Part completely unlinked from score—corruption of some sort?
In the attached score (open with the latest nightly build), the Bb Trumpet 3 part is somehow completely unlinked from the score—i.e., changes made to one are not reflected in the other. The easiest way to confirm this is to delete everything in the main score, then scroll to the Bb Trumpet 3 part and find everything is still there. This specific part was added using a recent nightly, and the reorder parts feature was used.
| Attachment | Size |
|---|---|
| Excellence.mscz | 610.61 KB |
Comments
Hmm, if you can figure out how you did that, I could add a menu item to do the same, document it and call it a feature :-). Seriously, steps to reproduce would be nice.
In reply to Hmm, if you can figure out by Marc Sabatella
I'll warily investigate. In the meantime, is there a simple way to edit the MSCX code so that changes will be linked again? (Conversely, where in the file is the link broken? Identifying that could be helpful for adding the feature.)
In reply to I'll warily investigate. In by Isaac Weiss
Just a shot in the dark to help going back to the linked situation: what about deleting the now unlinked part (if it is still listed among the score parts) and generating it again? I assume that editing the .mscx would be a nightmare (links are indicated in the file by numerical indices).
And, incidentally, if a reliable way to arrive at this result can be found on demand, I would rate it a feature too!
In reply to Just a shot in the dark to by Miwarre
I'm aware of that option, but I was hoping for a different solution not only so as not to lose formatting work, but also because understanding this seems essential. If links are indicated in the file by numerical indices, what's different about those indices? Once that's understood, might it be possible to implement a way to similarly change to such indices in the future?
Is there anybody who can look into this and learn something from the file? I've found a different bug with reordering parts, but I haven't been able to reproduce this yet.
In reply to Is there anybody who can look by Isaac Weiss
Only thing I can tell that wasn't initially obvious is that this file appears to have been improted from an earlier version of MuseScore, perhaps 1.3 or even a nightly build from more than a couple of years ago? The chord symbol hard coded into the file doesn't really match what I'd expect of a 2.0 score. 1.3 had its own way of recorded part information in score. Perhaps the code dealing with reordering of parts doens't deal well with scores imported from 1.3?
In reply to Only thing I can tell that by Marc Sabatella
I'm actually pretty sure it was inittially created with 1.x. The chord symbols is indeed one indication, but the lack of a workTitle meta tag is another and the fact that some texts are in Time New Roman rather than FreeSerif is a third
In reply to Only thing I can tell that by Marc Sabatella
Indeed, though it's been a long, long time since that score was imported from 1.3 (maybe a year, going back to 2.0 Beta 2), and I'm fairly sure that among the vast number of changes made with MuseScore 2 since then I've deleted all linked parts at least once or twice. If the parts were all generated with MuseScore 2, could the score's earlier history make a difference? (It doesn't with #89791: Reordering parts with new part moves wrong part, and other problems with Parts dialog.)
Thank you very much for looking into this. Those of us who are helpless when it comes to programming rely on the help of those with more learning for every scrap of knowledge.
In reply to Indeed, though it's been a by Isaac Weiss
The earlier history could probably make a difference. Hard to say for sure. But it does seem quite plausible that a bug in a pre-release nightly build might have corrupted something about the part list such that deleting parts in a released version didn't do exactly the right thing and things went south from there. It's also quite possible there are more bugs in the reordering facility; I haven't played with it at all.