[lltx] Re: luamicrotype?

Khaled Hosny khaledhosny at eglug.org
Wed Mar 3 05:48:20 CET 2010

On Wed, Mar 03, 2010 at 02:06:34AM +0100, Robert wrote:
> >So why tie ourselves with legacies, where all what we
> >need, to get microtype support in fontspec, is few tens of lines written
> >in an elegant, easy to understand programming language? (most of the
> >needed work is already done in luaotfload).
> I must admit that I'm a little bit puzzled as to what you're
> actually proposing. You cannot be seriously suggesting that 4000
> lines of code ("legacy" or not) and another 3000 lines of
> configuration can be replaced with "a few tens of lines"? With that
> you may be able to provide rudimentary support along the lines of
> good-old protcode.tex. The bulk of microtype, however, deals with
> the user interface, intended to be automatic, systematic,
> straight-forward and flexible. It's all there and time-tested, and I
> don't see why it should rewritten/replaced. Users won't be amused, I
> guess.

It can be very well my naïveness making me not realise the amount of
work needed. But, AFAIK, there is protrusion/expansion code already in
luaotfload (code extracted from ConTeXt) and what is missing is an
interface to it. Writing lua code is far far far straightforward than
TeX macros, so I'd even prefer a lua based solution even if it were 4
times the TeX macros solution. I'm not undermining the value of your
work, but, from maintenance point of view, I really doubt there are
many people, besides you, who can understand those ~7000 lines of TeX

I'm already struggling with fontspec having to support XeTeX too (I
really hate when I've to deal with complex, mind blowing TeX code that
one can replace by straightforward, easy to understand lua functions).


 Khaled Hosny
 Arabic localiser and member of Arabeyes.org team
 Free font developer

More information about the lualatex-dev mailing list