[MEI-L] clef as milestone element

[Andrew Hankinson, Mr]

On 2011-06-07, at 4:24 PM, TW wrote:

> 2011/6/7 Laurent Pugin <laurent at music.mcgill.ca>:
>> On Tue, Jun 7, 2011 at 11:22 AM, Johannes Kepper <kepper at edirom.de> wrote:
>>> 
>>> Am 07.06.2011 um 11:03 schrieb Laurent Pugin:
>>> 
>>> Although I have no example in mind currently, I'm pretty sure that I've seen multiple clefs on one staff for different voices (that is, layers).
>> 
>> Yes, I am aware of such cases.
>> 
> 
> I think I've seen something similar to this:
> 
> http://imageshack.us/photo/my-images/101/shiftedclef.png
> 
> (I made this up.)  I.e. strictly speaking wrong use of the F clef, but
> the piano player immediately knows what's meant.  Should this be
> covered?  If yes, should there be a clef/@line.ges attribute, and
> should the restrictions on clef/@line be relaxed (i.e. integer instead
> of positive integer in the range of the number of lines?

You will need to have some notion of ledger lines, so yes, you should be able to specify negative numbers as the line.


> 
>> 
>> It looks like a good solution to me. I would vote for <clef>, but
>> <clefIndication> would be fine too. Maybe we can think of another
>> @reason value. What about reason="sb"?
>> 
> 
> A fundamental question:  Why "sb" rather than "systembreak"?  I
> personally prefer descriptive naming where you don't have to guess.
> For example, I have no idea what note/@accid="tb" is meant to mean.
> It's not documented either.  Of course one can add an explanation to
> the documentation, but I find it much more accessible to have a
> meaningful attribute or tag name in the first place.

<sb /> follows the convention of the "b", or "breaking" indicators, like "<br />" and "<lb />" (for line breaks) and <pb /> for page breaks. This is largely started in HTML and inherited through TEI.

"tb" on accid stands for "Triple Flat" in MEI 2010. "tf" is now also allowed in the upcoming release for consistency with "f", "ff" and "nf".

Ultimately, and this is something that I think needs bearing in mind, XML is *not* supposed to be meant for human consumption first -- it's meant for computer consumption. I may be opening up a new can of worms, world of hurt, and pouring lemon juice in old wounds simultaneously, but my feeling is that if you want to create a system break, you (as a programmer) would call "createSystemBreak()" in some library and it will do all the necessary checking and insertion for you. Most people will then just need to click a button to insert a system break.

With MEI, though, we're at the stage where there are no real software libraries that will produce this, so we're left to hand encoding them. Frankly, I don't think this is a healthy position to maintain outside of a few specialized projects. My expectation is that we'll be producing software to automatically read and write MEI in the very near future.

All this is to say the difference between "<sb />" and "<systembreak />" is a human-oriented difference specifically for hand-encoders. I know that technically we could replace *everything* with numbers if all we're concerned with is computer readability, but I think it still needs to be somewhat legible for the early adopters and software implementers to get their heads around. After that, though, most people will be generating MEI through other means.

-Andrew

> 
> Thomas
> 
> _______________________________________________
> mei-l mailing list
> mei-l at lists.uni-paderborn.de
> https://lists.uni-paderborn.de/mailman/listinfo/mei-l