>Date: Thu, 09 Nov 2006 15:34:37 +0000
>From: Darren J Moffat <Darren.Moffat at Sun.COM>
>
>Don Cragun wrote:
>> The project team is not trying to set precedent by doing this, but I
>> would expect it to be used frequently in similar cases to avoid
>> backwards compatibility problems.
>
>Is there any reason why we shouldn't set precedent here ?

I don't think we need a precedent to use due diligence.  If a project
team or any review committee (including PSARC) believes that moving or
removing a utility will break applications, the project team either
needs to install a link (hard or symbolic) or find and fix applications
that may be affected by the change.  This just seems to me to be common
sense.

>
>Given what you said about standards compliant applications being allowed 
>to use the full path I think that says that we always want these 
>symlinks when we unify a /usr/bin and /usr/xpg?/bin utility otherwise 
>things will break.

If we end up removing a /usr/xpg*/bin utility due to a fix required by
an interpretation request (which intentionally changes behavior), a
project team may decide not to install a symlink.  (I'm not concerned
about any precedent if this issue comes up; any project that does this
would be granted an exception if needed.  I'm just saying that "always"
may be too strong.)

>
>In case it isn't clear I think this case should be setting precedent.

Is this a formal request to turn this into a fast track to set a
precedent?  (If I remember correctly, a self-reviewed case can't set a
precedent.)

 - Don
>
>-- 
>Darren J Moffat


Reply via email to