httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Chuck Murcko <>
Subject Re: Possible bugs in current version
Date Tue, 01 Oct 1996 20:06:51 GMT
Brian Behlendorf liltingly intones:
> On Tue, 1 Oct 1996, Robert S. Thau wrote:
> >   Agreed. I think each platform mentioned in Configure should
> >   have a default value explicitly set and we then change the
> >   default to be what we want (yes/no, either is fine)
> > 
> > Well, what we have now is some hosts having an explicit default to 'no',
> > and others relying on the implicit default 'no' --- whereas some hosts,
> > perhaps most, want "yes" (because they *don't* have a Posix regex package).
> > 
> > This could be fixed by changing the general default (as I did), or by
> > changing the defaults for specific hosts, but it seems to me that *something*
> > has to change...
> So we have platform settings for
>   *-apple-aux3*  (=no)
>   *-sgi-irix*    (=yes)
> we are missing settings for platforms I consider common:
>   *-ibm-aix*
>   OS/2
>   *-hp-hpux*
>   *-linux*
>   all bsd variants
>   *-next-nextstep*
>   *-solaris2*
>   *-sunos4*
>   *-unixware*
bsd is no.
I'll have to run the compat tests on (old) UnixWare and new Solaris, but it
seems like it's only safe to stick with the SysV regex on those.

> Regardless of which default we choose, we should come up with settings for
> these, at least for solaris, sunos, linux, and bsd.  If we come up with those
> settings I'm comfortable with the default being "yes".  How does one test a
> given platform?
I guess build Henry's tester with the system lib instead of his.

Chuck Murcko	N2K Inc.	Wayne PA
And now, on a lighter note:
In 1869 the waffle iron was invented for people who had wrinkled

View raw message