First of all, it's clear that the article is not talking about the QWERTY layout per se, but the character set on the keyboard-- switching to Dvorak would do nothing to improve the situation described.
Second, it's odd that he doesn't bring up the experience of APL, the only language that I know of that tried to break out of the traditional character set. Aficionados find it concise and elegant; personally, I find it cryptic. I fear the same might occur with other attempts in this vein.
I don't think QWERTY is actually harming language design - designers simply implement multiple-character operators (like ==). (however, I certainly agree it's awkward and extended symbol operators would be cleaner).
Also, as the author mentioned, I don't think it's really feasible to get all coders to agree on one 'coder keyboard' layout (hell - we've had 50 years to decide on an optimal editor and are no closer than when we started).
I think the best solution is to continue with this multi-char symbol syntax, but to have keyboards with programmable centrally placed macro keys that can be automatically redefined by the current application. For example, when coding lisp I'd love to have dedicated paren keys, and maybe even a dedicated 'setf', 'defun', and 'let' keys.
I think that as OLED keyboards ala the Optimus Maximus become more economically feasible and widespread over the next decade or two, this is exactly what will happen. Then, in 20-30 years when these keyboards become ubiquitous, language designers will start to implement non-ascii operators that can be mapped to the programmable keys.
It will certainly be prettier - but I don't think it will herald in a new era of better languages.
I'd favor blank keyboards outright, at least until we have those keyboards with fancy screens embedded in them, and even then.
1. I hardly ever typed on a keyboard where the labels on the keys corresponded to the actual current logical keyboard layout. There are so many potential issues, maybe you're typing canadian french on a US keyboard, maybe you're typing Dvorak, maybe you're typing special characters, maybe you're typing some foreign language, maybe you're typing in emacs with your special keybindings.
2. Looking at the keyboard while typing is a great way to type slow and inaccurately, and maybe even have neck pain if you're lucky. And you can't see what you're typing on the screen. Learning touchtyping takes just a couple days (and motivation!) and is a great investment.
And an extensive keyboard macro system embedded right in the OS would be great!
I don't think a syntax with a whole bunch more sigils, punctuation and the like would be an improvement in language design. Of course, I've always been more of an oatmeal and toenail clippings guy than a line-noise guy.
Look at Ruby. The guy who came up with it is a researcher at Osaka university; English is definitely not his native language.
Computers tend to use English for a number of reasons, some practical, but mostly historical. In this particular case, the practical reason involves the lack of a good, concise, 'if-then-else' structure in the Japanese language. Sure, the idea can be expressed, but it isn't really a common construct in everyday life.
good point this - Is it perhaps not the QUERTY keyboard that is limiting language syntax design but the language itself. Most (but not all) languages have been designed "in English" - using typical English language words (for verbose languages like BASIC) and making use of the punctuation of English to punctuate syntax structure.
Second, it's odd that he doesn't bring up the experience of APL, the only language that I know of that tried to break out of the traditional character set. Aficionados find it concise and elegant; personally, I find it cryptic. I fear the same might occur with other attempts in this vein.