Home
Reading
Searching
Subscribe
Sponsors
Statistics
Posting
Contact
Spam
Lists
Links
About
Hosting
Filtering
Features Download
Marketing
Archives
FAQ
Blog
 
Gmane
From: Peter Murray <peter-UnoDsp4/C3/2fBVCVOL8/A <at> public.gmane.org>
Subject: [Fedora-users] Re: Accessing a large JPEG2000 datastream
Newsgroups: gmane.comp.cms.fedora-commons.user
Date: Thursday 18th May 2006 01:39:40 UTC (over 11 years ago)
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 5/17/06 2:54 PM, Klein, Michael wrote:
> I¹ve been following this discussion with interest. I admit I don¹t know
a
> lot about the internal file structure of JPEG2000 files, but I assume
> there¹s some kind of header and internal directory that points to the
data
> and metadata streams within the rest of the file.

Any JPEG2000 file is a sequence of "boxes" and each box has 3 fields:
length, type and data (in that order).  A decoder navigates the file box by
box:  it reads the length followed by the type, then decides if the type is
something it is interested in, and if not reads forward the the length
amount to the start of the next box.  (In summary, JPEG2000 is a variable
length coder.)

> As such, it seems like HTTP byteserving might be an option to consider.

I don't think HTTP byte-serving is a good idea for JPEG2000 -- the overhead
of the HTTP header is quite a bit.  There is a specialized protocol defined
as Part 9 of the JPEG2000 standard called JPIP for streaming between a
client and a server.  (See http://www.jpeg.org/jpeg2000/j2kpart9.html
for
more information.)  For what its worth, we might be sponsoring an intern
for the Google Summer of Code to write either the JPIP disseminator on the
server end or a JPIP client as a Java applet for the browser.  (See
http://drc-dev.ohiolink.edu/wiki/ProjectIdeas#JPIPStreamingDisseminatorforFedora
for more info on that.)

> In addition, byte-range support would provide a second benefit ?  the
> efficient delivery of PDF files to browser plugins. Adobe¹s Reader
> plugin already understands how to request byte ranges in order to access
> pages randomly without waiting for the entire PDF to download, and I
> assume there are other clients which do as well.

Agreed -- the PDF use case is one where I'd like to see HTTP byte-range
serving available through a Fedora disseminator or in the Fedora core.


Peter
- --
Peter Murray                       http://www.pandc.org/peter/work/
Assistant Director, Multimedia Systems  tel:+1-614-728-3600;ext=338
OhioLINK: the Ohio Library and Information Network   Columbus, Ohio
The Disruptive Library Technology Jester           http://dltj.org/
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (Darwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFEa9Bc4+t4qSfPIHIRAkzUAJ4mhtfszfVJHvjiGRe6Q+fBmZjD+ACgswaX
TJbaBN+CNQg9D4pqI7p98uE=
=3b6l
-----END PGP SIGNATURE-----
 
CD: 4ms