Unofficial keyboard layout poll (bis)

[Saber]

This is the bad thing about a few unofficial keyboard polls...

I count a few people from my one who haven't chimed in on this one (Askarus, ouroboros, Zink, who all voted for Saber's layout; Inqui, Wilhelm, and nomax for _wb_'s).

I think those persons with layout proposals shouldn't be voting for their own or against any others, to keep things honest numbers-wise. 

 

[bzar]

I think those persons with layout proposals shouldn't be voting for their own or against any others, to keep things honest numbers-wise. 

I've never understood this viewpoint. How are people making things (be it competition entries or keyboard layout proposals) not part of the community voting for their favorites? Of course authors will probably vote for their own entry, and should vote honestly about the others as they would had they not created something themselves. Compare with the situation that someone else had made the exact same entry. Why should it matter who made the entry as long as the authors are fair toward others as well?

 

[_wb_]

This is the bad thing about a few unofficial keyboard polls...


I count a few people from my one who haven't chimed in on this one (Askarus, ouroboros, Zink, who all voted for Saber's layout; Inqui, Wilhelm, and nomax for _wb_'s).

I think those persons with layout proposals shouldn't be voting for their own or against any others, to keep things honest numbers-wise. 

To the contrary: I think that people who made the effort to make proposals, should definitely vote! After all, they most likely gave more thought to it than anyone else. And yes, obviously they will like their own proposal.

For your situation, as a start with the rest to figure out on your own, remap over the German umlauts you don't need for Circumflex. It won't cover all of them but that takes care of four of them(or you could change Dead Grave or Acute to Dead Circumflex, depending on which group has a greater number of diacritics needed). Dead key diacritic for umlauts, additionally, could be on Shift + AltGr + s and Shift + AltGr + n or m for Circumflex. There's also Shift + AltGr + Euro if you find Scroll Lock unnecessary.


Letters ç or ñ could go over Euro and the other you'll make with Compose, or place it and any others under Shift + AltGr, mapping over those, like en and em dash which you don't need.

OK so basically what you propose is that non-Germans will have to live with labels that are just horribly wrong, like they have to type € to get ç, Ü to get dead circumflex, and so on?

Shift + AltGr, as it concerns the F-keys, has a simple solution: we don't put any values there, that is, if people(more than solely you) require Shift + F1-12 shortcuts. You didn't answer me yet but are there alternatives to those Shift F-key shortcuts?

I don't often use Shift F-key shortcuts, and I only use FOSS software so of course there are always alternatives (even if that means I have to modify and recompile stuff). That's not the point. The point is that something is broken if Shift + F1 does anything else than Shift + F1.


Imagine I would say: let's put the symbol ≥ at Shift+X, because really, how often do you need capital X? Can't you just rephrase your sentences so they don't start with a word that begins with X? And if you really need capital-X that often, you can always remap € or § to capital X, you know.

Concerning BT and WiFi, you did notice their location way, way up top? How does one accidently hit those keys? As for GPS and HDMI, there are AltGr + top key slots vacant. 

Yes, I agree that you'll rarely hit them by accident. I just wonder why you would want separate dedicated keys for such toggles, with each specific hardware setting separately labeled (or unlabeled, which is even worse). It means we have to know in advance, and set in stone for now and forever, which toggles will be the most important ones (so they are primary hardware buttons), which ones are less important (shifted), and which ones are not so important (AltGr/Meta/Fn, with or without Shift).

To me, a key deserves to be a dedicated primary key if there are plausible use cases in which you want to repeatedly press that key, or in which you actually use that key many times (say 5 or 10 times) in a short period of time, say one minute. Maybe that's a rather arbitrary criterion, but to me it makes sense. A bluetooth toggle just doesn't fit that criterion.

 

[Saber]

I think those persons with layout proposals shouldn't be voting for their own or against any others, to keep things honest numbers-wise. 

I've never understood this viewpoint. How are people making things (be it competition entries or keyboard layout proposals) not part of the community voting for their favorites? Of course authors will probably vote for their own entry, and should vote honestly about the others as they would had they not created something themselves. Compare with the situation that someone else had made the exact same entry. Why should it matter who made the entry as long as the authors are fair toward others as well?

I can see truly you don't understand. In simpler terms let's just say that when you've cast your ballot already(by submitting your own proposal) you shouldn't have a second vote in the process counted for yours and against others.   

This is the bad thing about a few unofficial keyboard polls...


I count a few people from my one who haven't chimed in on this one (Askarus, ouroboros, Zink, who all voted for Saber's layout; Inqui, Wilhelm, and nomax for _wb_'s).

I think those persons with layout proposals shouldn't be voting for their own or against any others, to keep things honest numbers-wise. 

To the contrary: I think that people who made the effort to make proposals, should definitely vote! After all, they most likely gave more thought to it than anyone else. And yes, obviously they will like their own proposal.

For your situation, as a start with the rest to figure out on your own, remap over the German umlauts you don't need for Circumflex. It won't cover all of them but that takes care of four of them(or you could change Dead Grave or Acute to Dead Circumflex, depending on which group has a greater number of diacritics needed). Dead key diacritic for umlauts, additionally, could be on Shift + AltGr + s and Shift + AltGr + n or m for Circumflex. There's also Shift + AltGr + Euro if you find Scroll Lock unnecessary.


Letters ç or ñ could go over Euro and the other you'll make with Compose, or place it and any others under Shift + AltGr, mapping over those, like en and em dash which you don't need.

OK so basically what you propose is that non-Germans will have to live with labels that are just horribly wrong, like they have to type € to get ç, Ü to get dead circumflex, and so on?

Shift + AltGr, as it concerns the F-keys, has a simple solution: we don't put any values there, that is, if people(more than solely you) require Shift + F1-12 shortcuts. You didn't answer me yet but are there alternatives to those Shift F-key shortcuts?

I don't often use Shift F-key shortcuts, and I only use FOSS software so of course there are always alternatives (even if that means I have to modify and recompile stuff). That's not the point. The point is that something is broken if Shift + F1 does anything else than Shift + F1.


Imagine I would say: let's put the symbol ≥ at Shift+X, because really, how often do you need capital X? Can't you just rephrase your sentences so they don't start with a word that begins with X? And if you really need capital-X that often, you can always remap € or § to capital X, you know.

Concerning BT and WiFi, you did notice their location way, way up top? How does one accidently hit those keys? As for GPS and HDMI, there are AltGr + top key slots vacant. 

Yes, I agree that you'll rarely hit them by accident. I just wonder why you would want separate dedicated keys for such toggles, with each specific hardware setting separately labeled (or unlabeled, which is even worse). It means we have to know in advance, and set in stone for now and forever, which toggles will be the most important ones (so they are primary hardware buttons), which ones are less important (shifted), and which ones are not so important (AltGr/Meta/Fn, with or without Shift).

To me, a key deserves to be a dedicated primary key if there are plausible use cases in which you want to repeatedly press that key, or in which you actually use that key many times (say 5 or 10 times) in a short period of time, say one minute. Maybe that's a rather arbitrary criterion, but to me it makes sense. A bluetooth toggle just doesn't fit that criterion.

Always so dramatic. You do realize that none of the Shift + AltGr values have been set in stone yet(or if they'll even exist beyond the uppercase umlauts)? You also haven't given me a list of these supposedly crucial Shift + F-keys you're desperately and repetitively in favor of not losing. Care to share them, or will you sidestep the question yet again?  ^_^

WiFi and Bluetooth are regularly used toggles, and while your "wonky" hardware key does save space for Escape, Insert, and Delete positioned way up top(also to be accidently pressed as you've worried), it will require someone to code a gui popup to show present status for it with absolutely zero guarantees it won't clash with other opened software the individual may find themselves using at the time, if they want to render WiFi, Bluetooth, or the like active. 

 

[_wb_]

Here's another minor modification of my proposal.

I'm still undecided between having standard Shifted punctuation symbols, that is {}_+?~ at Shift+[]-=/`, and having direct Meta-keys for the shifted punctuation symbols (like on the Pandora). I'm moving back in the direction of direct Meta-keys though; I think it looks cleaner and it's slightly more efficient.

So here is my new main proposal:

Main proposal (labels as printed): http://www.keyboard-...add280f30b3d279

Main proposal (including unlabeled default mappings in blue): http://www.keyboard-...187094e2e220d6d

Changes:

• Swapped keyboard Meta and Shift
• Put dead greve at Shift+backtick (`), since tilde is not at Shift+backtick anymore. (I used to have dead grave at Meta+G, so now \ can move there, freeing up one more key)
• Illustrated the proposal by adding an unlabeled default mapping that covers German, French, Spanish, Portuguese, and the Scandinavian languages.

 

[Saber]

^

It'll be difficult for the owner to remember where all those keys are if they are blank, especially without any correlation: those on W, R, T, Y, I, D and F for instance.

 

[_wb_]

Always so dramatic. You do realize that none of the Shift + AltGr values have been set in stone yet(or if they'll even exist beyond the uppercase umlauts)? You also haven't given me a list of these supposedly crucial Shift + F-keys you're desperately and repetitively in favor of not losing. Care to share them, or will you sidestep the question yet again?  ^_^

Who said anything about crucial and desperation? I'm just saying it should be possible to do Shift + F-keys, like it should be possible to do Shift+space, Shift+enter, Shift+backspace and so on. If you want some Shift + F-key shortcuts to understand the point, I'll give you some:

In the text editor I'm using (MCEdit), "Save" is F2 while "Save as" is Shift-F2; "Search" is F7 and "Search again" is Shift-F7.

In the browser I'm using (Firefox), you can go navigate between frames with F6 (next) and Shift-F6 (previous), when searching for something (Ctrl-F or F3), you can use F3 to find the next occurrence and Shift-F3 for the previous one. Shift-F2 shows the dev toolbar, Shift-F5 shows the profiler, Shift-F7 the style editor, Shift-F4 the scratchpad.

In the word processor I'm using (LibreOffice), the default shortcuts include Shift-F3, Shift-F4, Shift-F5, Shift-F8, Shift-F9, Shift-F11, Shift-F12, Ctrl-Shift-F5, Ctrl-Shift-F8, Ctrl-Shift-F9, Ctrl-Shift-F11, Ctrl-Shift-F12.

Obviously there are alternative ways to do those things. You can use the menus, for example. Or e.g. in LibreOffice, you can easily modify the shortcuts. But that's not the point. The point is: why make things complicated and make Shift + F-keys do weird stuff?

WiFi and Bluetooth are regularly used toggles, and while your "wonky" hardware key does save space for Escape, Insert, and Delete positioned way up top(also to be accidently pressed as you've worried), it will require someone to code a gui popup to show present status for it with absolutely zero guarantees it won't clash with other opened software the individual may find themselves using at the time, if they want to render WiFi, Bluetooth, or the like active.

I think such a popup / on-screen-display / menu / notification thing has to be coded anyway (or perhaps / probably something like that already exists). Or do you want your toggle buttons to do their thing without any visual feedback? Do you want the power button to immediately shut down the device, without any confirmation dialog?

 

[_wb_]

^

It'll be difficult for the owner to remember where all those keys are if they are blank, especially without any correlation: those on W, R, T, Y, I, D and F for instance.

Yes, that's true. Which is why that would be just a default. It's fine if you're German to have Ä on A, but if you're French, you'll probably want à on A, if you're Danish you might want Å on A, and so on.

 

[Saber]

Always so dramatic. You do realize that none of the Shift + AltGr values have been set in stone yet(or if they'll even exist beyond the uppercase umlauts)? You also haven't given me a list of these supposedly crucial Shift + F-keys you're desperately and repetitively in favor of not losing. Care to share them, or will you sidestep the question yet again?  ^_^

Who said anything about crucial and desperation? I'm just saying it should be possible to do Shift + F-keys, like it should be possible to do Shift+space, Shift+enter, Shift+backspace and so on. If you want some Shift + F-key shortcuts to understand the point, I'll give you some:

In the text editor I'm using (MCEdit), "Save" is F2 while "Save as" is Shift-F2; "Search" is F7 and "Search again" is Shift-F7.

In the browser I'm using (Firefox), you can go navigate between frames with F6 (next) and Shift-F6 (previous), when searching for something (Ctrl-F or F3), you can use F3 to find the next occurrence and Shift-F3 for the previous one. Shift-F2 shows the dev toolbar, Shift-F5 shows the profiler, Shift-F7 the style editor, Shift-F4 the scratchpad.

In the word processor I'm using (LibreOffice), the default shortcuts include Shift-F3, Shift-F4, Shift-F5, Shift-F8, Shift-F9, Shift-F11, Shift-F12, Ctrl-Shift-F5, Ctrl-Shift-F8, Ctrl-Shift-F9, Ctrl-Shift-F11, Ctrl-Shift-F12.

Obviously there are alternative ways to do those things. You can use the menus, for example. Or e.g. in LibreOffice, you can easily modify the shortcuts. But that's not the point. The point is: why make things complicated and make Shift + F-keys do weird stuff?

WiFi and Bluetooth are regularly used toggles, and while your "wonky" hardware key does save space for Escape, Insert, and Delete positioned way up top(also to be accidently pressed as you've worried), it will require someone to code a gui popup to show present status for it with absolutely zero guarantees it won't clash with other opened software the individual may find themselves using at the time, if they want to render WiFi, Bluetooth, or the like active.

I think such a popup / on-screen-display / menu / notification thing has to be coded anyway (or perhaps / probably something like that already exists). Or do you want your toggle buttons to do their thing without any visual feedback? Do you want the power button to immediately shut down the device, without any confirmation dialog?

I don't have a confirmation dialog when I turn WiFi on my laptop yet I manage just fine. The dashboard's dock can offer status.  

As for your statement "Obviously there are alternative ways to do those things. You can use the menus, for example. Or e.g. in LibreOffice, you can easily modify the shortcuts. But that's not the point. The point is: why make things complicated and make Shift + F-keys do weird stuff?", that there are alternative ways shouldn't be dismissed when we can get even more distinct keys out of the Pyra layout(which you've rallied for as well).   

 

[Saber]

^

It'll be difficult for the owner to remember where all those keys are if they are blank, especially without any correlation: those on W, R, T, Y, I, D and F for instance.

Yes, that's true. Which is why that would be just a default. It's fine if you're German to have Ä on A, but if you're French, you'll probably want à on A, if you're Danish you might want Å on A, and so on.

Glancing at the wiki for French orthograghy, if I was French, I would be inclined to remap Dead Acute to Dead Circumflex given the number of accented keys for each subgroup. 

 

[Saber]

Thinking about it some more, in my layout Shift + AltGr + s for dead Umlaut and/or Shift + AltGr + x for dead Circumflex has a nice melody to it. Given the locations of s and x they should be easy and worthwhile to do, Circumflex in particular. 

 

[bzar]

I think those persons with layout proposals shouldn't be voting for their own or against any others, to keep things honest numbers-wise. 

I've never understood this viewpoint. How are people making things (be it competition entries or keyboard layout proposals) not part of the community voting for their favorites? Of course authors will probably vote for their own entry, and should vote honestly about the others as they would had they not created something themselves. Compare with the situation that someone else had made the exact same entry. Why should it matter who made the entry as long as the authors are fair toward others as well?

I can see truly you don't understand. In simpler terms let's just say that when you've cast your ballot already(by submitting your own proposal) you shouldn't have a second vote in the process counted for yours and against others.

Suppose five people want to make a choice that reflects their collective will. Four of the five people offer a proposal. Which of the following yields the best representative result:

1. Only people with no proposals get a vote
2. Everyone gets one vote (simple majority vote)
3. Everyone orders the proposals by preference, the one with the best average rank wins

Yes, I'm cherry picking the situation to prove a point. Making a suggestion is not the same as a vote. Everyone can make suggestions. If only people with no suggestions vote, the voting result will reflect the will of that segment of the total population, limited by the options presented by the rest. It's not representative of the entire population.

I agree with you that doing what we have seen in a previous vote, that is an author ranking his proposal the best, all the other the worst, is (and was) a bad thing. Average ordering is better in representing collective will.

 

[Grench]

As for your statement "Obviously there are alternative ways to do those things. You can use the menus, for example. Or e.g. in LibreOffice, you can easily modify the shortcuts. But that's not the point. The point is: why make things complicated and make Shift + F-keys do weird stuff?", that there are alternative ways shouldn't be dismissed when we can get even more distinct keys out of the Pyra layout(which you've rallied for as well).   

I'll simplify this.

It is a keyboard.  It must behave as a keyboard.

If you can't do F2, shift+F2, alt+F2, ctrl+F2, alt+ctrl+shift+F2 then the layout is broken.

ALL of the F1-F12 keys MUST work in all cases with and without modifiers or the layout is broken and must be re-thought.

Does that make it easier to understand?

 

[_wb_]

1. This is an unofficial vote, so who cares?

2. This is a public vote, so if anyone is voting in a dishonest/manipulative way, everyone can see that.

 

[ible]

Main proposal (labels as printed): http://www.keyboard-...add280f30b3d279

Huh, interesting, didn't think you'd go full Saber on us .

I noticed an interesting symmetry about the letters i,j,k, which are common indices in programming, with Meta+JK being [], shift+meta+JK being {} from your previous layout. For ultra efficiency in coding with arrays, you could modify:

http://www.keyboard-layout-editor.com/#/layouts/c86282eb5a2d4395d872c77ebc409ead

I agree with your removal of Grave accent being shift+meta+tab, requiring tilde to move. I've edited some of my layouts in response.

 

[_wb_]

Main proposal (labels as printed): http://www.keyboard-...add280f30b3d279

Huh, interesting, didn't think you'd go full Saber on us .


I noticed an interesting symmetry about the letters i,j,k, which are common indices in programming, with Meta+JK being [], shift+meta+JK being {} from your previous layout. For ultra efficiency in coding with arrays, you could modify:

http://www.keyboard-layout-editor.com/#/layouts/c86282eb5a2d4395d872c77ebc409ead


I agree with your removal of Grave accent being shift+meta+tab, requiring tilde to move. I've edited some of my layouts in response.

I don't know if we should really tweak for efficient array indexing , but if you like to have [ ] at Meta+JK, then I would do it like this:

http://www.keyboard-layout-editor.com/#/layouts/bbdb93339ca9922ac84ac831f77002fe

Reason: while { [ ] } is nice from an ascii-art perspective (and for Germans I guess), I don't like the separation between { and }. Also, with this modification, [ ] \ are in their normal order and \ is close to | which also makes sense (as they're on the same key on normal keyboards).   [normal being US-QWERTY, as usual]

My original reasoning was to have [ ] aligned with ( ), by the way.

 

[ible]

Ah, nice, I didn't see the () alignment. I guess pretty much any way has some patterns.

 

[comradekingu]

Saber, you are confusing keyboard labels and their functionality. Making it next to impossible for someone who doesnt know all these things to form an opinion.

Shift+altGr+nubers already does something. This is a reason why you cant call what it is that you are proposing to do, AltGr. It simply isnt.

Updated my layout again. In my sig, the final keyboard thread, or/and the official thread here. As you were. 

 
Nice to see that you're using the layout editor now, that makes it a bit easier to compare the different proposals.
 
Remark: you shouldn't put shift+AltGr (shouldn't call it AltGr either) things on the number row / F-keys, because you have to be able to do Shift-F-keys. Shift-F5 should be just Shift-F5, not ≥.
 
Also I wonder how anyone can remember all those shift+AltGr symbols if they aren't labeled. Not that I think that those symbols are very necessary. Most people don't need these symbols: ¡ ¿ § « » ² ³ ≤ ≥ ° ÷ — ¥ µ; if you do for some reason need some of them frequently, you can of course map them somewhere, but I don't think we need them by default. They're also easy to remember with compose:

The most important aspect to using a keyboard is position. Ask people to type on a keyboard and people will do so, even blindly. Ask them to recite what the buttons are, and they dont really know. New and experienced users alike. I can dig up the research if you like.

Why cant we ask ourselves how _few_ metrics are disturbed or made impossible, rather than to rationalize how they dont matter? Make the list, ask the question, are these compromises worth it?

In its essence, the second things are changed, it puts an additional burden of learning on the user. And it begs the question, are we smart enough to start playing dvorak. I can answer to that last thing myself, no we are not. Not individually, not collectivly.

Just because you cant think of a reason not to mess the possibility to mess up shift+anything, doesnt make it ok. Nor it is a guarantee against someone coming along later to make use of that functionality. The gnome-project is real, and it exists.

I think those persons with layout proposals shouldn't be voting for their own or against any others, to keep things honest numbers-wise. 

I've never understood this viewpoint. How are people making things (be it competition entries or keyboard layout proposals) not part of the community voting for their favorites? Of course authors will probably vote for their own entry, and should vote honestly about the others as they would had they not created something themselves. Compare with the situation that someone else had made the exact same entry. Why should it matter who made the entry as long as the authors are fair toward others as well?

If you conduct a vote, that is biased, made by a person who has a vested interest, the whole notion of fair voting is gone.

Same as if you conduct a vote, let the results be known, and the ones whos proposals are voted over, cast their votes right away. If people see a discussion almost devoid of reasoning and collective rationality, they wont have a reason to participate that outweighs the reasons not to.

Voting for yourself in a vote that is supposed to be about general perceptance is what i dont understand. Seeing as people design things that only befit themselves, the vote is already inherrent in the work.

 

[bzar]

If you conduct a vote, that is biased, made by a person who has a vested interest, the whole notion of fair voting is gone.


Same as if you conduct a vote, let the results be known, and the ones whos proposals are voted over, cast their votes right away. If people see a discussion almost devoid of reasoning and collective rationality, they wont have a reason to participate that outweighs the reasons not to.


Voting for yourself in a vote that is supposed to be about general perceptance is what i dont understand. Seeing as people design things that only befit themselves, the vote is already inherrent in the work.

The voting happens on equal ground. Everyone has equal opportunity to propose and vote on solutions. Because of this, the gathering of proposals and voting on them can be considered separate. First, a collection of proposals is requested, then those proposals are voted on. Considering them separate, to get the collective will they can be considered "given" with no regard to their origins. The author of a proposal should not affect the result of the vote in any way. Only the proposals and the collective will of the members of the voting community should have any bearing on the result. The result should reflect the optimal solution for the entire community. This is not possible if only a part of the community can vote.

If there are ten people in a community and you want to gauge the popularity for a proposal only the author would vote for, is the correct number 0% or 10%?

 

[comradekingu]

You jumped to conclusion, and didnt at all consider swaying polls and tactical voting. Not voting for oneself isnt a new concept.

That destinction isnt taking away peoples ability to vote. There is only a fixed representative that does vote. Same reasons the many worthless threads arent possible, the ones participating in them are making the vote a representation of that. And thus, the same people who dont want to participate in that manner, are left out.

Hidden elections were made for this purpose.

You dont solve the problem of not being able to vote having made a layout, you create the problem of few voters misrepresentation.

Again, show layouts to non-pandora people. There you have a virtually endless supply of unbiased prospective buyers. If you dont help them, they are on their own, and  a very real representation of themselves.

Its not the people at fault, because they cant be changed by magic. Design for, not against.

Edit: Wb: I am still interested in having my questions answered and returned in similar fashion. Open for anyone.