Redesign of dlang.org

[via Digitalmars-d]

On Wednesday, 23 April 2014 at 23:25:43 UTC, Andrei Alexandrescu 
wrote:
>> 1. you don't use a monofont and have to figure out what "cpl" 
>> is in "em"
>> or "rem".
>
> I'm talking averages.

It varies with the font used. :-P And if you use a webfont the 
rendering will be less legible than when you use a screen 
optimized system font. So if you want clarity, unicode coverage 
and performance you don't know which font you get. :-]

> For a given language (English in our case), the relationship 
> translates into average words per line.

Not necessarily… since groupings aren't neccessarily words. Also, 
you don't READ prose in programming documentation like a novel, 
if you are a proficient programmer. You scan, backtrack, read, 
scan, backtrack, read…

>> 3. quantitative human factors studies that go beyond your inate
>> capabilites tend to be full of severe methodological flaws.
>
> Uhm...

Uhm what? Large sections of psychology are fundamentally flawed 
as a predictive science, it's primarily interpretative.

Measures are contextual. Reading is a skill, you are better at 
what you are used to. Besides, reading efficiency is less 
important than reading comfort.

If studies are off by 50% then that suggest that there are NO 
HOPE of significant results here.

>> Waste of time.
>
> I doubt it.

You are better off setting something reasonable, test, modify, 
iterate. And if needed you can always provide a stylesheet 
switcher (takes 15 minute to implement).