Gearing up for 64-bit time_t on sparc64

[Tillman Hodgson]

On Sun, Jan 18, 2004 at 10:06:11PM -0500, Garance A Drosihn wrote:
> A few people have tried my script and followed the instructions
> in my special "UPDATING.64BTT" file.  No one as reported any
> problems to me, and at least a few people have reported that it
> has worked for them.

That's good to hear.

> I agree with your view, but I am not sure what is appropriate for
> me to do at this point.  I could just commit the "installworld_nk"
> and "UPDATING.64BTT" files to /usr/src, but it wouldn't surprise
> me if that caused a few people to scream at me.

Heh ;-)

> Also, as mentioned in the UPDATING.64BTT file, we need to have a
> binary-version of 'cvsup-without-gui' available, one which has been
> compiled on a 64-bit time_t system.  I have one of those, so it's
> a matter of where should we put it for people who will need it.

I'm waiing for such a plan to be announced, myself :-)

While I don't mind taking a /few/ risks with my Ultra 5, it runs a
public website and I'd like to minimize downtime. I also have 137 ports
installed and I'll need to rebuild them all (for safety) and having a
working cvsup is definitely a benefit.

That reminds me ... I've trying to move to a system where my i386
-STABLE build host holds the ports tree for my -CURRENT machines as well
(to offload the cvsup and INDEX building, but not the package building).
I've received a couple of great hints from the stable@ mailing list on
this. I wanted to confirm (if possible) that the move to 64bit time_t on
Sparc doesn't/shouldn't impact on this. Any opinion on this?

Thanks,

-T


-- 
If you are not happy here and now, you never will be.
	Taisen Deshimaru