On Wed Feb 16 2000 at 22:40:54 -0500, der Mouse wrote:

> Quite aside from the license issues...would it fix the speed problem?
> Ever since I went to a 1999-08-13 source tree, vi has been very slow to
> start, often to the point where emacs (my personal variant, at least)
> starts comparably fast.  (file also is slow almost to the point of
> uselessness; in that case I ktraced it and can say with confidence that
> the time is going into parsing the magic database.)
> 
> Since I tend to run old machines (SS1+, SS IPX, Sun-3/60, uVAX-II, etc)
> speed is a nontrivial issue to me.  (You lucky sods with half-gigahertz
> CPUs can't tell the difference between .1ms to start and 3ms to start,
> perhaps, but over here where it's the difference between .5s and 15s,
> it really matters.)

I think that it took at least 30s to start for editing the kernel
configuration file on a sparc IPX. After I had a new non-generic kernel
the speed was better and it took only about 5s, but it was still slow. I
never bothered to investigate, though.

-- 
Antti Kantee <pooka@iki.fi>