Author: michael Date: Fri Nov 10 18:47:01 2006 New Revision: 20838 Modified: trunk/DOCS/tech/nut.txt Log: stream_count a few more words for forward_ptr and file_id_string Modified: trunk/DOCS/tech/nut.txt ============================================================================== --- trunk/DOCS/tech/nut.txt (original) +++ trunk/DOCS/tech/nut.txt Fri Nov 10 18:47:01 2006 @@ -401,6 +401,8 @@ file_id_string "nut/multimedia container\0" + the very first thing in every nut file, usefull for identifying nut + files *_startcode all startcodes start with 'N' @@ -423,9 +425,15 @@ version NUT version. The current value is 3. All lower values are pre-freeze +stream_count + number of streams in this file + forward_ptr size of the packet data (exactly the distance from the first byte after the packet_header to the first byte of the next packet) + every nut packet contains a forward_ptr immedeatly after its startcode + with the exception of frame_code based packets, the forward pointer + can be used to skip over the packet without decoding its contents max_distance max distance between startcodes. If p1 and p2 are the byte
michael wrote:
file_id_string "nut/multimedia container\0" + the very first thing in every nut file, usefull for identifying nut + files
I guess that it could be pointed as the magic string for file (btw I could add a IANA formal request if you thing it's time) lu -- Luca Barbato Gentoo/linux Gentoo/PPC http://dev.gentoo.org/~lu_zero
Hi On Fri, Nov 10, 2006 at 09:50:50PM +0100, Luca Barbato wrote:
michael wrote:
file_id_string "nut/multimedia container\0" + the very first thing in every nut file, usefull for identifying nut + files
I guess that it could be pointed as the magic string for file (btw I could add a IANA formal request if you thing it's time)
please elaborate you mean some sort of official "files starting with "nut/multimedia container\0" are nut files as defined by nut.txt from IANA? if so iam in favor of that unless rich or oded or anyone else has any objections or suggestions to change the string? [...] -- Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB In the past you could go to a library and read, borrow or copy any book Today you'd get arrested for mere telling someone where the library is
Michael Niedermayer wrote:
Hi
On Fri, Nov 10, 2006 at 09:50:50PM +0100, Luca Barbato wrote:
michael wrote:
file_id_string "nut/multimedia container\0" + the very first thing in every nut file, usefull for identifying nut + files
I guess that it could be pointed as the magic string for file (btw I could add a IANA formal request if you thing it's time)
please elaborate you mean some sort of official "files starting with "nut/multimedia container\0" are nut files as defined by nut.txt from IANA?
Something like that http://www.rfc-archive.org/getrfc.php?rfc=4288 We could have a full rfc but I think I'll wait for Rich completing his document and then put his words in the right template.
if so iam in favor of that unless rich or oded or anyone else has any objections or suggestions to change the string?
lu -- Luca Barbato Gentoo/linux Gentoo/PPC http://dev.gentoo.org/~lu_zero
On Sat, Nov 11, 2006 at 11:29:01PM +0100, Michael Niedermayer wrote:
Hi
On Fri, Nov 10, 2006 at 09:50:50PM +0100, Luca Barbato wrote:
michael wrote:
file_id_string "nut/multimedia container\0" + the very first thing in every nut file, usefull for identifying nut + files
I guess that it could be pointed as the magic string for file (btw I could add a IANA formal request if you thing it's time)
please elaborate you mean some sort of official "files starting with "nut/multimedia container\0" are nut files as defined by nut.txt from IANA?
if so iam in favor of that unless rich or oded or anyone else has any objections or suggestions to change the string?
the only suggestion i would make is to mandate that the main header startcode come immediately following the null byte. this is required anyway for a valid nut file so anything doing filetype detection via magic numbers should require it as well.. and it might prevent false positives in some weird borderline cases (e.g. a text file beginning with "nut/multimedia container" where subsequent data is missing and replaced by null bytes). rich
participants (4)
-
Luca Barbato -
michael -
Michael Niedermayer -
Rich Felker