can't open score created with earlier development build

• Apr 12, 2018 - 07:43
Reported version
3.0
Priority
P3 - Low
Type
Functional
Severity
S4 - Minor
Status
active
Regression
No
Workaround
No
Project

This issue came up as part of the investigations for #124876: Lyrics: Underscore is too long and overlaps and #132801: Horizontal Scrolling Breaks, where sample scores for those issues don't open in current master (a819eb6) anymore with an error message reg. bad format.
In both cases MuseScore 2(.2.1) can open these files (after ignoring the message that these are for a new version)

1st file, Neusidler,_Buch_1.mscz, created with 3543170 (master branch, I'm not sure this information taken from the score is correct, as that commit is from before 2.0 came out), gives:
Neusidler,_Buch_1_error.png
The code in the mscx looks pretty innocent:

          <beginText>
            <text>1.</text>
            </beginText>

Here it can be seen, in 2.2.1, that this volta 1 spans 2 measure (measures 7 and 8) and as such overlaps with volta 2 (measure 8)
Neusidler,_Buch_1.png
Possibly due to the fact that it is an irregular measure?

2nd file, Dont stop.mscz, created with bcc4136 (master branch), gives
Dont stop error.png
Here too the code in the mscx looks pretty innocent:

          <beginText>
            <text>cresc.</text>
            </beginText>

Here it seems that maybe MuseScore doesn't like the combination of measure repeats with hairpins (in measure 30)?
Dont stop.png
Very strange that these hairpins have a beginText?!

While it is documented that newer versions of the development branch may not be read ofiles created with older versions of that branch, I think here something is really going wrong that wasn't intentional

Attachment Size
Neusidler,_Buch_1.mscz 9.96 KB
Dont stop.mscz 51.6 KB

Comments

That Fugue? Pretty different, that file had been created with MuseScore 1.1 and current master detects corruptions not detected by 2.x):
Measure 3, staff 1, voice 2 too long. Expected: 4/4; Found: 17/16
Measure 4, staff 1, voice 2 too long. Expected: 4/4; Found: 17/16
Measure 5, staff 1, voice 2 too long. Expected: 4/4; Found: 17/16
Measure 6, staff 1, voice 2 too long. Expected: 4/4; Found: 17/16
Measure 7, staff 1, voice 2 too long. Expected: 4/4; Found: 18/16
Measure 8, staff 1, voice 2 too long. Expected: 4/4; Found: 17/16
Measure 9, staff 1, voice 2 too long. Expected: 4/4; Found: 17/16
Measure 9, staff 2, voice 2 too long. Expected: 4/4; Found: 17/16
Measure 10, staff 1, voice 2 too long. Expected: 4/4; Found: 20/16
Measure 10, staff 2, voice 2 too long. Expected: 4/4; Found: 17/16
Measure 11, staff 1, voice 2 too long. Expected: 4/4; Found: 18/16
Measure 11, staff 2, voice 2 too long. Expected: 4/4; Found: 17/16
Measure 12, staff 1, voice 2 too long. Expected: 4/4; Found: 17/16
Measure 12, staff 2, voice 2 too long. Expected: 4/4; Found: 10/8
Measure 13, staff 1, voice 2 too long. Expected: 4/4; Found: 20/16
Measure 13, staff 2, voice 2 too long. Expected: 4/4; Found: 17/16
Measure 14, staff 1, voice 2 too long. Expected: 4/4; Found: 20/16
Measure 14, staff 2, voice 2 too long. Expected: 4/4; Found: 20/16
Measure 15, staff 1, voice 2 too long. Expected: 4/4; Found: 10/8
Measure 15, staff 2, voice 2 too long. Expected: 4/4; Found: 17/16
Measure 16, staff 1, voice 2 too long. Expected: 4/4; Found: 20/16
Measure 16, staff 2, voice 2 too long. Expected: 4/4; Found: 18/16
Measure 17, staff 1, voice 2 too long. Expected: 4/4; Found: 17/16
Measure 17, staff 2, voice 2 too long. Expected: 4/4; Found: 18/16
Measure 18, staff 1, voice 2 too long. Expected: 4/4; Found: 18/16
Measure 19, staff 1, voice 2 too long. Expected: 4/4; Found: 18/16
Measure 19, staff 2, voice 2 too long. Expected: 4/4; Found: 17/16
Measure 20, staff 1, voice 2 too long. Expected: 4/4; Found: 5/4
Measure 20, staff 2, voice 2 too long. Expected: 4/4; Found: 17/16
Measure 21, staff 1, voice 2 too long. Expected: 4/4; Found: 17/16
Measure 21, staff 2, voice 2 too long. Expected: 4/4; Found: 10/8
Measure 22, staff 1, voice 2 too long. Expected: 4/4; Found: 17/16
Measure 22, staff 2, voice 2 too long. Expected: 4/4; Found: 10/8
Measure 23, staff 1, voice 2 too long. Expected: 4/4; Found: 17/16
Measure 23, staff 2, voice 2 too long. Expected: 4/4; Found: 10/8
Measure 24, staff 1, voice 2 too long. Expected: 4/4; Found: 20/16
Measure 24, staff 2, voice 2 too long. Expected: 4/4; Found: 17/16
Measure 25, staff 1, voice 2 too long. Expected: 4/4; Found: 5/4
Measure 25, staff 2, voice 2 too long. Expected: 4/4; Found: 18/16
Measure 26, staff 1, voice 2 too long. Expected: 4/4; Found: 17/16
Measure 26, staff 2, voice 2 too long. Expected: 4/4; Found: 17/16
Measure 27, staff 1, voice 2 too long. Expected: 4/4; Found: 17/16
Measure 27, staff 2, voice 2 too long. Expected: 4/4; Found: 20/16
Measure 28, staff 1, voice 2 too long. Expected: 4/4; Found: 20/16
Measure 28, staff 2, voice 2 too long. Expected: 4/4; Found: 5/4
Measure 29, staff 1, voice 2 too long. Expected: 4/4; Found: 2/1
Measure 29, staff 2, voice 2 too long. Expected: 4/4; Found: 2/1
but then loads it. Yes, this too needs to get looked at, but it seems an entirely different issue.

master seems to move the 1st voice 1 16th in those measures to the end, compare how it looks in 2.2.1 (and 1.3 too):
Fugue-2.2.1_and_1.3.png
vs. master:
Fugue-master.png

Now reported in #271284: 1.1 score gets corrupted in current master

Attachment Size
Fugue.mscz 9.28 KB

Given that we have changed a fileformat and dropped a support of the old 3.0-dev format, what should we do about this? Fix a branch which is able to read that format or just not fix it?

Our 3.00 compatibility build still has the issue with reading those files

Instead of

          <beginText>
            <text>1.</text>
            </beginText>

it should now (?) be
<beginText>1.</beginText>
for voltas, I guess the same is true for hairpins.
just ignoring that extra text tag should solve the issue?

Severity S3 - Major S4 - Minor

Not sure it is really a major priority since it affects only developers... and testers... and old nightlies users :)

UPD: reread the issue. The initial request is about old scores created in stable release, so it may affect users, but not sure this is still a major.