Hi Ingo,
> seemed to indicate that the execution time benefit is roughly
> consistent with what one would expect if it were mainly due to the
> simple reduction of the macro file size:
The rule of thumb I've often heard is the lexer is the bottleneck. I
haven't time to dig right now, but I've one more datum to throw in.
$ strace -fc man bash >/dev/null
% time seconds usecs/call calls errors syscall
------ ----------- ----------- --------- --------- ----------------
42.86 0.439904 14663 30 4 wait4
26.86 0.275675 13 20625 2 write
25.39 0.260614 13 20443 read
...
That's a large page, and it's measuring man, tbl, col, and the rest of
the gang, not just nroff processing macro files, but it shows I/O
dominates.
> a) You mean that savings of 5-20% in execution time are
> "significant", even though formatting with mandoc would save about
> 60-90% of formatting time instead?
Yes, because we're not moving to mandoc. We are going to format with
nroff and troff because we like it. :-)
--
Cheers, Ralph.
https://plus.google.com/+RalphCorderoy
