Re: [Denemo-devel] Re: Delete destroys the selection

[Richard Shann]

On Sun, 2010-05-16 at 03:24 +0200, Nils Gey wrote:
> It works now. Unfortunatly I compiled before I read this mail, so I could not 
> check if Crtl-X was failing.
> I've uploaded a new Delete script which replaces the old command bound to the 
> "Delete"-key. It deletes empty measures and selections, too. Selections get 
> only deleted if the preferences switch "Apply to Selection" is on.
> 
> 1) I want to remember you, Richard, to enforce >1 measure selections to whole 
> measures only. There might be cases where people want only parts of measures 
> in the beginning and in the end, but I think the more stable behaviour, both 
> for scripts and users, is more important.
> 
I didn't know you had this idea. There are clear reasons why multi-staff
is always whole measures (the way it is represented internally prevents
anything else anyway). But no such constraint applies to single staff.
One of the very few uses I make of selections is when I have started
entering notes from the wrong part - I select and delete the wrong stuff
and paste it into the right part. It may well start mid-measure.

It would be trivial to provide a command to extend the selection to
whole measures - it is just a matter of setting the first object marked
to 0 and the last to G_INT_MAX. But I guess you need to explain why you
don't want to handle a selection that straddles a barline. Recall that
these are really just "chunks" of objects used for display, usually
corresponding with measures.

Richard

> 2) It is obviously a difference for anything related to copy, cut, delete and 
> paste if you select barlines or not, especially the first and last one. Is it 
> possible to show graphically that barlines are selected, too? 
> 
> Nils
> 
> On Sat, 15 May 2010 18:12:53 +0100
> Richard Shann <address@hidden> wrote:
> 
> > I have pushed fixes - can you check that Ctrl-X was failing to delete
> > one measure in multiple staff case?
> > Richard
> > 
> > On Sat, 2010-05-15 at 09:56 +0100, Richard Shann wrote:
> > > On Fri, 2010-05-14 at 20:39 +0200, Nils Gey wrote:
> > > > 1.My last checkin was correct. I checked again and it is just what I 
> > > > added, a oneliner of code, which works, and one line of comments for 
> > > > the visual aspect in denemo.scm. 
> > > > 
> > > sorry, my mistake
> > > > 2.Thank you for d-DeleteSelection. But there is still something wrong 
> > > > and I think different things than you mentioned:
> > > > Sometimes the deletion does not work, it is dependent how the selection 
> > > > was created. Try to:
> > > > a)Select with the mouse, especially from right to left
> > > > b)shift+end 
> > > > There seems to be a "random factor" in this bug, meaning one I cannot 
> > > > reproduce at will. Sometimes it works, sometimes not. The more I try in 
> > > > one session the more it gets broken.
> > > > 
> > > I think I have seen this too,  but I can't find a pattern
> > > > 3. Is deleteSelection supposed to delete the measures, too, or leave 
> > > > empty measure behind? 
> > > > I cannot tell because of the bug mentioned, but I hope it is supposed 
> > > > to delete the measures, too.
> > > > 
> > > it will be whatever d-Cut is but without altering the clipboard.
> > > 
> > > ("will" here does not mean future tense, I just mean that the code of
> > > d-DeleteSelection is the last bit of d-Cut without the copy to clipboard
> > > bit that d-Cut does before deleting the selection.)
> > > 
> > > Richard
> > > 
> > > 
> > > 
> > > > 
> > > > Nils
> > > > 
> > > > 
> > > > On Fri, 14 May 2010 14:08:39 +0100
> > > > Richard Shann <address@hidden> wrote:
> > > > 
> > > > > On Fri, 2010-05-14 at 13:29 +0200, Nils wrote:
> > > > > > I think it doesn't matter: Choose whatever is easier internally, 
> > > > > > optional d-deleteObject parameter or d-DeleteSelection. 
> > > > > > Well.. just as I write this i like "d-DeleteSelection"
> > > > > I have implemented this.
> > > > > Two things:
> > > > >       There is some strange behavior around selecting empty measures. 
> > > > > In some
> > > > > circumstances this seems to "revive" old clipboards, or some such. I
> > > > > haven't investigated, but trying to test d-DeleteSelection lead me to
> > > > > try and remove the contents of the clipboard and I noticed that doing
> > > > > Ctrl-C with no selection visible did not work (?).
> > > > >       Nils last check in seems in error, there is only a commented 
> > > > > out proc
> > > > > def and a stub added to denemo.scm (?)
> > > > > 
> > > > > Richard
> > > > > 
> > > > > 
> > > > > >  better.
> > > > > > For the user it makes no difference in the end.
> > > > > > 
> > > > > > Nils
> > > > > > 
> > > > > > 
> > > > > > On Fri, 14 May 2010 09:10:10 +0100
> > > > > > Richard Shann <address@hidden> wrote:
> > > > > > 
> > > > > > > Have I understood you aright? What shall we call the command? Of 
> > > > > > > should
> > > > > > > it just be a parameter to the d-Delete command? (We need a 
> > > > > > > complete set
> > > > > > > of commands sketching out; but as we go on, we get a better idea 
> > > > > > > what
> > > > > > > they would be...)
> > > > > > > 
> > > > > > > Richard
> > > > > > > 
> > > > > > > 
> > > > > 
> > > 
> > > 
> > > _______________________________________________
> > > Denemo-devel mailing list
> > > address@hidden
> > > http://lists.gnu.org/mailman/listinfo/denemo-devel
> >