Re: [MMUSIC] An error in SDP syntax regarding r= lines?

[Colin Perkins <csp@isi.edu>]

--> Federico Montesino Pouzols writes:
>Hello everyone,
>
>  On section 6 in draft-ietf-mmusic-sdp-new-10.txt the following
>examples of r= lines are given:
>
># r=604800 3600 0 90000
># r=7d 1h 0 25h
>
>  From this examples, I understand that "0" is a valid value for the
>offsets. However on appendix A r= lines are defined like this:
>
># repeat-fields =       "r=" repeat-interval SP typed-time
>#                      1*(SP typed-time)
>
># repeat-interval =     typed-time
>
># typed-time =          POS-DIGIT *DIGIT [fixed-len-time-unit]
>
># POS-DIGIT =           %x31-39 ; 1 - 9
>
>  The two examples above would violate this syntax, since the first
>offset is "0", which is not a POS-DIGIT. In RFC 2327, typed-time
>started with DIGIT, so there was no problem.
>
>  It seems to me, correct me please if I am wrong, that the first
>offset, "0", is not necessary and could be assumed. Anyway, to keep
>backward compatibility it has to be there. Would the following rule
>solve the problem?
>
># repeat-fields =       "r=" repeat-interval SP typed-time SP "0"
>#                      1*(SP typed-time)

I agree that it looks like a bug in the ABNF. A better fix might be to
change typed-time to be

	typed-time = 1*DIGIT [fixed-len-time-unit]

to avoid mandating a leading zero?

Colin

_______________________________________________
mmusic mailing list
mmusic@ietf.org
https://www1.ietf.org/mailman/listinfo/mmusic