[followups to trademark-policy-dev]

I. Szczesniak wrote:
> On 10/31/07, Alan Coopersmith <[EMAIL PROTECTED]> wrote:
>> I. Szczesniak wrote:
>> No one has ever promised /usr/gnu/bin would be POSIX compliant.
> 
> I disagree. POSIX violations are BUGS and have to be fixed, otherwise
> interoperability between systems is completely gone. Or are you saying
> that Sun intends to ship the GNU tools as cheap 'outsourced'
> alternative to the Sys5 tools and then refuse support, even in
> violation of our support contract? 

IANAL.

I'd be very surprised if the definition of support for these tools
is "Sun will rewrite them to do things that they don't do (and aren't
documented to do) in the reference version available from the supplier".

In fact, since this is an open source discuss list, and we are talking
about the GNU toolset, you might recall the longstanding tradition of
"if you find something that you don't like, you are empowered to fix
it and encouraged to contribute your work back to the community".


The point where your comments are a bit on-topic is in the compatibility
discussion we are having over in <trademark-policy-dev>.  Obviously,
simply having a tool called "/usr/bin/tar" isn't enough to be
"compatible", you also need to adhere to some behavioral standard.

The real question is "which standard?"

   Ship "Sun's old crufty Sys5 tools" in /usr/bin and you alienate
   the GNU-Linux crowd who consider being different from GNU to
   be a bug;

   Ship the "cheap 'outsourced' alternative GNU tools" instead,
   and you now have the "must have POSIX" crowd up in arms.

Maybe the best answer is to go work on a project that delivers the
best of both by doing a concerted slice/dice/merge job on both
source bases - and contribute back the result to both parties...

   -John (who still NAL)

_______________________________________________
indiana-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/indiana-discuss

Reply via email to