Hacker News Reader

Comment by jpfr | Hacker News Reader

This is all good and well. But why XeTeX as a starting point instead of LuaTeX?

Most distributions started the transition from pdfTeX to LuaTeX as the main implementation. And the code from the new translation from the original Pascal to C is pretty nice.

And LuaTeX also has UTF8 support and modern font handling.

View on HN

> And LuaTeX also has UTF8 support and modern font handling.

I always thought this was the whole point of XeTeX?

I thought the point of XeTeX was bidirectional support (e.g. Arabic and English text in the same line), hence the palindromic name.

LuaTeX is a kind of successor to XeTeX. Biggest thing is that xelatex / latex / etc all had to use fixed size memory management while luatex supports dynamic allocation (which means no more "TeX capacity exceeded, sorry." dreaded error message)

Dynamic memory allocation is not always a desirable feature. If I were running a service like Arxiv, I would much rather have fixed memory management than dynamic.

Similarly, if I was running TeX in web assembly, in a tab in a browser I'd much rather hit "TeX capacity exceeded, sorry." than have it try to scuttle off with all the swap...

> UTF8

I remember reading that the unicode-enhanced Computer Modern isn't exactly the same as the original (latin) Computer Modern that Knuth designed with the former having a few cosmetic issues. If that is indeed the case, is it possible to use the original Computer Modern with a UTF8 engine?

It's been a bit since I tested the two, but for my purposes, XeTeX was twenty to twenty-five percent faster in the worst-case scenarios.

Reply on news.ycombinator.com