Re: sibcall/tailcall terminology

[Phil Edwards <phil at jaj dot com>]

On Tue, Sep 17, 2002 at 08:55:46AM -0700, Richard Henderson wrote:
> On Tue, Sep 17, 2002 at 07:59:10PM +1000, Fergus Henderson wrote:
> > Proposal A:
> > I propose that we use "tail call" to mean (3)
> > and "sibling call" to mean (2).
> > 
> > As a consequence:
> > 	- sibcall.c should renamed tailcall.c
> > 	- -foptimize-sibling-calls should be renamed -foptimize-tail-calls
> > 	  (or perhaps -foptimize-tailcalls?)
> > 	- try_tail_call in calls.c should be renamed try_sibling_call
> 
> While I agree that this would be more consistent, I don't think
> it's worth renaming files.

Then it should at least get a big hefty comment at the top, pointing
out the descrepancy between the file's name and the file's contents.
Perhaps a link to Fergus' message in the archives.


> And certainly we can't change the
> user interface without pain.

I would submit that this -foption isn't as well-known as others.  We could,
at minimum, do the traditional "provide both, warn & deprecate on old"
before removing the old one.


Phil

-- 
I would therefore like to posit that computing's central challenge, viz. "How
not to make a mess of it," has /not/ been met.
                                                 - Edsger Dijkstra, 1930-2002