When comparing Carpalx QGMLWB vs qwpr, the Slant community recommends Carpalx QGMLWB for most people. In the question“What are the best keyboard layouts for programming?” Carpalx QGMLWB is ranked 10th while qwpr is ranked 13th. The most important reason people chose Carpalx QGMLWB is:
If you already know QWERTY you don't have to relearn the punctuation, with the single exception of the `;:` key which is in QWERTY's `P` position, like Colemak.
Specs
Ranked in these QuestionsQuestion Ranking
Pros
Pro Punctuation is in the same location as QWERTY
If you already know QWERTY you don't have to relearn the punctuation, with the single exception of the ;:
key which is in QWERTY's P
position, like Colemak.
Pro ZXC don't move compared to QWERTY
Like Colemak, the common undo, cut and copy Ctrl-commands don't move from their QWERTY positions. The V
(paste) key does though.
Pro Uses a colemak-like character layout
Uses the colemak character layout of moving P on a standard QWERTY keyboard layout one step down, extending the home row.
Pro Has low consecutive finger use
See the source here.
Pro Letter positions optimized
Via a quantitative effort model.
Pro Good for vim users
Qwpr is pretty close to QWERTY, and even the HJKL keys are in the same left-to-right order (though on different rows). Alternatively, the AltGr plane also has arrow keys in a sensible position under the right hand.
Pro Common shortcuts don't move
A, Z, X, C, and V are in the same positions as QWERTY and Colemak.
Pro As easy as Dvorak
It's 32% easier than QWERTY by the Carpalx metric, which is slightly better than Dvorak's 30%. This is probably within Carpalx's margin of error though.
Pro Alternate plane with CapsLock key
CapsLock is pretty useless for most people, but qwpr layout uses it to shift to another plane with easy access to punctuation and arrow keys. This is especially useful for programmers.
Pro Minimal retraining from QWERTY
11 keys move, but except for P and E, they don't change fingers.
Cons
Con Effort model is speculative
The quantitative effort model central to the optimization is based on armchair speculation, rather than a scientific biomedical study. The chosen metrics and weighting for them are likely partially correct. But no-one is really sure how correct.
Con Doesn't take finger length into account
Unlike layouts such as norman and workman, QGMLWB doesn't take the length of fingers into account, for example on a standard QWERTY layout, it's easier to reach E than C.
Con Punctuation is not optimized
Programmers have to use punctuation a lot, but (except for the ;:
key, like Colemak) punctuation hasn't been moved from their positions on QWERTY. In fact, the non-letter characters , . - " _ ' ) ( ; 0 1 = 2 :
are used more than the least-frequent letter z
in a reasonable English corpus. Not optimizing punctuation at all, especially for programmers, is nonsense.
Con V key has moved compared to QWERTY
The common paste shortcut used in for example windows has been moved to the right hand, making the layout harder to learn. This however is fixed in the QGMLWY variant of the carpalx series.
Con A very small user base and community
Con Doesn't favor the right hand
For right handed users, this keyboard layout doesn't use the usually stronger right hand more than the left, infact it sometimes favors the left hand more.
Con Puts E on the pinky
'E' is the most used English letter by far, at almost 13%. That's almost as much as the spacebar. It needs to be on a strong finger. The pinky is the weakest finger and on the right side it is already overtaxed from Ctrl, Shift, and Enter.
Con P and E change fingers from QWERTY
Which makes it harder than necessary to learn from QWERTY. (And makes no sense. 'E' was arguably better in its QWERTY position on a strong finger.) This is due to using the flawed Carpalx effort model.