Hi Harm, On Thu, Nov 1, 2012 at 9:57 AM, Thomas Morley <thomasmorle...@googlemail.com> wrote:
> Hi David (N), > > attached my first try. > > If the markup-property full-length is set #t, some calculations ensure > that no space occurs at line-end/begin. Some dashes will only be > partly visible, but I think its ok. > In addition I wrote a top-level-testing-function to minimize typing. > > Some comments are added. > But I have no clue why I need 13 as a factor in a specific calculation. I'm not sure. Could it have something to do with 'blot-diameter of the rounded edges? > > What do you think? I like the option to have the dashed-line stencil function available for the user to plug in whatever values and get those exact results (full-length = #f). As far as the adjusted line goes, however, I'm thinking that the segments at beginning and end could be complete, not partly visible in some cases. I see that you modify on and off in one circumstance; why not do this whenever full-length is set to #t? The interface could take the settings of on and off as a wish list, and shorten and lengthen them to fit the actual situation. I suppose in this circumstance phase would be set to 0. You could assemble the dashed line from round-filled-boxes as dashed bar lines are (as you mentioned earlier), but would there be any need for that if all dashes and spaces are consistently sized? I'll keep thinking about this, but I hope my comments here are helpful... Best, David _______________________________________________ lilypond-devel mailing list lilypond-devel@gnu.org https://lists.gnu.org/mailman/listinfo/lilypond-devel
