[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message] Re: PSVI formalization
[Alaric Snell] > On Friday 10 May 2002 13:53, Thomas B. Passin wrote: > > > True (see SOAP+attachmemts). XML is a good candidate for sending > > metadata about binary data. > > > > As in MPEG-7, if I understand things right. > > But why does any data need to be 'binary'? XML is an interchange format, but > it falls apart for certain types of information such as images... large > arrays and so on. Other interchange formats don't. You can encode images in > PER quite happily, it's just a "SEQUENCE OF SEQUENCE OF SEQUENCE { red > INTEGER, green INTEGER, blue INTEGER }" from memory (might want to constrain > those R/G/B integers into a sensible range to make them one or two bytes > each). > > Those giant arrays can fit seamlessly in with the "SEQUENCE OF Paragraph" > describing the image to humans and the "ImageInformation ::= SEQUENCE { name > UTF8String, width INTEGER, depth INTEGER }" and be processed with the same > tools. You can pull out the colour of the top left pixel of the image with > what passes for XPath in ASN.1-land. > I don't know much about MPEG-7, but my impression is that the metadata they want to deal with is much more ambitious than describing the number of lines and frames. I think they want to describe more abstract characteristics of scenes, and generally to be able to do SMIL-like things. These properties don't seem to me to properly belong in with the image data. If this view is roughly right, it makes a lot of sense to me to have the metadata separate, just like it makes sense for card catalogs to be separate from books. Cheers, Tom P
|
PURCHASE STYLUS STUDIO ONLINE TODAY!Purchasing Stylus Studio from our online shop is Easy, Secure and Value Priced! Download The World's Best XML IDE!Accelerate XML development with our award-winning XML IDE - Download a free trial today! Subscribe in XML format
|