screwup

[email protected] Sun, 4 May 2003 19:07:10 -0600
Newsgroups gmane.comp.video.mcf.general
Message-ID <[email protected]>
I said 2 meg instead of 2 gig, sorry.. 

Windows has a 2 gig limitation on files in Windows 98.. 

My idea was just to encapsulate the media formats in an object that 
knows how to handle its own storage and retrieval. The object would 
require special (open sourced) libraries initially. The idea is 
rather than having a data format which is interpretted by libraries, 
why not put a special kind of media-knowledgable language in 
libraries and make it accessible to the data format, put methods 
(limited instruction set accessors) for the data format included 
with the data format, so that when an object is passed 
 from platform to platform, it can determine its method of storage, 
it can describe itself, it can maintain itself.. 

The particular advantage of this is that AVI, Quicktime, other file 
formats would overtime become a single media object, the media 
object would make these proported to be unique and proprietary file 
formats (hey we know the real story guys, its just a method of 
branding right?) the same format, and it would end data formats (as 
data formats are interprettive) and bring about media objects.. 

The functions at AVI and Quicktime and such use that are similar 
(DCT's , Wavelet transforms, etc..) would becoming functions usable 
by the objects to construct methods, the objects would use these 
methods to convert the particular flavor of media format to the 
format expected by the user of the object [the application]). The 
object interface (what the programmers use) would have "methods" 
like "getFrame()" and "storeFrame()", "compressMovie
()", "compressFrame()". The media object would have not reliance on 
pixel based imagery, alternatively images could be draw-based or 
vector-based or Postscript, frames could eventually include layers, 
but at the very least a media object can get/store frames, insert 
frames, remove them.. Eventually it can stream, compress.. 

Media objects are not file formats because file formats do not 
contain specialized instructions on how to interpret and store data. 
Data formats without this information become interpretative and open 
to assumptions on the part of the programmers(who must backward 
engineer data formats when they are not openly described). Even 
when they are openly described, if the engineer is not able to 
obtain libraries to read the media format in a language available to 
him, he will create his own libraries based on obsolete information 
or will add his own features to the data format. In some cases 
such needs need to be addressed, but with a static data format its 
impossible to allow new features to be added to a file format while 
still maintaining compatibility across platforms. The engineer would 
then have to propagate his changes to the media format to others who 
will either include the changes (making the media format bloated) or 
or deregard the changes (make the media format potentially obsolete 
if the changes are recognizeably better). With a media object, 
the changes are made to the object not to the library one every 
platform. Only when there is anything functionally unique in 
the file formats, will there be a need to change the libraries, 
so leveraged file formats which are functionally equivalent (the 
data contained is relatively similar, it only looks different) only 
organizationally different (how the data is stored not what it 
stores).   

A media format as such would require more understanding, is 
possible, but here is the glory in doing it "it hasn't been done 
yet"!!!

Why? Because objects when open sourced and working properly are 
impossible to leverage commercially, thus there is no commercially 
viable reason to do it because its not very marketable (to sell 
something you must have some proprietary or unique edge, this is 
the reason why proprietary software is, not because it is 
proprietary its used as an excuse to investors to produce "special 
treament" in the market, so Quicktime is a brand its not a 
collection of file formats that use DCT's and flavors of MPEG with 
motion vectors and frame searches, same with AVI). And when both 
file formats are changed, you can bet there is a lot of commercial 
activity as a result (apple engineers training companies on how to 
convert to the new file format, commercials for what new 
technologies quicktime supports, etc.). 

Once the file formats become unified behind this media object (or 
under it, or inside of it, however you like to look at it), 
the branding/marketing advantages of file formats fade into history 
and the uses for file formats (storing video, photos, sound, 
specialized combinations, layered video/photos, etc..) become the 
focus.. This may also reduce the amount of work engineers have to 
do, but then again who here wants to be working with file formats 
for the rest of their lives.. Wouldn't you rather be working on 
software that makes movie making easier or, photography easier, 
etc.. 
 


Sign up today for your Free E-mail at: http://www.canoe.ca/CanoeMail 


-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf