Preparing the release of MuseScore 1.3

classic Classic list List threaded Threaded
10 messages Options
Reply | Threaded
Open this post in threaded view
|

Preparing the release of MuseScore 1.3

lasconic
Administrator
Hi,

A version on the 1.X branch was not meant to happen but "Only fools don't change their minds" (french proverb)

So why making MuseScore 1.3 and what should be in.

* A very nasty bug causing ugly playback after a couple of measures AND we have a fix for it. http://musescore.org/en/node/19148
Going through the forum on a daily basis, I encountered a question about this bug several times per week.

* Another nasty bugs with tuplet handling. MuseScore fails to open some files with invalid tuplets. There is already a fix in the SVN and I answered several emails and forum posts "My file can't be opened anymore" by just loading the file in a fixed version of MuseScore and save it again.
See http://musescore.org/en/node/19357 for example. 80% of the "I can't open the file" problems are due to this bug.

* Updated manuals, translations

* Nothing more. For almost a year, all major development is happening on what is to become MuseScore 2.0, it's not worth spending a lot of development effort on this intermediary release.

I will take on me the packaging on Mac OS X and on Windows. We got several requests to have a MSI package on Windows, so the Windows installer will also be provided as a MSI.


Lasconic

------------------------------------------------------------------------------
Free Next-Gen Firewall Hardware Offer
Buy your Sophos next-gen firewall before the end March 2013
and get the hardware for free! Learn more.
http://p.sf.net/sfu/sophos-d2d-feb
_______________________________________________
Mscore-developer mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/mscore-developer
Reply | Threaded
Open this post in threaded view
|

Re: Preparing the release of MuseScore 1.3

Marc Sabatella
On 2/15/2013 9:09 AM, Lasconic wrote:
> * Another nasty bugs with tuplet handling. MuseScore fails to open
> some files with invalid tuplets. There is already a fix in the SVN and
> I answered several emails and forum posts "My file can't be opened
> anymore" by just loading the file in a fixed version of MuseScore and
> save it again.
> See http://musescore.org/en/node/19357 for example. 80% of the "I
> can't open the file" problems are due to this bug.

I didn't realize there was a single known cause for those; wonderful to
see that and the playback issue addressed.

So, I realize you didn't throw this open for suggestions, but I'm not
going to let that stop me from making a couple :-)

I absolutely agree there is no reason to spend a lot time on this
branch.  The biggest and easiest-to-fix showstoppers only.  For this
reason, you might consider whether it makes more sense to call this
1.2.1 rather than 1.3.  Not that it matters much, but I think 1.2.1 sets
expectations better.

Now, I really don't know if there is *any* chance of other fixes
sneaking in, but off the top of my head, the two that I would nominate
as most worth looking at are:

- bug on Windows where you can't save a file if title has a colon

- bug on Mac where saving as PDF just doesn't work in some cases (not
sure to what extent this is fully understood)

I gather these are probably really Qt issues that are only be truly
solved by recompiling with a different version of Qt then was used for
1.2, and that probably introduces too much risk.  But what about adding
code to strip out colons from the title before trying to open the file
dialog?  And if the Mac / PDF issue is understood to the point where the
code can know when it is going to fail, what popping up a dialog letting
the user know this isn't supported and describing a workaround?

These strike me as the most frustrating known points of failure besides
the two already being addressed.

OK, one more: crash on generating parts if voltas are used on staves
other than the first.

What these have in common is a situation where a user might have
invested a significant amount of work into a piece and then appear to be
dead in the water unable to continue, where potentially a simple fix
would prevent that.

Marc


------------------------------------------------------------------------------
Free Next-Gen Firewall Hardware Offer
Buy your Sophos next-gen firewall before the end March 2013
and get the hardware for free! Learn more.
http://p.sf.net/sfu/sophos-d2d-feb
_______________________________________________
Mscore-developer mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/mscore-developer
Reply | Threaded
Open this post in threaded view
|

Re: Preparing the release of MuseScore 1.3

lasconic
Administrator
Hi Marc,

Thanks for your suggestion.

The bug with colon on Windows is fixed. Same for the volta bug with part extraction.

For the bug on Mac, we don't have any clue and I can't reproduce on any mac I tried...
Qt will be updated to 4.8.4 though. It's just 4 bug fixing release from Qt 4.8.0, the one used in MuseScore 1.2.

Regarding the version number, I like them short since the whole 0.9.6.X serie :)

lasconic

2013/2/15 Marc Sabatella <[hidden email]>
On 2/15/2013 9:09 AM, Lasconic wrote:
> * Another nasty bugs with tuplet handling. MuseScore fails to open
> some files with invalid tuplets. There is already a fix in the SVN and
> I answered several emails and forum posts "My file can't be opened
> anymore" by just loading the file in a fixed version of MuseScore and
> save it again.
> See http://musescore.org/en/node/19357 for example. 80% of the "I
> can't open the file" problems are due to this bug.

I didn't realize there was a single known cause for those; wonderful to
see that and the playback issue addressed.

So, I realize you didn't throw this open for suggestions, but I'm not
going to let that stop me from making a couple :-)

I absolutely agree there is no reason to spend a lot time on this
branch.  The biggest and easiest-to-fix showstoppers only.  For this
reason, you might consider whether it makes more sense to call this
1.2.1 rather than 1.3.  Not that it matters much, but I think 1.2.1 sets
expectations better.

Now, I really don't know if there is *any* chance of other fixes
sneaking in, but off the top of my head, the two that I would nominate
as most worth looking at are:

- bug on Windows where you can't save a file if title has a colon

- bug on Mac where saving as PDF just doesn't work in some cases (not
sure to what extent this is fully understood)

I gather these are probably really Qt issues that are only be truly
solved by recompiling with a different version of Qt then was used for
1.2, and that probably introduces too much risk.  But what about adding
code to strip out colons from the title before trying to open the file
dialog?  And if the Mac / PDF issue is understood to the point where the
code can know when it is going to fail, what popping up a dialog letting
the user know this isn't supported and describing a workaround?

These strike me as the most frustrating known points of failure besides
the two already being addressed.

OK, one more: crash on generating parts if voltas are used on staves
other than the first.

What these have in common is a situation where a user might have
invested a significant amount of work into a piece and then appear to be
dead in the water unable to continue, where potentially a simple fix
would prevent that.

Marc


------------------------------------------------------------------------------
Free Next-Gen Firewall Hardware Offer
Buy your Sophos next-gen firewall before the end March 2013
and get the hardware for free! Learn more.
http://p.sf.net/sfu/sophos-d2d-feb
_______________________________________________
Mscore-developer mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/mscore-developer


------------------------------------------------------------------------------
The Go Parallel Website, sponsored by Intel - in partnership with Geeknet,
is your hub for all things parallel software development, from weekly thought
leadership blogs to news, videos, case studies, tutorials, tech docs,
whitepapers, evaluation guides, and opinion stories. Check out the most
recent posts - join the conversation now. http://goparallel.sourceforge.net/
_______________________________________________
Mscore-developer mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/mscore-developer
Reply | Threaded
Open this post in threaded view
|

Re: Preparing the release of MuseScore 1.3

Marc Sabatella
On Feb 16, 2013, at 7:01 AM, Lasconic <[hidden email]> wrote:

> The bug with colon on Windows is fixed. Same for the volta bug with part extraction.

Fantastic, thanks for considering my suggestions.  And FWIW, I think that knocking out these out helps justify the 1.3 number.

I made it a point not to actually search the forums or issue tracker to come up with my suggestions but to limit it to things that have come up often enough or seemed serious enough to have actually stuck in my head.

With another day to think about it, the only other issues that come to mind are any known actions that cause score corruption.  But I can't recall the details of what is reproducible and fully understood and what is or is not already fixed.  Seems there was something about copy paste involving a multivoice context, and copying empty 5/4 or 7/4 bars.  Seems we talked about fixing the latter right before 1.2 but it didn't happen; maybe it turned out to be be too hard to deal with on the branch.  Here again, even a dialog that pops up saying to retyr tpwith no empty measures selected could be an improvement.  But then, realistically, not that many people wrie in 5 or 7.

Marc


------------------------------------------------------------------------------
The Go Parallel Website, sponsored by Intel - in partnership with Geeknet,
is your hub for all things parallel software development, from weekly thought
leadership blogs to news, videos, case studies, tutorials, tech docs,
whitepapers, evaluation guides, and opinion stories. Check out the most
recent posts - join the conversation now. http://goparallel.sourceforge.net/
_______________________________________________
Mscore-developer mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/mscore-developer
Reply | Threaded
Open this post in threaded view
|

Re: Preparing the release of MuseScore 1.3

jorgk3
In reply to this post by lasconic
I have another wish:

http://musescore.org/en/node/18728
With a few easy steps you can corrupt a score beyond repair.

You don't need to write 5/4 or 7/4 to do it, it happens in 4/4 ;-)
Reply | Threaded
Open this post in threaded view
|

Re: Preparing the release of MuseScore 1.3

Tony Mountifield
In reply to this post by Marc Sabatella
In article <[hidden email]>,
Marc Sabatella <[hidden email]> wrote:
>
> With another day to think about it, the only other issues that come to mind are any known actions that cause score
> corruption.  But I can't recall the details of what is reproducible and fully understood and what is or is not already
> fixed.  Seems there was something about copy paste involving a multivoice context, and copying empty 5/4 or 7/4 bars.
> Seems we talked about fixing the latter right before 1.2 but it didn't happen; maybe it turned out to be be too hard to
> deal with on the branch.  Here again, even a dialog that pops up saying to retyr tpwith no empty measures selected could
> be an improvement.  But then, realistically, not that many people wrie in 5 or 7.

Marc, don't you ever transcribe any Dave Brubeck? :)

Tony
--
Tony Mountifield
Work: [hidden email] - http://www.softins.co.uk
Play: [hidden email] - http://tony.mountifield.org

------------------------------------------------------------------------------
The Go Parallel Website, sponsored by Intel - in partnership with Geeknet,
is your hub for all things parallel software development, from weekly thought
leadership blogs to news, videos, case studies, tutorials, tech docs,
whitepapers, evaluation guides, and opinion stories. Check out the most
recent posts - join the conversation now. http://goparallel.sourceforge.net/
_______________________________________________
Mscore-developer mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/mscore-developer
Reply | Threaded
Open this post in threaded view
|

Re: Preparing the release of MuseScore 1.3

lasconic
Administrator
In reply to this post by jorgk3
As discussed in the bug report, it's not a corruption since you can get the rest back by exchanging the voices twice.

Thank you for your suggestion though.

lasconic

2013/2/16 jorgk3 <[hidden email]>
I have another wish:

http://musescore.org/en/node/18728
With a few easy steps you can corrupt a score beyond repair.

You don't need to write 5/4 or 7/4 to do it, it happens in 4/4 ;-)




--
View this message in context: http://dev-list.musescore.org/Preparing-the-release-of-MuseScore-1-3-tp7577786p7577805.html
Sent from the MuseScore Developer mailing list archive at Nabble.com.

------------------------------------------------------------------------------
The Go Parallel Website, sponsored by Intel - in partnership with Geeknet,
is your hub for all things parallel software development, from weekly thought
leadership blogs to news, videos, case studies, tutorials, tech docs,
whitepapers, evaluation guides, and opinion stories. Check out the most
recent posts - join the conversation now. http://goparallel.sourceforge.net/
_______________________________________________
Mscore-developer mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/mscore-developer


------------------------------------------------------------------------------
The Go Parallel Website, sponsored by Intel - in partnership with Geeknet,
is your hub for all things parallel software development, from weekly thought
leadership blogs to news, videos, case studies, tutorials, tech docs,
whitepapers, evaluation guides, and opinion stories. Check out the most
recent posts - join the conversation now. http://goparallel.sourceforge.net/
_______________________________________________
Mscore-developer mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/mscore-developer
Reply | Threaded
Open this post in threaded view
|

Re: Preparing the release of MuseScore 1.3

jorgk3
Hi,

I think you got confused by the long-winded bug report.

This is corruption of the finest quality.

Add four notes in voice two in the first bar, then delete the last three. Delete the resulting rests. Add four notes in bar 2. This is the result:

Now select the remaining note in voice 2 in bar 1 and change it to a minim (1/2). Result is:


It's easy to find many variations on the theme, for example by changing the note to a semibreve. Also try doing something to the remaining notes in bar 2. The second bar appears to be "shortened" in the second voice, you can no longer place four quarter notes there.

How much more corrupt would you like it to be?

Jörg.
Reply | Threaded
Open this post in threaded view
|

Re: Preparing the release of MuseScore 1.3

jorgk3
I different example was mentioned in the bug report:
http://musescore.org/en/node/18728#comment-68097
Reply | Threaded
Open this post in threaded view
|

Re: Preparing the release of MuseScore 1.3

lasconic
Administrator
In reply to this post by jorgk3
And we go again... It's discussed extensively in the bug report...

Select the two measures and press delete, and everything is fine. In my book, corruption is "there is nothing a user can do to fix the score".
The behavior you describe is nasty, shouldn't happen and it is a bug but it's not a corruption. In any case, the behavior is too much anchored in MuseScore 1.2 to make a safe fix for 1.3.
If you want to make a fix, your patch is welcome but for MuseScore 2.0.

Case closed for 1.3.

lasconic

2013/2/16 jorgk3 <[hidden email]>
Hi,

I think you got confused by the long-winded bug report.

This is corruption of the finest quality.

Add four notes in voice two in the first bar, then delete the last three.
Delete the resulting rests. Add four notes in bar 2. This is the result:
<http://dev-list.musescore.org/file/n7577808/delete-rest-5.png>
Now select the remaining note in voice 2 in bar 1 and change it to a minim
(1/2). Result is:
<http://dev-list.musescore.org/file/n7577808/delete-rest-6.png>

It's easy to find many variations on the theme, for example by changing the
note to a semibreve. Also try doing something to the remaining notes in bar
2. The second bar appears to be "shortened" in the second voice, you can no
longer place four quarter notes there.

How much more corrupt would you like it to be?

Jörg.



--
View this message in context: http://dev-list.musescore.org/Preparing-the-release-of-MuseScore-1-3-tp7577786p7577808.html
Sent from the MuseScore Developer mailing list archive at Nabble.com.

------------------------------------------------------------------------------
The Go Parallel Website, sponsored by Intel - in partnership with Geeknet,
is your hub for all things parallel software development, from weekly thought
leadership blogs to news, videos, case studies, tutorials, tech docs,
whitepapers, evaluation guides, and opinion stories. Check out the most
recent posts - join the conversation now. http://goparallel.sourceforge.net/
_______________________________________________
Mscore-developer mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/mscore-developer


------------------------------------------------------------------------------
The Go Parallel Website, sponsored by Intel - in partnership with Geeknet,
is your hub for all things parallel software development, from weekly thought
leadership blogs to news, videos, case studies, tutorials, tech docs,
whitepapers, evaluation guides, and opinion stories. Check out the most
recent posts - join the conversation now. http://goparallel.sourceforge.net/
_______________________________________________
Mscore-developer mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/mscore-developer