Thank you for your detailed response, Alan. Also thanks for explaining what in these JDS diffs you were referring to (I had no nerves to look more than 10 secs). Also I had thought you may have (rightfully) misunderstood me because of my truncated misspelled msg. And hence expected that your diffs would show exactly the opposite. Not so, excellent! I must add: userland did not get rid of gcc support in Oracle's latest diffs. In theory the framework is still fully functional. The problem is that at more and more places in components/*/Makefile's some Studio (often new 12.3 or 12.4 only) CFLAGS and CXXFLAGS are being set locally for a specific module. But not in a clean manner and compiler clean, but carelessly into the global CFLAGS/CXXFLAGS (as if done by some newcomers who don't know how to do it)! And that's simply a pity. Because I find it dumb to change 500 Makefiles for such nonsense I have brought the cw to the next level. And I modified it so to be usable with userland (with which it normally doesn't have a thing to do). It now takes all gcc or Studio-flags and always defaults to using gcc. All options get translated or (where not possible) ignored. It meanwhile even works with some of the most sophisticated autotools scripts, one of few exceptions where is fails is Perl 5.20 or higher. Although it is by no means complete yet and I didn't even advance its logic. That's why I didn't simply upload it, but I must clean it. Now it is possible to cheackout any tip Oracle userland and only modify a small handful of things, then it builds mostly without interaction on OI/Hipster or (unpublished instable hacked) "OpenSXCE.next".
Thanksgiving (ah, what's that? Not celebrated here) : Also: Aha, I was not even aware that we had weekend, let alone any special days off / holidays. But you are right, on zerohedge I saw these crazy buyers on black friday almost killing each other for getting the cheapest things they never need (us-only), but never knew it is a public holiday. Whatever - in a few days I probably and hopefully respond with some first TB 42.0b2 diffs, with last week's KMS diffs replacing the post-modern gfxp functions and after that we can finally deal with the Illumos code which causes flash to crash (or rather: with removing it). After my alternative gfxp functions are online, I bet Larry Ellison will contact us here personally and tell us what's wrong. Then finally we also have the Sun/Oracle's DRM/KMS port functioning on lower than 11.1 %martin
_______________________________________________ oi-dev mailing list [email protected] http://openindiana.org/mailman/listinfo/oi-dev
