Open Goldberg Variations don't load in latest version
I just downloaded a copy of the Open Goldberg Variations from the Musescore sheet music site and tried to open it in Musescore 2.0 beta 1. The corruption appears to begin at variation 26. I'm not sure what's happening, but e.g. the time signature is 18/16, which I'm pretty sure isn't right, and the treble and bass clef notes appear to be out of sync.
This has actually been a problem for at least the past six months, but I didn't get around to reporting it until now because I assumed you guys were already aware of it—sorry about that. The fact that you released a beta with this bug in it suggested to me that perhaps you weren't aware of it. Or possibly it only happens on the Mac version of the software?
| Attachment | Size |
|---|---|
| Open Goldberg Variations.mscz | 154.01 KB |
Comments
I'm not surprised at all that they don't load correctly. They were made with a much older development version, when the MSCZ file structure was in flux. Older 2.0 sample scores usually show some issues when loaded.
Just on the subject.
For a while (on my system anyway), everything in the score sounded drumset notes.
It suddenly sounds fine now in the beta.
The Open Goldberg Variations (OGV) score will have to be revised. Basically this means remove all custom layouting from the score, check for syntax issues, and then relayout the score again. Any takers?
That's a separate bug. Load a score, play it, load another score, play it, the second score plays with the first score's voices.
I don't know how to do the "remove all custom layouting" bit, and I'm not convinced that when you do that you'll wind up with something that works, but I'm certainly willing to be one of the proofreaders. I think you need more than one—it's a lot of notes. :)
The original effort is still valid in many aspects (e.g. audio recording, printed book, MuseScore improvements), but here are my thoughts:
You would probably get a decent, albeit imperfect result in 2.0. I think it would be better to wait until the 2.x (or even 3.x) series has matured (later next year?), as it will be more resilient (bugs fixed, more future-proof, and possibly more).
Maybe it should be typeset from scratch in a stable version (to be safe) and another public review undertaken? I know there might be some hesitation in revisiting the project, but since it would be a special one-off (being the only score produced in a developing version), I'd say it's worth it. I also think the (exquisite) iOS app could benefit from one or two fixes as well - maybe that could be ported to other platforms too.
You can get as perfect a result in 2.0 as you care to spend the effort to get. perhaps you mean, decent but imperfect *by default with no manual adjustments*, but that will remain true forever. I think it unreasonable to expect major changes to layout after 2.0. We do virtually everything one could reasonably expect, the rest will probably "always" require manual adjustments. And we made the defaults predictable so those manual adjustments should always remain "safe" in future versions. I know you'd like MuseScore to one day become LilyPond, but i think that's very unlikely, and if it ever does happen, you'll be able to simply reset all to default and get the new layout improvement automatically.
So I really don't see anything to gain in waiting.
With regards to manual adjustment, my motto has been "intervention only if there's ambiguity". It's probably true what you say about reset. I suppose my concern is MuseScore might not know how to handle surrounding elements that would adapt (or not), especially if the overall environment changed - maybe I don't articulate this well. It might not be a good investment of time, particularly if it's not preserved in the next version, and I don't imagine most people would install multiple just to have one score looking good.
Maybe I am unjustified, but I just think it would be better to wait until this series is mature (2.x might be good enough).
It's not the main reason (especially since MS is a notation program), but we could potentially do stuff like this, for example: #13811: Select playback method of Grace Notes, Arpeggio & Glissando and Articulations & Ornaments
Again, I think it is unwarranted to think there will be any significant layout changes after 2.0. Again, unless you can provide specific counter examples, there is no benefit in waiting. At most, some might change might happen in spacing in some rare corner case situation, so maybe if hat happens and it affects Goldberg, we update it. But when you use terms like "mature" this implies you are under the mistaken impression that signficant layout changes are planned for some mythical 2.x release. But this is simply not the case.
I don't know if I can properly demonstrate examples of my concerns (other than just offering my caution). They may only be realised later by typesetters, reviewers and critics.
If the defaults are improved (e.g. beaming, ties) in maybe 2.1 (or whatever), I could be more convinced.
I can now hear sounds other than piano in the beta (e.g. violin) - it seems unpredictable.
My experience of the "other sounds" issue is that if you load a score and play it, and then you load another score and play it, the second score will play with the sound settings from the first score.
@mellon it would be great if you could share the steps to reproduce this sound problem in another post.
As I keep saying, there really are no plans to make significant improvements to default layout any time soon. MuseScore is not LilyPond. The few cases where our layout is not as good as it could be are almost all cases where we would need a tremendously more sophisticated (read: slow) layout algorithm that took interaction between elements into account. I could imagine such a thing being run as an optional "i'm feeling lucky" pass, but the defaults for basic layout of individual items is unlikely to change much. And for whatever few corner cases we do improve, if Goldberg happens to include manual adjustments made to work around those corner cases, we update the score. It's not that hard.
Also see this: #179761: Problems with Open Goldberg scores on MuseScore.com
I think the purpose of series 2 will be ensuring it has the elements/features to enable the input of key information, which can then be used as a means of improving the next (potentially the pinnacle of layout - see improvements ).
I'm sure it could be re-visited at a later date in a stable version of series 3, but for now my suggestion would be to remaster and publish in the upcoming 2.1 version (stable enough to refer/fallback). Could the app be re-released?
I've already extracted these and saved them in 2.1 (from about Feb 22, 2017) and need to know how to upload them since Open Goldberg isn't my account.
@mike320: I'd suggest emailing the files to support@musescore.com so Thomas can update them from the back end.
Automatically closed -- issue fixed for 2 weeks with no activity.
Automatically closed -- issue fixed for 2 weeks with no activity.
Automatically closed -- issue fixed for 2 weeks with no activity.
Automatically closed -- issue fixed for 2 weeks with no activity.
Automatically closed -- issue fixed for 2 weeks with no activity.
Automatically closed -- issue fixed for 2 weeks with no activity.
Automatically closed -- issue fixed for 2 weeks with no activity.