Garrett D'Amore wrote:
> I think it doesn't qualify for automatic approval. I suspect the
> special numbering will be the cause of some consternation. I'm not sure
> what number we're up to on our patching, but when will we run into the
> problem where our patch numbers don't work?
Patch IDs currently start with a leading '1' and we at almost
40,000 right now, so we've made ~40k patches over the course
of Sun's life. I'm not sure when patches were introduced; 20 years
ago? Maybe someone I cc'ed knows. Since patch IDs started with 1 and
we've carved out 9, that leaves ~760k to go, so while this was discussed,
it was not considered to be an issue.
> What about expanding on the idea used by T-patches, where a special
> prefix is used instead... ("Sxxxxxx" or somesuch), that lives outside of
> the numeric space associated with patch numbers?
This was discussed among the group but was not the chosen
alternative.
> Of course, since we have no architecture (nor product!) for patching
> anything more recent than S10, maybe its not an issue. (It seems like
> IPS changes the paradigm of patching in non-trivial ways, such that it
> isn't clear to me if it makes sense to patch a Nevada system using
> normal patches. If we ever release a Solaris 11 built upon SXCE instead
> of OpenSolaris then the problem may remain...)
>
> If the idea of using a special patch prefix letter (outside of the
> numbering space) is not acceptable for some reason, then I'm willing to
> hold my nose and give this a +1. I'd still like confirmation from the
> project team as to whether the idea was considered, and if it was
> rejected, I'd like to know why.
I don't recall all of the reasons why this idea wasn't the
preferred solution, but I cc'ed most everyone who was involved
in these discussions so hopefully one of them will chime in.
Thanks,
Jerry
