2) the current program is Alpha. In addition to videos, other functions are still being tested, so there is nothing to say about stability and compatibility. At present, bugs are also found, still in debug;

 

Future plans:

1) release a Windows evaluation program later, and a Linux evaluation program;

2) Hardware acceleration (Video Codec and rendering) will be started at the end of the year)

3) I will release a secondary development kit as appropriate. You can use this secondary development kit to develop your own player (note that it is not a plug-in ), in this way, developing your own Flash Player or flash-based interface applications on a specific platform will become possible (due to personal energy problems, I may only open the PC version-too many embedded platforms)

4) The subsequent work focus will not be discussed for the time being. I want to explain some of the preparations;

 

Supplement:

I have seen some arguments that question the efficiency of flash player in the embedded system, but I want to explain several questions:

1) Although vector computation has performance problems, most of the 32-arm platforms now can cope with simple flash even if the MHz operating system is the same as the MHz operating system. At least there is no problem with the UI. See the link later;

2) The current open-source flash, such as gnash, has a big problem in its rendering efficiency (this is why I didn't use gnash). For example, gnash uses the rendering engine transform, rendering performance is poor, especially in embedded systems. Therefore, it takes some time for gnash to run better on the embedded platform;

The actual videos of flash running in the embedded system are as follows:

7x7 platform: (high quality, 4x4) http://www.hygraphics.com/download/FlashPlayer/video/GP32500Demo.wmv

Arm9-platform: (high image quality, 4x4 anti-aliasing, use colordither) http://www.hygraphics.com/download/flashplayer/video/flash000020demo.wmvuses jxd 990 Platform

In actual use, image quality is even disabled, and the performance will be improved a lot (scanline algorithm, the performance improvement is largely related to vertical scanning ), in addition, colordither has a significant impact on rendering efficiency-because we feel that there is still a lot of surplus in processing UI time, so I added colordither.