Re: [Lilypond-auto] Issue 4036 in lilypond: Patch: Revert "Issue 3978: Merge alignment cleanup"

[lilypond]

Comment #10 on issue 4036 by address@hidden: Patch: Revert "Issue 3978: Merge  
alignment cleanup"
http://code.google.com/p/lilypond/issues/detail?id=4036

Well, "self-alignment-X does has sensible effect if the object's X-offset  
is set to some other self-alignment-interface property (such as  
aligned-on-x-parent) but the effect is different from that assumed in those  
doc examples (as we see after Janek had Fingering align-on-parent)."

I mean, isn't it pretty clear what we want here?  I don't see that we have  
anything to gain by setting X-offset to something like align-on-parent  
which interprets self-alignment-X for aligning not the "self" but the  
parent, and then mushing together a parent-aligner with a self-aligner by  
using make-simple-closure.  Or whatever.

So what we should have is just a single callback for all those aligners,  
and two properties
parent-alignment-X (which can be ##f for just taking the reference point 0  
or a number between #LEFT and #RIGHT for locating the reference point  
between left and right edge), and self-alignment-X (which can be ##f for  
just taking the reference point 0 or a number between #LEFT and #RIGHT for  
locating the reference point between left and right edge).

If we do that, meddling with self-alignment-X will result in a predictable  
shift of the self-aligned object around the parent reference point, and  
meddling with parent-alignment-X will change the parent reference point.

Yes, we'll have one more property lookup for alignment.  But it should be  
worth the trouble because we then have exactly one meaning for  
self-alignment-X, and we just have a single callback for X-offset when we  
design aligning to the parent.

--
You received this message because this project is configured to send all  
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings