Conclusion / synthesis of the keyboard poll results


_wb_

Microbe
Staff member
Joined
Apr 5, 2012
Messages
5,390
Age
40
Location
Brussels, Belgium
The final keyboard poll (http://boards.openpandora.org/topic/18315-keymat-layout-the-final-poll//URL]) is now over.

192 members have voted. These are the top three layouts:

Layout 7: 83 votes

post-1-0-82807200-1437096011.png


Layout 1: 66 votes

post-1-0-74656700-1437096001.png


Layout 10: 53 votes

post-1-0-94727400-1437096015.png


Several people have indicated their reasons for voting or not voting for certain proposals, which helps a lot to interpret the results and combine the best elements from the most popular layouts into one final layout.

For example, one of the things people like in layout 7 is the dedicated Esc key and the principle of staggered US-QWERTY Shift-keypairs on the bottom two Fn-rows, though they can be reordered horizontally for more efficiency.

One of the things people like in layout 1 is the positioning of the German letters üöäß, which is close to their location on a German keyboard.

We can even take specific aspects into account of layouts that didn't get lots of votes, like layout 11, which has F11 as a dedicated key (and that could be useful since it is often used to toggle full-screen, an important toggle on any small-screen device). Or layout 2, which introduced the idea of dead Greek and dead Math keys (an idea that can be kept as unlabeled Fn+Shift+G/M functions, provided that the Fn+G/M symbols do not need a Shifted version).

Here is my attempt to derive a conclusion/synthesis from the discussion in all of the keyboard threads, the final poll results, and the arguments given in the final poll thread:

http://www.keyboard-layout-editor.com/#/layouts/4f6ce9cc204bf2037187d57c9485f740

Obviously it is impossible to make everyone happy in matters like this. Everyone has different needs, is used to different 'standards', and has different tastes. De gustibus non est disputandum.

Still, I think the above proposal would make as many people as happy as possible, and it would in any case be a major improvement compared to the Pandora layout.

I'm not starting this thread to restart the keyboard discussion all over. This thread is only meant to discuss the poll results, how to interpret them, and how to distill one final layout out of it. It is meant to help ED, to make his job easier, not harder. So please don't (re-)introduce new/different layout ideas here or let this thread degenerate into walls of text of keyboard vendettas.

If you want to contribute constructively to this thread, I suggest to do it in the following format:

  • People like or dislike aspect XYZ of layout N  [add pointer(s) to discussion thread]
  • This aspect can be improved in the synthesis layout as follows  [give link to modified layout]
  • This change does not break any of the desired properties  [justification needed, which will be easier for minor changes than for big ones]
 
Last edited by a moderator:

_jr_

Advanced Member
Joined
May 5, 2013
Messages
1,170
Having the brightness control surround the power key and F11/F12 on the rightmost key would better reproduce the easily reachable position of layout 11 (and arguably be more symmetrical). Now that backspace is free of a meta binding one could argue that it is the more natural position for ß, but then the other umlauts would have to be moveded, too, which doesn't look feasible without moving shift to the left side, which being a quite large departure from the top voted layout is probably out.


keyboard vendetta: I'd still prefer to not label the international keys:)
 
Last edited by a moderator:

pmprog

DNF (Did Not Finish)
Joined
Apr 25, 2011
Messages
4,150
F11 as a dedicated key (and that could be useful since it is often used to toggle full-screen, an important toggle on any small-screen device)
Not really. Considering the Pyra is likely to be like the Pandora, and that GLES/SDL/3D-accelerated apps cannot be windowed with the current driver. Sure there's Firefox, but not really something needing a dedicated key IMO.
 

_wb_

Microbe
Staff member
Joined
Apr 5, 2012
Messages
5,390
Age
40
Location
Brussels, Belgium
F11 as a dedicated key (and that could be useful since it is often used to toggle full-screen, an important toggle on any small-screen device)
Not really. Considering the Pyra is likely to be like the Pandora, and that GLES/SDL/3D-accelerated apps cannot be windowed with the current driver. Sure there's Firefox, but not really something needing a dedicated key IMO.
Why do you include SDL in that list? I don't see why SDL apps couldn't be windowed.

It's not just web browsing that benefits from an easy full-screen toggle. I use it also for terminal windows, movie players and text editors, for example (too bad LibreOffice doesn't follow the F11 convention, it has "Styles and Formatting" on F11 and uses a very obscure Ctrl-Shift-J to toggle full-screen, though you can configure it differently).

Anyway, I agree that we don't really need a dedicated F11 key, but if we can have it, then why not.
 

pmprog

DNF (Did Not Finish)
Joined
Apr 25, 2011
Messages
4,150
I don't see why SDL apps couldn't be windowed
Because on the Pandora, even if you specify windowed, it will not be windowed, it will consume the fullscreen; and in one of the other threads, they spoke about the same issue on the Pyra
 

_wb_

Microbe
Staff member
Joined
Apr 5, 2012
Messages
5,390
Age
40
Location
Brussels, Belgium
I don't see why SDL apps couldn't be windowed
Because on the Pandora, even if you specify windowed, it will not be windowed, it will consume the fullscreen; and in one of the other threads, they spoke about the same issue on the Pyra
Pandora System Info is an SDL application that goes windowed without problems. As long as you don't use notaz' omapfb or GL|ES, there is no issue with windowed SDL on the Pandora. I expect the same on the Pyra.
 

pmprog

DNF (Did Not Finish)
Joined
Apr 25, 2011
Messages
4,150
Pandora System Info is an SDL application that goes windowed without problems. As long as you don't use notaz' omapfb or GL|ES, there is no issue with windowed SDL on the Pandora. I expect the same on the Pyra.
Hmmm, I never got any of my SDL code to go windowed. I must have been using the omapfb then (I'll assume it's the default then?), because I certainly wasn't using GLES. I guess none of this really matters. It feels a bit weird to have two F-keys away from the others but whatever, I guess
 

Grench

Forum Addict!
Joined
Oct 3, 2008
Messages
6,629
The final keyboard poll (http://boards.openpandora.org/topic/18315-keymat-layout-the-final-poll//URL]) is now over.

192 members have voted. These are the top three layouts:

Layout 7: 83 votes

post-1-0-82807200-1437096011.png


Layout 1: 66 votes

post-1-0-74656700-1437096001.png


Layout 10: 53 votes

post-1-0-94727400-1437096015.png


Several people have indicated their reasons for voting or not voting for certain proposals, which helps a lot to interpret the results and combine the best elements from the most popular layouts into one final layout.

For example, one of the things people like in layout 7 is the dedicated Esc key and the principle of staggered US-QWERTY Shift-keypairs on the bottom two Fn-rows, though they can be reordered horizontally for more efficiency.

One of the things people like in layout 1 is the positioning of the German letters üöäß, which is close to their location on a German keyboard.

We can even take specific aspects into account of layouts that didn't get lots of votes, like layout 11, which has F11 as a dedicated key (and that could be useful since it is often used to toggle full-screen, an important toggle on any small-screen device). Or layout 2, which introduced the idea of dead Greek and dead Math keys (an idea that can be kept as unlabeled Fn+Shift+G/M functions, provided that the Fn+G/M symbols do not need a Shifted version).

Here is my attempt to derive a conclusion/synthesis from the discussion in all of the keyboard threads, the final poll results, and the arguments given in the final poll thread:

http://www.keyboard-layout-editor.com/#/layouts/4f6ce9cc204bf2037187d57c9485f740

Obviously it is impossible to make everyone happy in matters like this. Everyone has different needs, is used to different 'standards', and has different tastes. De gustibus non est disputandum.

Still, I think the above proposal would make as many people as happy as possible, and it would in any case be a major improvement compared to the Pandora layout.

I'm not starting this thread to restart the keyboard discussion all over. This thread is only meant to discuss the poll results, how to interpret them, and how to distill one final layout out of it. It is meant to help ED, to make his job easier, not harder. So please don't (re-)introduce new/different layout ideas here or let this thread degenerate into walls of text of keyboard vendettas.

If you want to contribute constructively to this thread, I suggest to do it in the following format:

  • People like or dislike aspect XYZ of layout N  [add pointer(s) to discussion thread]
  • This aspect can be improved in the synthesis layout as follows  [give link to modified layout]
  • This change does not break any of the desired properties  [justification needed, which will be easier for minor changes than for big ones]
So - you introduce yet another new layout, incorporating features from layouts people didn't care for, then set a rule that nobody else can introduce another new layout. Interesting.

Here's another idea? How about a truce. How about no new layouts - not even permutations of existing, until ED states what he wants done next? I realize you're doing that whole free speech thing, but the poll wasn't even over and you started creating new layouts.

Let it breathe a bit? Let's find out what ED wants done from here forward?
 
Last edited by a moderator:

ible

professional vim user
Joined
Mar 24, 2014
Messages
2,568
Location
Seattle, WA
I think it would be nice to put a "final evolution" form in the OP for each layout, so ED can see what we've since iterated/improved upon. _wb_'s is a nice combo of 1 and 7, with improved locations for all the symbols. IMO it would be a shame to regress to 7 when _wb_'s latest is a very excellent forward direction.

I would like to put forward 1's final form, with improved locations for the non-international Meta key symbols (i.e. easier to type /? with the keyboard only, no shoulders, very minimal modifications):

http://www.keyboard-layout-editor.com/#/layouts/81311d37a08ca831c053d206836cfdc1

Taking inspiration from 7 and 10 (dedicated escape):

http://www.keyboard-layout-editor.com/#/layouts/b9d15fd98241cbb5ab86b5b734d77752

And with a final augmentation of keeping enter/backspace meta-free:

http://www.keyboard-layout-editor.com/#/layouts/fa838e5aebcafdb391d2b23b5e35e687

Nothing new here, but just wanted to collect them for the interested parties.
 
Last edited by a moderator:

_wb_

Microbe
Staff member
Joined
Apr 5, 2012
Messages
5,390
Age
40
Location
Brussels, Belgium
Here's another idea? How about a truce. How about no new layouts - not even permutations of existing, until ED states what he wants done next? I realize you're doing that whole free speech thing, but the poll wasn't even over and you started creating new layouts.


Let it breathe a bit? Let's find out what ED wants done from here forward?
So you want us to stop trying to make progress until ED gives us the green light? What would be the advantage of such a waiting period?
 

Linux-SWAT

Forum Addict!
Joined
Feb 13, 2010
Messages
9,010
Damn, I wanted to take time to vote wisely, the poll is already closed :/ .
 

_wb_

Microbe
Staff member
Joined
Apr 5, 2012
Messages
5,390
Age
40
Location
Brussels, Belgium
Damn, I wanted to take time to vote wisely, the poll is already closed :/ .
Feel free to share what you would have voted, or more importantly, the wise rationale behind it.
I would have vote for none of the proposed layouts.
OK, and now the wise rationale behind it please ;)

Seriously, it is useful to know why you wouldn't vote for the layouts. It could very well be the case that e.g. you consider some aspect X of layout 1 to be bad, and layout 2 is good in aspect X but bad in aspect Y and Z so you can't vote for that either. This information could be useful to construct a combined layout that is good on all aspects X, Y and Z, if possible.

Of course there's no guarantee that something like that is possible, e.g. if you would have voted for none of the proposals because, say, none of them are AZERTY, then there is no way to make you happy. It all depends on your reasons.
 

ekianjo

Hardcore Member
Joined
May 7, 2012
Messages
8,261
Location
神戸市、日本 (Japan)
OK, and now the wise rationale behind it please
I did not vote either and one rationale is : You don't know what's good and bad until you actually try and test it. Especially regarding keyboards and very touch-sensitive objects like that. It's easy to discuss on paper but that's when you actually have it in hands that the flaws become all the more obvious. 

For me choosing a layout without thorough hands-on testing is borderline crazy, but well I understand there is a timeline to meet.
 
Last edited by a moderator:

mrpalmtop

Very Active Member
Joined
Mar 14, 2014
Messages
133
OK, and now the wise rationale behind it please
I did not vote either and one rationale is : You don't know what's good and bad until you actually try and test it. Especially regarding keyboards and very touch-sensitive objects like that. It's easy to discuss on paper but that's when you actually have it in hands that the flaws become all the more obvious. 

For me choosing a layout without thorough hands-on testing is borderline crazy, but well I understand there is a timeline to meet.
That's why I wish I'd have thought of my idea of a separate keyboard PCB layer sooner. It would have enabled us to have multiple physical layouts. Unfortunately, it was too late by the time I asked ED about it.

It's heresy to say this but I actually prefer a keyboard for gaming as I think it allows for more precise control than what you get with a joystick (too much registering of inadvertent directional movements w/joysticks imo). That's why in my ideal world, we could have a keyboard option that had only keys.
 

Neelix

Insecticidal Maniac
Joined
Jan 8, 2011
Messages
3,281
Location
Melbourne, Australia
That leads to an interesting question... Would it be practical to get multiple keymat layout prototypes made so ED can try each of them out and see what works and what needs tweaking? Perhaps let people try out a couple of layouts at GamesCom to get feedback based on real life usage...

Just a thought.

-Neelix
 

ekianjo

Hardcore Member
Joined
May 7, 2012
Messages
8,261
Location
神戸市、日本 (Japan)
That leads to an interesting question... Would it be practical to get multiple keymat layout prototypes made so ED can try each of them out and see what works and what needs tweaking? Perhaps let people try out a couple of layouts at GamesCom to get feedback based on real life usage...


Just a thought.


-Neelix
Yeah, that would be ideal. But again, it costs money and time... :/ and ED does not seem to want to wait further at this point, but I may be wrong.
 
Top