Saber
Advanced Member
- Joined
- Jan 23, 2012
- Messages
- 1,303
The only criteria I concerned myself with. From that core foundation where it was actualized in my layouts.Saber: ... You have to consider actual use.
The only criteria I concerned myself with. From that core foundation where it was actualized in my layouts.Saber: ... You have to consider actual use.
That's where colour coding comes in. It's fairly well established convention that the top symbol on a key is the shifted version. It's also fairly a well established convention that if some symbols are printed in a different colour, you can access them by pressing the modifier key of that colour. (regardless of what the modifier is called)Thats only how it works in your head.
You arent accounting for this:
Show someone a keyboard where a key has 4 prints, ask them to produce all 4 without help. It fails at _regular_ AltGr. Shift+AltGr, when that isnt even a standard keyboard-feature... Try it.
Shift+AltGr, when that isnt even a standard keyboard-feature...
See, this is the part that bothers me most about your arguments. You are already violating standards. I seriously don't know how else to put this. I'll try one last time, but that's it. I mean it this time.The Pyra does not have enough keys for a standard keyboard. This is obvious. Some compromises must be made entirely out of necessity. There will be parts of known standards we can follow exactly, there will be parts that will look kind of close, and there are parts that we must make completely different.Even if that _were_ the case, standard is better, on a multitude of levels.
Which im sure ible did too, the reason he arrived at a different conclusion was he considered actual use, as in pressing the buttons, rather than a more widely derived aspect of typing. If you look into the argument you will notice how it supports one thing, but not the other. Actual use is definite proof.The only criteria I concerned myself with. From that core foundation where it was actualized in my layouts.Saber: ... You have to consider actual use.
Colour-codes are ugly. And non-standard. Colour-codes are always a hack for when other options couldnt be found. Two colour-codes are uglier than one. Position is what normal keyboards rely on, and that works.Thats only how it works in your head.
You arent accounting for this:
Show someone a keyboard where a key has 4 prints, ask them to produce all 4 without help. It fails at _regular_ AltGr. Shift+AltGr, when that isnt even a standard keyboard-feature... Try it.
That's where colour coding comes in. It's fairly well established convention that the top symbol on a key is the shifted version. It's also fairly a well established convention that if some symbols are printed in a different colour, you can access them by pressing the modifier key of that colour. (regardless of what the modifier is called)
Between the two conventions if I see 4 symbols on a key, and the 2 symbols on the right are (for example) blue, then I'm going to assume that I need to hold shift for the symbols on the top and the blue modifier for the symbols on the right.
- Neelix
The blue ones arent printed. If you tried to sell that to anyone, they would opt for the default instead. I think if they are on AltGr+shift, they should be. Changing it is bad, and relying on AltGr+shift is bad.Shift+AltGr, when that isnt even a standard keyboard-feature...
![]()
Oh... look at all those symbols/characters which are most easily accessed with AltGR+Shift...
¡£÷ÄÅÉËÞÜÚÍÓÖ“”¦Á§ÐÏØ°ÆŒ¢ÑÇ
There's quite a lot for something which apparently isn't a standard feature.
- Neelix
No one said anything about more than two colours. As has been explained it is entirely possible to print four characters with two colours and have them all understood. As for clutter, that's entirely subjective, what one person sees as "clutter" another sees as helpful guides and not cluttered at all. I already demonstrated with the cut-and-paste job that it doesn't take up any additional room, nothing needs to be made smaller than it already is on the Pandora which everyone seems to have agreed is acceptable.Any more than two colours on the keymat could make it quite expensive... Also limiting the number of symbols printed on a key means its likely to look less cluttered
Start here.Less specifically, there's also a nebulous understanding of what constitutes a "standard", things that just kind of are even if they aren't officially written down. This should have been obvious from context but I'm stating it explicitly in case it isn't.What are these standards people keep making oblique references to... I'd love to read these standards....
When someone asks you why you made a certain choice and you respond with information, that is defending.When you tell someone that they are wrong because they have done something you disagree with, and then when they explain they've merely followed the standard differently from you and you still reject that as "not making sense" is attacking.I am not dictating. I am defending the differences my layout has to any standards i have chosen to derive them from. With other layouts it sometimes seem (to me) as if standards are invented, so i feel it neccisary to ask, because i could be missing something.
Actually they can be printed. And if they're not, you can get something like this:It also makes it exceed ED's budget, pretty sure he said 3 colours at the most. Same as he said 3 prints each key as the most.
When its easy to make good arguments for these choices, why is it that some people break them without even touching on the logic behind it?
"People could be able to use Shift+AltGr" isnt dealing with the issue, its dealing with the issue created by going in that route.
Still dont see why 2 colours are better than 1, when even 1 is ugly.
WizardStan:Variations due to neccesity are very different from inventing something entirelly new. When you say "must make completely different" What is the extent of doing that?
Point out what you actually mean instead of just drawing up similarities which arent dealing with the issue.
Shift+; isnt : in "the standard", look at a german keyboard.
Instead of inventing an argument, you could make one. What is egocentric about designing a layout based on clear metrics? If i werent sharing them, then what is the point in discussion? All im asking is the exact same of others. I am very open to real arguments.
I am not dictating. I am defending the differences my layout has to any standards i have chosen to derive them from. With other layouts it sometimes seem (to me) as if standards are invented, so i feel it neccisary to ask, because i could be missing something.
When have i said "must be standard" what are you reffering to?
3per key isnt falsely derived. But do provide an argument as to why you think so. Its the amount of symbols you need to put on numbers before you start putting symbols on all letters (2-per key) Since it is possible at 3, there is no point in doing 4.
Which im sure ible did too, the reason he arrived at a different conclusion was he considered actual use, as in pressing the buttons, rather than a more widely derived aspect of typing. If you look into the argument you will notice how it supports one thing, but not the other. Actual use is definite proof.The only criteria I concerned myself with. From that core foundation where it was actualized in my layouts.Saber: ... You have to consider actual use.
Colour-codes are ugly. And non-standard. Colour-codes are always a hack for when other options couldnt be found. Two colour-codes are uglier than one. Position is what normal keyboards rely on, and that works.Thats only how it works in your head.
You arent accounting for this:
Show someone a keyboard where a key has 4 prints, ask them to produce all 4 without help. It fails at _regular_ AltGr. Shift+AltGr, when that isnt even a standard keyboard-feature... Try it.
That's where colour coding comes in. It's fairly well established convention that the top symbol on a key is the shifted version. It's also fairly a well established convention that if some symbols are printed in a different colour, you can access them by pressing the modifier key of that colour. (regardless of what the modifier is called)
Between the two conventions if I see 4 symbols on a key, and the 2 symbols on the right are (for example) blue, then I'm going to assume that I need to hold shift for the symbols on the top and the blue modifier for the symbols on the right.
- Neelix
Pretty well established, if we assume it to be true, is still worse than established. When i said try it, i meant on someone who isnt you. Preferably someone who only/if barely knows how to use a regular keyboard.
Which one of the blue ones is altGr, and which one is just AltGr?
The blue ones arent printed. If you tried to sell that to anyone, they would opt for the default instead. I think if they are on AltGr+shift, they should be. Changing it is bad, and relying on AltGr+shift is bad.Shift+AltGr, when that isnt even a standard keyboard-feature...![]()
Oh... look at all those symbols/characters which are most easily accessed with AltGR+Shift...
¡£÷ÄÅÉËÞÜÚÍÓÖ“”¦Á§ÐÏØ°ÆŒ¢ÑÇ
There's quite a lot for something which apparently isn't a standard feature.
- Neelix
Yup, exactly. This is exactly what I've been trying to impress upon comradekingu, that we cannot follow the "standard" exactly, we need to make sacrifices and that different people have different ideas of what sacrifices need to be made: we can't do everything the user expects but we can at least do some stuff they expect. When Grench suggested the "fn"+"shift" combination it made sense as one of those trade-offs.There is no standard I know of for keyboards with just 41 or so keys
Says you. I counter that it makes it more intuitive, less hunting and pecking, because users already have ":" encoded into their brain as "shift"+";" so you only need to remember where one punctuation mark is, not two.putting 4 symbols on a large number of the keys will look crowded and off putting making it very time consuming to "scan and peck", very off putting to new or prospective users.
I've tried explaining it to you, very slowly, using walls of text, and with bullet points. You still don't get it. You just can't get it, I don't know what else to tell you.I'm just going to leave it at that. I know others have understood. They may disagree and have better ideas but at least they understood. Some understood and did agree. When the time comes we'll just have to see what wins out. I don't claim to have a perfect plan, but I am convinced that at least some of the ideas are good. If you still can't see that then I don't see why I should keep trying to explain it to you.So what _more_ of user-expectancy is it you bring to the table?
By follow I think you mean approximate. I remember how adamant you were for an uncentered "left-centric" lettered keyboard. P+2 and L+3. How it was "the way". Then it was Greek letters instead of ABXY. Then the single key spacebar protesting. Back to Greek again. None of those came to be. So in practice the act of respecting a convention, any convention, needs to be diregarded when developing a layout for the Pyra. Otherwise we end up nowhere fast by imposing these limits that no one seems to know why are being followed in the first place, only that others have before us and vaguely we must as well, for a handheld that isn't a desktop.I'm not saying that US-QWERTY makes more sense than German QWERTZ or French AZERTY. It doesn't. None of the frequently used keyboard layouts really make sense. They just are what they are.
But I do think that it's better to try to follow just one of them (and I would argue it should be US-QWERTY), and follow it as closely as possible (within the constraints of having limited keys etc). I think it's a bad idea to do the letters like in the US, the shifted comma and period like in Germany and the shifted number row like something else again.
So in practice the act of respecting a convention, any convention, needs to be diregarded when developing a layout for the Pyra. Otherwise we end up nowhere fast by imposing these limits that no one seems to know why are being followed in the first place, only that others have before us and vaguely we must as well, for a handheld that isn't a desktop.![]()
On the nature of four characters per key, I took the liberty of gimping this together
![]()