Using notes from this thread and the referenced pages, I image-updated build 111b to 132, installed osnet, redistributable, and sunstudio from the 2009Sep tarball to set up for later builds. The image-update yielded SUNWonbld at ..0.132, and I secured the core tarballs and crypto bits from the osol/on download page for b132.
On the first boot after all this, the system (after a gratuitous bootadm update-archive), came up and ran just fine, thank you =] I did hit a few of the expected bugs that David Comay pointed out in the indiana-discuss list at 2010-February/017516, but these were either easy to work around or were trivial. FWIW, I suspect (guess?) that bug 13534 may have been responsible also for the 'StaticSeat1' directory being placed in /, owned by root:gdm, which I think would otherwise have been put in /var/lib/gdm, had I done the workaround for the bug before booting. Also, I hit bug 14135 (which I don't think David called out), a trivial bug involving the xdt driver not being applicable to the i86pc platform. Workaround == ignore. Now at b132, the system runs fine. My only problem is that attempts to build the pristine b132 ON tree fail, and fail in a way that suggests I may have a toolchain mismatch of some sort. All relevant packages that I know of are at 132, the patched sunstudio product is 2009Sep, and everything else is from the b132 download page. I'm using nightly(1) in the "normal" way (a la the README that unwraps into the workspace). Since around Flag Day 2005 I have never had a problem building the pristine ON tree, but those were all Nevada builds against itself at the same level that was shipped. If anyone sees an obvious omission I made here, I'd appreciate some input! --Jim -- This message posted from opensolaris.org _______________________________________________ opensolaris-discuss mailing list [email protected]
