On Thursday, March 10, 2016 at 4:21:15 PM UTC+5:30, Marko Rauhamaa wrote: > Ben Finney : > > > As for how solved it is, that depends on what you're hoping for as a > > solution. > > > > [...] > > > > Hopefully your operating system has a good input method system, with > > many input methods available to choose from. May you find a decent > > default there. > > I don't have an answer. I have requirements, though: > > * I should be able to get the character by knowing its glyph (shape). > > * It should be very low-level and work system-wide, preferably over the > network (I'm typing this over the network). > > The solution may require a touch screen and a canvas where I can draw > with my fingers. > > The solution may have to be implemented in the keyboard. > > Or maybe we'll have to wait for brain-implantable bluetooth tranceivers. > Then, we'd just think of the character and it would appear on the > screen.
Lets say you wrote/participated in a million-line C/C++ codebase. And lets say you would/could redo it in super-duper language 'L' (could but need not be python) Even if you believe in a 1000-fold improvement going C → L, you'd still need to input a 1000 lines of L-code. How would you do it? With character recognition? OTOH… I am ready to bet that on your keyboard maybe US-104, maybe something more exotic: - There is a key that is marked something that looks like 'A' - Pounding that 'A' produces something that looks like 'a' ' And to get a 'A' from the 'A' you need to do a SHIFT-A IOW its easy to forget that typing ASCII on a us-104 still needs input-methods Its just that these need to become more reified/firstclass going from the <100 chars of ASCII to the million+ of unicode. Or if I may invoke programmer-intuition: a. If one had to store a dozen values, a dozen variables would be ok b. For a thousand, we'd like a -- maybe simple -- datastructure like an array/dict c. For a million (or billion) the data structure would need to be sophisticated The problem with unicode is not that 100000 is a large number but that we are applying a-paradigm to c-needs. Some of my --admittedly faltering -- attempts to correct this: http://blog.languager.org/2015/01/unicode-and-universe.html http://blog.languager.org/2015/03/whimsical-unicode.html http://blog.languager.org/2015/02/universal-unicode.html -- https://mail.python.org/mailman/listinfo/python-list
