[tekacs]

Scanning over the page and the 'explanation' page, I found a headache developing as I tried to read everything there.

My experience so far is that I find this far harder to read. For reference's sake, I don't do parameter alignment in my own code (see Stratoscope's comment), but still find it easy to scan code in most-any monospace, probably because my eyes just know where to jump to all the time...

For all the words in favour of proportional fonts, I would venture that when the words on a page are in fact ultimately forming macroscopic symbols (identifiers), often with PascalCase or camelCase or underscore_words, uniformity no matter the characters used is perhaps a more useful feature than making the words easily comparable to those read in prose?

[krick]

Agreed. If somebody likes that — well, that's nice. I even see on screencasts sometimes TextMate with proportional fonts, so obviously somebody will find it desirable. But it won't be me for sure. First off, what I've seen there doesn't seems readable by itself. Second, I don't really see any real benefits of using proportional fonts for coding, while drawbacks are quite big: when I'm tired or just reading quickly through the code I often find myself depending on being able to see that lines with the same number of characters have the same width.

Even more than that (maybe it's something like "professional deformation" already, but I'm fine with it): I've considered using non-monospaced fonts automatically for stuff like LaTeX or markdown, or my own txt notes, but didn't find it pleasant enough. For all the european languages, anyway — typing japanese in vim is a bit painful for me and I don't know languages with more exotic writing system, be it arabic or hindi. I even hard-wrap natural language texts, which would be crazy using proportional font — and am glad I did so, when I want to see it through `less` on fullscreen.