Renaming "timewise delete"

• Sep 30, 2016 - 03:38

In the nightly builds currently, Ctrl+Delete (Cmd+Delete on Mac) will delete a note or rest and shorten a measure; delete a whole measure; or delete a barline and merge measures. This function is named "Timewise delete," though it appears nowhere visibly yet.

Though "timewise" appears to be a legitimate word, it's awkward, rather like "bigly." Thoughts for alternatives? "Hard delete" is what I'm thinking of right now.

Note also that I'm adding it to the Edit menu and relevant context menus in an upcoming PR.


Comments

Sorry to hijack the thread, but what do you think about the whole thing?

Cmd + delete will delete a note or rest and shorten a measure; delete a whole measure; or delete a barline and merge measures.

I think it's great that we have the possibility to add/delete time and have the measure grow. It could open the door to a specific "scratchpad" view for example. It's also great that it uses the same code.
But, from a user perspective, I really wonder if it makes sense to have the same command for all:

1. Deleting one or more measures is something particular and it should be easier to do. I believe it should be a first class class citizen in the menu, like it was in MuseScore 2. It should also be in the contextual menu when one or more measures are selected.

2. Inserting/deleting notes and having the measure shrink/grow. Currently, it has some limitation, for example it's not possible to insert notes with the keyboard, you need to use ctrl + click. I wonder if it wouldn't be better to have an insert mode, like we have a repitch mode. The cursor shape could change and notes would be inserted. I'm not a huge fan of modes in general but here it seems to fit.

3. Deleting a barline. Again, from a user perspective, this is different. It would be great if it could share the same shortcut than delete measures but I believe it should be a separated command from a user point of view.

In reply to by lasconic

You can insert notes with keyboard: use shift+ctrl+notename.

Deleting measures as a special command is problematic as there is no way to select a measure. You can click on a measure and it looks like it is selected but what you really get is a selected time range which happens to span a measure.
More general i dont't like special commands which hide more generic ones as they give a wrong model about the structure of the program. To implement "easy to use" for me means to find a minimal set of commands/concepts to achieve the desired functionality.

What about "Resect", "Resection" tool. Implies removal of something and joining up the remainder. Translates into several common languages:

réséquer, resecare, resecar, ressecar, resezieren, resekera
résection, resezione, resección, resseção, resektion, reseksjon

In reply to by Isaac Weiss

"Delete section" is confusing since we are talking about a tool here that can be applied to a measure, a single note, a barline etc. to not only erase it and leave a blank space behind but to remove it entirely (like a tumour). I say we steal the medical word - since it isn't being used for other common things it is less likely to to be confused with other terms. It's a form of "cut" but distinct from "cut and paste". The word "section" comes from the same root as resect but what used to be a verb (to section means to cut) got turned into a noun (a section being a cut piece).

I didn't know if this was an issue or not, so I didn't want to formally address it yet, but I noticed that in nightly build 2016-09-30-1931-master-70b00e6 on Windows 10, I was unable to remove the last note or rest (of any value) of a measure with the "timewise delete". Is this by design? I am able to delete from any other point in the measure just fine; just not the last note/rest.

~Happy trails, Flyingninja77

In reply to by ericfontainejazz

But it works for single notes anywhere else in the measure, even if you select a single note when there are more notes in the chord. E.g.:
timewise-del (1).PNG
becomes
timewise-del (2).PNG
However,
timewise-del (3).PNG
doesn't change after the so-called "timewise delete" unless, as you said, you select the entire chord. This seems like unexpected behavior to me.

Happy trails, Flyingninja77

Attachment Size
timewise-del (1).PNG 11.63 KB
timewise-del (2).PNG 10.99 KB
timewise-del (3).PNG 12.45 KB

In reply to by Isaac Weiss

Apparently it only removes barlines of measures enclosed completely in the selection.

For example, the following selection:
no-full-measure-selection.png

will after timewise delete becomes:
after-time-wise-delete.png

I can't imagine any easy description for what this does...the best I can come up with is:
"Remove segments of partially-selected measures, remove the duration of those segments from the actual duration of those partially-selected measures, and remove entire measures (including barlines) of completely selected measures"

The previous Ctrl->Del function is much easier to describe: "Delete Measures Containing Selection"

I object to having a different function for Ctrl->Del because the behavior of the keystroke Ctrl->Del has changed when only part of a measure was selected. Most users have keystroke behavior memorized, so shouldn't have default behavior of keys change between versions. The only thing that should be changed about Ctrl->Del is that it's description should be changed to the unambiguous "Delete Measures Containing Selection", since that is what it does, instead of the previous slightly less clear "Delete Selected Measures". I find this function useful since don't have spend extra effort to select the entire measures first.

Please use a different keystroke other than Ctrl->Del for this timewise function. Alt->Del and Ctrl->Shift->Del are available for this timewise function. At the very minimum, musescore should still have an action for "Delete Measures Containing Selection" in the Preferences->Shortcut list, so user can configure however want.

(Separate issue is name "Timewise delete", which does not unambiguously describe what the functionality is.)

In reply to by ericfontainejazz

If you select a measure, the time wise delete does the same thing as before the change. Only now you can also delete a barline or partial measure or a selection that doesn't start or end on a measure boundary. So there funtionality has been increased, just the name is not optimal

In reply to by Jojo-Schmitz

My two cents: I would say to include both, provided it doesn't blow the whole system, and allow the user to set his preferred delete/remove command to whatever shortcut he wants. Now, as to which one would be the default, that is another point worth discussion.

I don't know how much I would end up using the timewise delete; maybe a lot, maybe a little. As a MS 2.0 user, I have gotten by just fine copy-pasting around the undesired note, splitting the barline as appropriate (Shift+\; don't remember if that was the default), and then deleting the resultant empty measure. This method can be a little slow, and the timewise delete would certainly speed it up, but I don't generally remove a note from the middle of a measure that often. I'm sure others do more than me, or we wouldn't be having this discussion in the first place.

I'm sure that many users would be benefitted by the timewise deletion command, but a case can be made that most inexperienced users would benefit by having the "Delete measures containing selection" command as the default (this will also help those users switching from 2.x to 3.0). As those users gain familiarity with the program, they can then explore the rich depth of customization that MS provides them, and assign the timewise delete to a more natural command if they feel the need.

Arguments could be made both ways. These are just my thoughts.

~Happy trails, Flyingninja77

tl;dr: I think that both are useful in different situations.

In reply to by ericfontainejazz

But why would anyone go out of their way to select only a partial measure then try to use a measure deletion function? I think you are worried about a case that essentially never happens. And if the behavior changes in those rare exceptions, so what? It's one way people will become aware of the new functionality. A new major version of a program is pretty much *expected* to change some things, hopefully for better. And in every way I can see, this *is* for the better, even if there could still be some tweaks to make it better still. So those few people who occasionally for some reason select only a partial measure before deleting will need to adjust to selecting a full measure. Since that's easier anyhow, I can't imagine anyone complaining.

In reply to by Marc Sabatella

Several random thoughts:

1. Timewise Delete is a particularly poor functional descriptive. It doesn't make much sense in English, and I can't even imagine how to translate it meaningfully into French (my second working language): 'Supprimer Temporalement' makes no more sense in French than 'Timewise Delete' does in English. And just for gits and shiggles, here's how Google translates that phrase into some of the other languages in which MuseScore is available: In German, 'vom Zeitbedarf gelöscht werden'. In Dutch, 'tijdsopzicht schrappen'. In Spanish, 'en cuanto al tiempo Eliminar cualquiera'. In Japanese, '削除時間的'. Do I need to go on...?

As an alternative, I would suggest Delete Selected Beats which is specifically descriptive of what the command actually does.

2. It seems important to me that this new command must be subject to a subsequent 'UNDO' command. I would presume that it would be structured that way, but in case I'm wrong about that, I thought I had better mention it. This command, which would remove individual beats and modify the overall structure of measures or whole sections of a piece of music, will almost certainly be used in error until people get used to it, and if it can't be undone by a CTL+Z command, a lot of people will be gnashing their teeth and muttering heinous imprecations about the developers under their breath....

3. Continuity of major commands is important in the evolution of any program, so changing what happens when one uses a well-established command such as DEL or CTL+DEL is a bad idea unless there is an overriding reason to do so. I don't see one from where I sit. In the current stable versions, these are two distinct 'delete' commands which users have become accustomed to.

The first command, DEL, replaces all the notes in the selection with rests, but does not delete the actual measure nor any of the acutal beats. This offers the user the opportunity to start over on a measure which he did not write to his satisfaction the first time. It happens to the best of us, and is thus a common use-case the program should continue to support.

The second command, CTL+DEL deletes all the measures included in the range selection the user has made before typing the command. This is useful when a user has written measures which need to be removed entirely from the score, for whatever musical reasons might apply. It is also particularly useful when a measure has become corrupted somehow and needs to be deleted and replaced with a 'clean' measure. The user must make a range selection prior to typing the command, and in the event that range selection contains portions of a measure, the entire measure will be deleted. Users understand this; to change it would be a bad idea, IMO.

The proposed new command, 'Timewise delete' would do something different, as I understand it from reading this thread, so IMO, a different keyboard shortcut should be used to invoke it. Instead of deleting the whole measures contained (even partially) in the range selection, 'Timewise Delete' would remove the actual beats within that range selection (or even the beats contained in a list selection? Is that correct??) and in so doing, it would modify the duration of the measures involved. So it is potentially useful, but only for those users who need to depart from the established durational structure for a short section of their score. This could be extremely valuable for producing notation for textbooks and method books, to say the least, so there appears to be a reasonable use-case for including it in 3.0. However, as stated above, the keyboard shortcut to invoke it should NOT be the same as the existing 'delete notes' or 'delete measure' commands, and a better descriptive term should be found for it.

In reply to by Recorder485

To relieve some of your fears, the current implementation of timewise delete includes the ability to undo this action as well.

I think that Delete Selected Beats is a better approximation of what we want it to describe than "Timewise Delete," but to me it still feels like it is not emphasized enough that the measure duration is changed in value. Here , underquark suggested "Resect," which I very much like, but it is certainly not an intuitive term. It is unfortunate that we haven't found another such succinct and descriptive term with which more people would be familiar. I personally like Remove Selected Beats from Measure(s), but to many, that may not be any different from yours.

~Happy trails, Flyingninja77

In reply to by flyingninja77

Thank you.

I think that part of the problem with naming this command is that it does more than one thing, so any truly descriptive name would necessarily require more than one verb clause, making it too long and complex for easy comprehension. 'Remove selected beats and reset measure duration' would be the shortest descriptive I can think of at present, but even that does not comprise the idea that barlines can be eliminated and former distinct measures combined into one.

In that this command takes us fairly into 'scratch-pad mode' territory--that is to say, gives us the ability to produce what is in essence unmeasured notation--I am wondering if it should function only if the user has specifically opted to work in unmeasured mode.

This ability is something I have argued in favour of more than once, but I am not sure that this is the best way to provide that. It would be far simpler and much more intuitive, IMO, if the user were given a command which put him into ummeasured notation mode, and once in that mode, he would (among other things) be able to use this rather 'wibbly-wobbly, timey-wimey' mega-delete command to 'exterminate' any single beat (or group of beats) completely. ;o)

In reply to by Recorder485

I would submit that "delete measure containing selection" is and always has been a very strange and inherently non-intuitive idea for a command. Who in their right mind would expect that they could select a range starting, say, on the "and" of beat 2 in one measure and ending on beat 3 of the next, and then hit some magic command that would result in both those measure being completely removed? If you think about it, that makes no sense at all really. Just because some users are accustomed to doing something a certain way doesn't mean it really makes sense, and this to me is a prime example. The behavior of the current 2.0.3 Ctrl+Del command looks, frankly, like a bug - a command that operates not on the selection itself but on something else. Bad design. We don't see it that way because we're used to it, but I think to the new user, the 3.0 behavior will look a *lot* more logicial.

i think all this worry about people doing that is unwarranted. Yes, in certain rare cases there might be some reason that some people have relied ion the current behavior where MuseScore deletes something other than the selection. But I think it's a tiny minority who have even noticed that this is the case, and I think they can adjust to the new more logical behavior.

In reply to by Marc Sabatella

it is not going out of the way...actually often it is harder to select the entire measure because you have to carefully make your selection by clicking from very first chordrest of a measure to very last chordrest of a measure, or have to be able to find somewhere on the staff where there aren't notes (can be hard in crowded measures) to select the entire measure.

Re discoverability, sure people will discover this different behavior, but a lot of people who are used to prior behavior will think that there is a bug when Ctrl->Del doesn't behave the same as previously.

Also Delete Measures Containing Selection is a much cleaner cut and unlikely to mess up users score, if they don't realize that the measure durations have *changed* when user thought was just "deleting" things.

Like I said, I don't think it is a bad function, I actually think it is useful. But I don't think a different function should replace the same keystroke as a previous differently-behaving function. Regarding whether it is ok to change behavior of keystrokes on major version changes, I don't think so unless there is overwhelmingly good reason, which is not the case here. Most people have to upgrade to a new version, but shouldn't have to rewire their brains for keystrokes just to be current.

What's wrong with two keystrokes for two different functions?

All excellent points by Recorder.

In reply to by ericfontainejazz

I would agree there are occasions where it can be slightly hard to select a full measure, but why should "delete measure containing selection" be the one and only command that provides this magic "do what I mean, what I say" mode? (DWIMNWIS) Why not also have a DWIMNWIS version of the delete (as in, replace with rests) command? Or the "copy" command? Or make the transposition functions (eg, up/down arrow) also come with separate DWIMNWIS versions? Unfortunately, we have limited shortcuts available. Again, if you use the command the way it was designed to be used in the first place, the behavior has *not* changed. You only see a change if for some reason you choose to rely on a very oddball (and undocumented as far as I know) quirk of the command.

At some point, I think we should simply accept the old command was pretty illogical, the new one is just plain better, and if you were in the habit of relying on this oddball non-intuitive behavior, just make the adjustment. It's not *that* hard to select full measures, and if it were, maybe we should spend effort making that easier so *all* commands can benefit, not just this one.

All: Don't miss that you can also remove a barline with this command, merging the measures on each side.

Myself, I greatly disagree with the idea that this should be separated from the "remove measure" functionality that it includes.

In reply to by Isaac Weiss

This is one of the things that makes renaming timewise delete difficult: it can do a lot with a single function call. Most beginning users and users new to 3.0 will have to get used to its name and behavior, but if it is well-documented on the website (which, in time, I'm sure it will be), then most people will be fine. The main issue is making sure the above users can identify the general concept of this command by its name, to get an intuition of what it is and how it is used. There is no need to call it "Delete beats in selection, and if it happens to cross a barline and not include the beginning of the first measure and the end of the last measure, then merge them into a single measure."

Ultimately, I personally won't care whatever it ends up being called--I'll just use the shortcut anyway. Similarly, I probably won't care whether the original behavior is reimplemented as another command or not; I can already see how the current behavior is useful, and I'm sure I can adapt my behavior accordingly.

~Happy trails, Flyingninja77

In reply to by Marc Sabatella

I'm going to try to look at this from a non-mucisal perspective, because that's the way most new users, who are used to the operating principles of word-processing programs, DO look at things until they get used to the idea of a measure being an immutable durational entity.

We are stuck with the fact that the physical button on the keyboard is labeled 'delete' on almost all hardware. But 'delete' in a text-editing environment means essentially the same thing as 'erase', 'wipe-out', or 'remove'--get rid of the selected material completely!--while in a score-writing context, it can mean two different things, both of which I explained in my earlier post. That being said, it would probably have been better if MuseScore had originally chosen the word 'erase' to define the command executed by a single DEL key push, and reserved 'Delete' to describe what happens when one types CTL+DEL. But it didn't...so now we have to deal with that.

Users are fairly accustomed to CTL+[x] commands 'upping the ante' by one notch, so that isn't a major issue, I think. The original commands can, and should, stand as they are.

The new command, the one provisionally labelled 'Timewise Delete' needs two things:

1. A better descriptive name
2. A different shortcut than either DEL or CTL+DEL.

'Remove selected beats' could be the best answer to #1, although a nasty little devil is whispering in my ear that it should be 'COMPLETELY remove selected beats' (I hate that!!). Question: could we provide a popup 'are-you-sure?' warning when that command is invoked, something along the lines of 'This command will remove the selected beats from the measure(s) and change the underlying measure duration accordingly.'?

As to the shortcut, by upping the ante by one more computer-typical notch, we could use CTL+SHIFT+DEL, if that doesn't conflict with any overriding OS shortcuts.

In reply to by Isaac Weiss

There is no need for SHIFT+CTL+DEL to create full-measure rests; simply select the material and type DEL all by itself. (Unless that has changed since 2.0.1...?)

As to the rest, the thought occurs to me that if a command does so many things that it is impossible to define succinctly in plain langauge, perhaps the structure of the command itself is the problem.

Again, I am not AT ALL against providing the ability to depart from the rigid durational structure of the program for producing such things as illustrations in theory or method books; but I am not sure that this is the best way to approach that. Werner states that he is not a fan of modes in general, but in this case a 'non-counting' mode may indeed be the best answer. Why would any ordinary user need to delete beats and barlines absolutely, other than to create graphics-only examples for insertion to a text document? It would be far easier to do that if one could type a command to enter a scratch-pad or unmeasured entry mode.

In reply to by Recorder485

This works only for complete measures. For measures with actual duration different than nominal, regular delete is not supposed to create a measure rest (although older versions had a bug where it would). Ctrl+Shift+Delete is the command to force a full measure rest even in these situations.

In reply to by Marc Sabatella

You got me wondering, so I looked it up in Preferences, and you're correct that CTL+SHIFT+DELETE is the default shortcut for "Full Measure Rest". Funnily enough, however, the command does not work except on irregular measures. For normal measures, typing that combination on a range selection did absolutely nothing. So it's a good thing a plain old garden-variety DEL does the job of putting a whole-measure rest into a score, even if it isn't getting 'screen credit' for its labours. ;o)

I did some test-driving to find out what I'd been missing. As it turns out, not much, but there are differences between the two commands worth noting.

1. Typing DEL on an irregular-measure selection replaces all the elements in the selection--including any notes--with appropriate durational rests, as expected. It does not create full-measure rests in the irregular measure (which would be incorrect notation in most cases, anyway).

2. OTOH, for those few cases where F-M rests are wanted even in irregular measures, typing CTL+SHIFT+DEL will generate them, and do so without erasing any notes within the selection. So, in the following selection, wherein the first measure has been made irregular (3 beats actual/4 beats nominal), this:

CTLSHIFTDEL 1.png

yields this on a CTL+SHIFT+DEL command:

CTLSHIFTDEL 2.png

...whereas it yields THIS on a simple DEL command:

CTLSHIFTDEL 3.png

As to DEL working to produce whole measure rests only for complete measures, I'm honestly scratching my head trying to understand why that even needs to be said. I must be missing something, but I can't help but ask where else other than in a complete measure would normal users want a full-measure rest? It really is poor notation to have F-M rests in such things as pick-up measures and split-measure voltas. Without the durational rests to tell the performer what the actual beat count is, he's just guessing.

As a matter of fact, aside from that one niche-case, DEL works just fine to insert F-M rests, no matter what one selects, in that it does exactly what one expects it to do: erase and replace. If I select a combination of full measures and partial ones, DEL erases all notation in them and generates full-measure rests for the full measures in the selection, and appropriate durational rests for the partial measures. It is able to figure this out when confronted with a change of time signatures, even when faced with several different local time signatures in the selection (note that first measure for both parts is in 4:4):

This:

DEL command 1.png

yields the following result on a DEL command:

DEL command 2.png

This, on the other hand:

DEL command 3.png

yields full-measure rests in every measure selected, because the entire measures were included in the selection:

DEL command 4.png

Anyway, sorry about the long virage, but you piqued my curiosity about a command I didn't know existed. I don't actually care what the new shortcut for the "Timewise Delete" command is; I suggested CTL+SHIFT+DEL because (a) I didn't know it was already assigned, and (b) it seemed the logical progression after CTL and CTL+DEL. The important thing is that the new command, whatever it is finally named (and we're pretty much in agreement on 'Remove Selected Beats' by now, I think, but I would like that popup warning), has a distinct shortcut to distinguish it from the two (now three, I learn) other commands using the DEL key either alone or in combination with a modifier key.

In reply to by Isaac Weiss

For completely empty measures, a single F-M rest in voice one suffices. Adding a second (or third, or fourth) full-measure rest to an empty measure is unnecessary unless something distinctly unusual is going on. For places where one or two voices contain notes while one or two others are silent for a measure or two, the need to show F-M rests (or indeed rests of any sort) for those silent voices is somewhat disputable. In general, multi-voice staves are already so crowded that whenever secondary-voice rests aren't absolutely necessary for clarity they are often made invisible.

There are exceptions, of course; I find a number of F-M rests in voices in the Schott/Weiner urtext edition of the Goldbergs, some written by Bach, others added by the editor in brackets.

------------------

I thought the consensus was settling on "Remove," without mentioning selected beats, because it will also remove a selected barline or measure.

As pointed out by ericfontainjazz earlier in this thread , the barline(s) will be removed and measures merged only when entire measures are contained in the selection, so again, the multiple different results that could be obtained by using the same command would make an absolutely precise descriptive name for this beastie sound like the title of a doctoral dissertation. I think we're going to have to break this into two parts.

What is wrong with using "Remove Selected Beats" and then giving the 'long version' in a pop-up warning?

In reply to by Recorder485

I'm not sure what use case there is for wanting a full measure rest in an irregular measure, but we didn't want anyone to complain it was impossible :-)

Anyhow, the crux of the discussion is whether there is any good reason to consider the new timewise delete to be a separate command from the old measure delete command. I have given my reasons why I think the answer is "no" - it's just a superset of the original, completely backwards compatible when used as intended.

The only argument I've seen for making it separate is that some hypothetical user might have accidentally the old command would delete full measures even with only a partial measure selected, may have chosen to rely on that bug/quirk, might be surprised by the new more logical behavior, and might for some reason be unwilling to change his habits to simply select full measures.

To me, that's not even close to a good reason to make this a separate command (thus requiring a separate shortcut, separate naming, separate documentation, etc).

In reply to by Marc Sabatella

I'm not sure what use case there is for wanting a full measure rest in an irregular measure, but we didn't want anyone to complain it was impossible :-)

Thanks, Marc--you just added a good laugh to my day! :)

As for the the rest, you've made me start to re-think if I've actually got nailed down all the intricate details of how and under what circumstances this new beastie would do what. IOW, is this really a superset containing all the various original 'delete-somehow-or-other' commands? If that's truly the case, there is less to worry about, I think.

I am not clear on what will happen when someone uses an 'old' command in its original context once this new superset command is enabled:

--Will DEL on a range selection of normal measures still 'erase and replace' everything in those measures with either F-M or durational rests? Or will it 'Remove Selected Beats' instead? If the latter, in what context will that happen?

--Will CTL+DEL still 'exterminate' selected measures? If so, will that rather wacky (I agree with you on this one) hiccup--the one that currently eradicates whole measures of which only parts have been selected--continue to function? Or will that be 'fixed' to provide a more logical behaviour? (I really don't see anyone depending on that one, although I definitely see people depending on the others.)

--Finally, if either the DEL or CTL+DEL shortcuts were re-assigned to this new 'Remove Selected Beats, etc.' beastie, what command would a user need to use to simply 'erase and replace' or 'exterminate'? If the command is context sensitive, it could conceivably do different things under different circumstances...but is it? It's really not clear.

In reply to by Recorder485

The behavior of Delete has not changed. It erases the contents of the selection, replacing with rests, including the use of full measure rests for regular empty measures.

The behavior of Ctrl+Delete has not changed in the usual case: select one or more full measures, Ctrl+Delete, and those measures are removed completely.

All that has changed is what happens if you do Ctrl+Delete with a partial measure selected (and thus the complementary partial measure *not* selected). Previously this first extended the selection to the full measure, then removed it - which I say is an illogical quirk. Now, the command simply removes the selected partial measure - thus leaving you with just the completentary partial measure (ie, an irregular measure), just as you asked for.

In reply to by Marc Sabatella

That is the clearest explanation of this new command that I have seen yet. Thank you, sir. On that basis, I can withdraw most of my objections and rest easy.

It's obvious this is complex enough to require a page in the Handbook, but that should be no problem. The potential confusion I see is that the same shortcut (CTL+DEL) would do two different things under differing circumstances, so in theory it should appear in two separate entries in the Preferences>Shortcuts page: One for "Remove Measure(s)" and another for "Remove Selected Beat(s)." That's less than ideal, IMO.

However, a simpler idea occurs to me, and that would be to list it only once--as "Remove Measure(s) or Beat(s)"--and then (and please bear with my continued lobbying for a pop-up) have a warning appear, but only when partial measures are selected:

"Warning: This command will remove all full measures in the selection and will also remove all selected beats from partial measures within the range box. The actual duration of those partial measures will be modified accordingly and some barlines may be removed as well. Do you wish to continue? [Yes] [No] [Don't show me this again.]"

What do you think? Does that cover all the necessary bases?

In reply to by Recorder485

I don't see it as two different things at all. It removes the selection, period. Doesn't matter if the selection is a full measure partial measure or any combination thereof - it is removed from the score.

I think warnings like the one you suggested make it seem far more complex than it is. We don't warn when you try to remove a selection that happens to be a full measure; why should we warn when you try to remove a selection that happens to be a partial measure? Either you intended to remove the selection or you didn't.

BTW, there is one other thing the new command does: if you select a barline, Ctrl+Delete removes it, thus performing the equivalent of the old "join" command. The old command did not do anything with just a barline selected - it required a range selection.

In reply to by Marc Sabatella

The functions are different because the result is different. In one case, the user is left with a score that contains only ordinary, full-count measures. In the other case, the user winds up with a score that contains a number of irregular measures that could potentially throw off everything else in his composition or arrangement. If I am composing a piece in 4:4, I do NOT want to find myself looking at a measure of 3:4 or 1:4 smack in the middle of my modulation to the relative minor.

I think there is a high probability that many users will try to use this 'Remove Beats' tool in the belief that it will simply 'move' the remaining music backwards to fill the deleted beats. Maybe they wrote a bum measure, with an extra beat out of place and just want to get rid of it. You or I would know enough to cut-and-paste onto the proper beat to solve the problem; most less experienced users will interpret 'remove beats' as a way to slide everything backwards by the number of beats removed, but they will expect the program to re-bar the rest of the music to maintain the nominal number of beats in their chosen time signature. That's not what will happen, and they will be confused (and probably highly annoyed as well).

As for a popup warning making things seem more complex than they are, I disagree. This IS a complex concept, given that it flies in the face of the basic durational structure of MuseScore. Heretofore, everything else in MuseScore has been based on the immutability of measures: You couldn't, in the normal way, delete a rest or a note, but now we're providing something which makes that possible. It's a useful concept--heck, how many times have I argued in favour of this ability?--but it changes how the user needs to think about what is happening...and a simple warning that 'Here be dragons, sirrah!' is not, IMO, out of place.

In reply to by Recorder485

The only reason you get an irregular measure is that the region you selected was irregular. It's not that the tool did something different different just because you selected a partial measure - it did what it always does. It's like a pair of scissors; it simply cuts what you ask it to. You don't say a pair of scissors cuts horizontally if you move it across the paper left to right but it cuts vertically if you move it top to bottom - just saying that sounds confusing and strange. it simply cuts what you ask it to, period. Acting like it does two different things just makes it seem confusing when it need not.

I do agree that it is conceivable a user will mistakenly try to use this feature instead of cut and paste to move things earlier. They'll learn that lesson immediately, but OK, if a warning saves them the trouble of noticing the problem and hitting Undo, that's OK. i'd just ask that a) the warning have a checkbox to not show again, and b) the warning not make the command sound more complex than it is. Something straightforward like "partial measure selected; irregular measure will result - continue?" would be much more accurate and far less confusing.

In reply to by Marc Sabatella

Agreed. I am well-known for saying things in more words than strictly necessary--old habit of a free-lance writer who is paid by the word!--you are better at being concise and to the point. (And I absolutely agree about having a 'don't show this to me again' checkbox; I included that in my original suggestion.)

In reply to by Marc Sabatella

For the record, I disagree with Marc and Werner on this one.
I believe deleting measures and deleting beats should be two different commands, for more or less the same reason than Recorder485. I understand that conceptually and in the code both features are the same but from a user perspective, deleting a measure and deleting time in a measure or spanning several measure is very different.

One of the reason why it's different: the complementary operation is different. If the user deleted one or more measures, he will look for a way to insert one of more measure. If the user deleted some beats, he will try to find a way to add more beats and it will be a different feature than the one to insert measure. He will have to use the measure properties dialog (which is not easy to find) or figure out how to insert notes in a measure (Shift+Control + note name or Ctrl + Click).

Personally, I would prefer to keep Ctrl + del to delete measure (we can discuss if it should work even if the full measure is not selected or not... but it's not really the place). And, then I would prefer to have an insert/delete mode (and you know, I hate modes as much as I hate message dialogs...), similar to the one in a text editor, maybe added to the repitch mode and others. In this mode, the cursor would change and the user would always delete beats or add notes when pressing Del or note names (or using the mouse). In this mode, we can imagine that the user could enter notes without barlines and then add the barlines later if he wants. IMO, having both insert and replace in the same mode, and delete notes/measure in the same mode, will confuse users.

In reply to by lasconic

Also for the record, I started out uncomfortable with this, but grew to like it. That said, the point out how different the complementary operations are is a good one. And I definitely agree an insert/delete mode would be better still. Trouble is, probably 90% of the people who would be looking for that would expect it not to create irregular measures but to move subsequent notes later/earlier in time (and then disagree about how many notes to move). Ideally, we'd have a nice consistent story for *all* of these use cases: deleting beats to leave an irregular measure, deleting beats while pulling existing notes to the left, and also both forms of insertion.

In reply to by lasconic

Nicolas, it sounds like you're arguing in favour of that 'scratch-pad' mode that's been on the back burner for a while. I agree wholeheartedly, and haven't changed my original position on 'timewise delete', really, but given that it appeared this new beat-delete function was more or less a fait accompli, I thought it better to look for the best compromise we could find on that basis. Marc's ideas helped us do that.

Do you suggest that 'timewise delete' be scrapped as a command that would function in the 'normal' counting mode, and that the ability to remove beats that it would provide be enabled through a dedicated 'unmeasured entry' mode?

In reply to by Isaac Weiss

I agree. "Resect" might be OK if anyone non-medical understood it - many spell checkers don't even recognise it; I have been English all my (fairly long) life, am quite well educated, and I don't instantly know what it means. "Remove" seems adequate, even without further explanation, and its intended use can easily be found out by using the command, and a section in the Handbook on the "Remove command" can spell out all the use cases. It is also presumably fairly easy to translate - the only problem would be if a language doesn't have common words for "Delete" and "Remove".

Interesting discussion:

flyingninja77 wrote:
It is unfortunate that we haven't found another such succinct and descriptive term with which more people would be familiar.

Recorder485 wrote:
I think that part of the problem with naming this command is that it does more than one thing, so any truly descriptive name would necessarily require more than one verb clause, making it too long and complex for easy comprehension.

Maybe it can be designated a 'tool' - something like 'Scissor' (verb) or 'Scissors' (noun).
Thereafter, the handbook explains the tool's usage.

Regards.

I am quite the late comer to this discussion but the implementation of the timewise delete or what ever it ends up being called for a single note is insane. No one but a computer programmer would expect to remove a singe note from a piece of music that is totally written in 4/4 time and suddenly get a measure that is unmarked and in 3/4 or 7/8 (or whatever) time. If I remove a note, it's because I accidentally double tapped my letter and realized that the note needs to be removed and the entire score needs to be shifted the proper amount to fill the "hole" created by the removed note. The more likely use is that I made the first note an 1/8th note by mistake and realized that I wanted a 16th note so I shorten the first note and have to cut/paste the rest of the measure I put in. I would rather just SHIFT-delete the rest created with Q and have the measure SHIFT internally and place a rest at the end to pad it. I still want a measure in 4/4 (e.g.) time. It would be better if the entire score shifted the distance of the rest (because I probably found my mistake when the last note of the current measure automatically tied itself to the beginning of the next measure), but I realize there are problems with that, especially when it comes to tuplets crossing barlines. This is the feature I see continually asked about in forums as well.

How often do you forget that you wanted a measure in 6/4 rather than 2 measures in 3/4 time so you merge the 2 measures? - and you still have to insert the time signature in that measure if you want one plus the 3/4 to return the remaining measures to normal. Why would you want to easily remove a barline and merge a measure? The only reason I can think of is to create a renaissance song with no measure lines and that is already somewhat available in 2.0.3. You really can't use this feature to remove all barlines if you are going to look at the score in anything but continuous view because the measure would span more than a page. You would have to go back split the measure where you want a new line.

You can already remove a measure and the score will shift so adding that capability again is redundant. You get a very different result from that than with the other results of this feature.

In reply to by mike320

To be clear, this won't exactly be something that could easily be done accidentally—it's a two-key shortcut. So while it's a new feature that won't be of use to everyone, it's not a destructive trap, either. This also has laid the foundation that may allow the function you described with Shift-Delete to be added.

Do you still have an unanswered question? Please log in first to post your question.