libg++ as a checkin requirement?

[Robert Lipe <robertl at dgii dot com>]

H.J. Lu wrote:

> may be related. It will be nice if people run test on libg++ before
> they check in changes. I have suggested it many times before. It went
> nowhere. The result is everyone wastes time on those bugs which can be
> caught by libg++. Should that be a check-in requirement?

If it is to be a check-in requirement, then we're going to to have dust
off libg++ and fix some (admittedly minor) portability problems with it.
It should probably be made part of the CVS tree.  We'd need to find the
dejagnu test stuff that was floating around for a while.  In all, we'd
basically have to accept that the package isn't dead.

I do run libg++ tests once in a while since you suggest we do so.  But
they're so clearly a second-class citizen that it's just too frustrating
to do it frequently.

RJL