Mediawiki Score Extension.

[Eric]

I don't think 2.16 code works with 2.18, which would be a problem. I wonder if the convert routine provided with 2.18 Lilypond is transportable (it converts, not always necessarily 100% exactly well but there you are, 2.16-or-earlier code to 2.18-code - so long as the code ("code"? I should know what Lilypond stuff is called...) contains the \version line (which most of the incipit codes don't... may I suggest including the line

\version "2.14.2" or as appropriate in incipits from now on?... )

My mistake, 2.14.2. I tried inserting that line and got a "program too old" error. We're not using 2.16, even ,we're using 2.14! So insert

\version "2.14.2" early in the text, near \key ... etc...

[imslp]

OK, will upgrade to 2.18 by tomorrow. Better to fix it early.

It's a pretty bad decision to break compatibility between minor versions, though.

[Eric]

the change from 2.16.1 to 2.16.2, etc. (or the unstable branch, more, since there usually isn't a 2.16.2, but lots of 2.17.x, or 2.15.x, on the way between 2.16 to 2.18 or 2.14 to 2.16) is a change between minor versions - the change from 2.14 to 2.16, from 2.16 to 2.18 is actually regarded (admittedly, not from the usual point of view- it's not a LilyPond 2 to LilyPond 3 change- more regarded by LilyPond "fans", I think) as a major change, I think... though I do understand your point - hadn't thought about it previously, it's true.

[horndude77]

It's a pretty bad decision to break compatibility between minor versions, though.

In Lilypond-land the even minor numbers are the stable versions (2.14, 2.16, 2.18, etc.). And these aren't frequently released (2.18: 2013 Dec 29, 2.16: 2012 Aug 24, 2.14: 2011 Jun 5). Breaking changes do occur between these. However in this case (2.16->2.18) most people won't run into issues. The highlight of these changes are here: http://lilypond.org/doc/v2.18/Documenta ... index.html. Most of the syntax changes are tame compared with the differences between some previous versions (manual bar line changes, an articulation change, tempo range change, etc.). I have heard of talk of an even greater syntax stabilization with more stringent rules about what can change and calling that 3.0 (either a rumor or a very long term goal).

The gist: It's a good idea to upgrade to 2.18 sooner than later (while the number of <score> sections are few). Most short snippets won't require changes.

[imslp]

Understood. I was just surprised because in most cases there is backwards compatibility going back a while (I asked MuseScore devs a few years ago, and they promised any files created by the beta release will be readable in v1).

I'm sure the Lilypond devs have their reasons (insofar as the programmer is concerned it is always cleaner to break compatibility - it's just that usually the users complain so we can't do it). In any case, we'll upgrade to 2.18 and keep it at that version.

[Eric]

Hrm. Your reference to groups of "Musescore devs" and "Lilypond devs" made me wonder how the former operates, but I see that the former, like the latter, is an open-source project with, it seems, also a more or less open developer group ...

[imslp]

I think MuseScore has a lead developer (actually, they do: http://musescore.com/team), I don't know about Lilypond. Each project is different and the development dynamics are frequently different as well (albeit all open source), e.g. some projects are more centrally regulated/moderated than others.

[Eric]

here's some information from the lilypond.org page, and I don't know if the background is somewhat relevant...

[Eric]

well, now that the base software has been updated, I was concerned that my (by now quite a few) incipits would break (causing the relevant pages to break too, as I recall), but it looks like the relevant things may be backward-compatible after all, and that in 2.18 one -can- use

"\once \override Hairpin.to-barline = ##f" (the new way of writing that the crescendo/decrescendo that's about to start should end after the barline, not just before it, but that this override should only apply this "once") rather than the old

"\once \override Hairpin #'to-barline = ##f" - but that this (and similar tweaks one can find in the documentation...) which I tend to put in whenever it looks like the composer really wanted it - still actually work(s), they don't break the file. ... neat.

[imslp]

Nice. Yes, Lilypond is now 2.18. Tell me if there are any problems.

[Eric]

There does seem to be that problem now, yes. I wonder if it's fixable with a couple of added (though bothersome) things- will have to check.

[imslp]

OK, so I wasn't imagining it. Tell me if you find a fix to the oversized noteheads.

[Generoso]

These musical examples could also be used to show wrong (or differences in the) notes in the various editions.
See the bottom of
http://imslp.org/wiki/Cello_Concerto,_O ... on%C3%ADn)

[horndude77]

These musical examples could also be used to show wrong (or differences in the) notes in the various editions.
See the bottom of
http://imslp.org/wiki/Cello_Concerto,_O ... on%C3%ADn)

You might consider something like this instead:

Code: Select all

\new Voice \with { \consists "Balloon_engraver" } \relative c' {
  \clef bass
  \key b \minor
  \once \override Score.TimeSignature.stencil = ##f
  <b dis, g,-\balloonText #'(2 . 6.5) "Difference 1">4 <a dis, fis,> <g c, ees,> <fis b,-\balloonText #'(-2 . -3.5) "Difference 2" dis,>
}

I think it's clearer keeping it with a bit of context.

[Eric]

In general, would people linking back to our main site please test the links before going live with them? (hint- right-parentheses. The solution is, I think, either to use the "URL" button up on above the box, or to use %29 instead of an actual ")", since %29 won't get dropped by the software en route. If you don't know what I'm talking about (i.e. that there's a problem/why there's a problem/that links fail), fair enough - but also don't post links to IMSLP workpages.