Atte Andr� Jensen wrote:
>On Thu, 12 Oct 2000, Derek Lane-Smith wrote:
>
>> I dunno, I think having two separate header fields would be cleaner and
>> less confusing than having several options for the current K: field. If
>> the two fields were both filled it provides some redundancy. If they were
>> in conflict, you can generate an error message, that would be missed if
>> there were only one.
>
>Which is exactly a good reason for having only one field. Redundancy is
>not good! Like in a batabase where you strive to register your information
>exactly once.
Yes, I agree with that. However, the same argument applies if you put
the information into a single field twice.
You do have to be careful with proposals to mess with the K: field,
because it serves two purposes; it not only supplies key/mode info
but serves to delimit the tune header. Introducing a k: field before
the existing K: will work:
X:1
T:test
M:6/8
k:^c^f
K:D
abc abc||
BarFly (and probably most other programs) ignores unknown fields in
the header, so this causes no problems. If you do it the other way
round:
X:2
T:test
M:6/8
K:D
k:^c^f
abc abc||
it will cause an error, and of course if you leave out the K: field
it will break all existing programs.
I can't see that this serves any purpose which is not already served
by the long-established K:D %^c^f or just K:D % two sharps.
Phil Taylor
To subscribe/unsubscribe, point your browser to: http://www.tullochgorm.com/lists.html
