[omniORB] [PATCH] "Safe" shortcut calls

[Nathaniel Smith]

On Fri, May 10, 2002 at 10:54:35AM +0100, Duncan Grisby wrote:
> On Tuesday 7 May, Nathaniel Smith wrote:
> 
> > This is actually a patch that I sent to Duncan a bit over a month ago,
> > just a few days before we learned about the closing of the labs --
> > so, of course, he's had plenty of things on his mind besides reviewing
> > my patch :-).
> 
> I have indeed been rather too busy to look at it. I don't think it's
> sensible to try to integrate it with omniORB 4 before the release,
> since it involves tweaking quite a lot of things, and I don't have
> time to make sure it's not breaking anything. It will be fine to

That makes sense.  I was careful to make sure that all the changes
only affect people who explicitly turn them on (ie, none of my code
should get executed at all, unless people make special arrangements),
so I think it would be all right to include... but by all means, be
paranoid.  I can't say I wouldn't do the same in your position :-).

> include the patch in the patches directory of the distribution,
> though, so it's easy for people to try it if they want.

Nod.

> I'm not sure there's a huge amount of demand for this shortcut stuff
> anyway. Is anyone other than the Fresco crowd using it or thinking of
> doing so?

Well, "safe" shortcuts give considerably more guarantees than the
current shortcut code -- one hope I had is that a lot more
applications will be able to switch them on safely, without
significant changes to application code.  All it takes to turn them on
or off is a few line change to your POA policy setting code; they're
also safe to use in most any situation where you don't also want to
set some contradictory POA policy (like SINGLE_THREAD or something).
I also made sure that they work consistently across platforms, and
always give some speedup.

In other words, they're quite non-intruisive, don't trap you into a
particular design (since you can trivially stop using them if you
ever want to do something they don't support), and hence there may be
lot of people who wouldn't go out of their way for these, but would
still take advantage of the essentially free speedup if they were
available.

That's what I'm hoping, anyway.  Does this sound plausible to
anyone else?

-- Nathaniel

-- 
"Of course, the entire effort is to put oneself
 Outside the ordinary range
 Of what are called statistics."
  -- Stephan Spender

This email may be read aloud.