> Date: Tue, 26 Sep 2006 10:24:50 -1000 (HST)
> From: Joseph Kowalski <Joseph.Kowalski at eng.sun.com>
> Subject: Re: Issue Summary Re: Korn Shell 93 Integration [PSARC-EXT/2006/550
Timeout: 09/27/2006]
> To: PSARC-EXT at sac.sfbay.sun.com, April.Chin at eng.sun.com
> Cc: ksh93-integration-discuss at opensolaris.org, roland.mainz at
> nrubsig.org,
don.cragun at sun.com
> MIME-Version: 1.0
> Content-MD5: NL4l4bo/oqR2UDNp46rsVA==
>
>
> > From: April Chin <April.Chin at eng.sun.com>
> ...
> > 4) There was a lengthy discussion concerning merging the existing
> > Solaris libcmd Sun Private interfaces and the new AT&T libcmd b_*()
> > built-in command interfaces (Project Private now, but the project team
> > expects that they will be made committed Committed Public in a future
> > case) into the same libcmd library.
> >
> > The Solaris libcmd interfaces are widely used by many consolidations,
> > including unbundled products, so renaming the Solaris libcmd would
> > introduce an incompatibility that is difficult to resolve.
> >
> > AT&T will not be renaming the library nor changing its location.
> > External applications will want to use the AT&T interfaces, so the
> > project team does not want to rename the AT&T libcmd on Solaris
> > systems. The project team will keep the new interfaces merged with the
> > old ones in the existing libcmd on /lib.
>
> I'm not agreeing to this (yet). We will discuss it tomorrow it seems.
> I think all that needs to be said in e-mail has been said.
>
> However, because of all the discussion, I just want to verify the following.
>
> 1) This case approves these interfaces (libcmd::b_xxx) as Consolidation
> Private. Some later case may make them some form of Public.
libcmd::b_xxx are proposed for Project Private in the case.
Yes, some later case may make these interfaces Public.
>
> 2) It has been suggested that this case formalize the libcmd::def*
> interfaces as Sun Private. I know I agreed to this, but I may
> want to reconsider as Consolidation Private might help to limit
> the expansion of use. (Again, discuss tomorrow.)
ok..discussion tomorrow...
>
> 3) The mapfile (etc.) associated with libcmd will mark all of the
> interfaces it contains as SUNW_private (or whatever variant is
> appropriate). (Code review, not architecture, but considering
> the controversy, I want to be clear.)
Yes, all the libcmd interfaces are SUNW_private in the mapfile-vers.
>
> 4) These routines will NOT be documented in man pages and a reasonable
> effort will be made to clarify the private nature of these in any
> other documentation.
Agreed.
Thanks,
April
>
> - thanks,
>
> - jek3
>
