6

I have been searching in vain for statistics about the runtime of pdfTeX, LuaTeX and XeTeX. My impression is that (ordering by increasing runtime), LuaTeX > XeTeX > pdfTeX, but exactly how big is this difference?

Martin Schröder
  • 15,156
Gaussler
  • 12,801
  • With luatex still a work in progress, the question seems premature. – Thérèse Apr 12 '15 at 20:32
  • Without a reference document the question doesn't make sense. E.g. here the lua solution is faster than the tex solution: http://tex.stackexchange.com/questions/87497/tikz-distribute-evenly-and-randomly-circles – Ulrike Fischer Apr 12 '15 at 20:59
  • I meant how fast LuaTeX is to compile an ordinary, Lue-free TeX document. Any document would do. – Gaussler Apr 13 '15 at 07:34
  • Does "lua-free" means without luaotfload and system fonts? A simple \lipsum-documents needs ~0.7s with lualatex, ~0.8s with pdflatex and ~2s with xelatex on my system. – Ulrike Fischer Apr 13 '15 at 15:18
  • I guess it means both with and without fontspec, because using system fonts is one of the main features of LuaTeX. In other words, I would also like to know if it is just fontspec that slows down LuaTeX. So far, from your numbers, it sounds that way. – Gaussler Apr 13 '15 at 15:25
  • But fast calculations (and other things you can do better with lua than with tex) are also one of the main features of luatex. Why don't you want to take into consideration only the one but not the other? – Ulrike Fischer Apr 13 '15 at 16:12
  • I would love to measure that as well, if someone can find a way to compare a Lua solution to a problem directly with a TeX solution. But I assume that is harder than to compare a pdfTeX document in Computer Modern with a LuaTeX document with, say, Linux Libertine. – Gaussler Apr 13 '15 at 17:15
  • 3
    It is easy to compare the compilation time of two documents. But it is up to you to decide if this comparision has any meaning and gives you the values you are interested in. Sorry but how can you expect to find some "exact" statistics when you own ideas regarding how and what to compare are so very vage? – Ulrike Fischer Apr 13 '15 at 18:43
  • I'm really not sure what a full answer could possibly be here! – Joseph Wright Dec 17 '15 at 13:58
  • @JosephWright Well, if the question is that bad, then why did you bring it back to life after half a year of inactivity? – Gaussler Dec 17 '15 at 14:01
  • @Gaussler I'm looking over unanswered questions and working downward by number of votes. This one is near the top: looking at it I'm not sure what an objective answer would be. – Joseph Wright Dec 17 '15 at 14:02

0 Answers0