Re: JIT stealth font-lock tuning parameters

[Stefan Monnier]

> > > >      (setq jit-lock-stealth-time 16)
> > > >      (setq jit-lock-stealth-nice 0.5)
> > > Would you object to changing the defaults as the first approximation, and 
> > > adding the more sophisticated operation as the second one?
> > 
> > I would personally find 16 seconds for jit-lock-stealth-time rather
> > annoying.  The reason is that it makes the jit-lock-defer-contextually
> > refontification "very slow".
> 
> The logic behind that is that, as long as you type, the syntax of the
> modified text can not be trusted.  If you don't consider short breaks as
> ``real'' idleness, the effect on jit-lock-defer-contextually should not
> annoy, I think.

When I change the syntax in a buffer, I occasionally wait for those
3 seconds to see the result after refontification.  Waiting 16 seconds
would be a real hassle.

For the real stealth fontification, I don't care whether it's
3 seconds or 3 minutes, but I think that as long as defer-contextually
uses the same delay, we should keep it somewhat short.

> > Note that I find the current settings quite usable on my 266Mhz Pentium-2
> > so I'm not sure how important it is to change them
> All I know is that quite a few people complain.

Weren't all the complaints in the w32 version of Emacs only ?


        Stefan