>> Anyway, as of my webrev changes earlier today, the onnv version of >> webrev will only work with ksh93. This works on OpenSolaris systems, >> where /usr/bin/ksh is ksh93. It breaks on Solaris Nevada systems, where >> /usr/bin/ksh is ksh88.
> Erm... why is this a "flag day" ? OS/Net has ksh93 since Solaris 11/B72 > and right now SXCE won't even boot without /usr/bin/ksh93 (don't worry, > ksh93 is (like ksh88) delivered via the "SUNWcsu" package and therefore > available on all systems). I know that ksh93 is widely available since snv_72. And I'm clearly taking advantage of that. Perhaps it would have been more clear if I had said "flag day for anybody using onnv-tools from Solaris 10 or earlier systems." In the two hours following my integration, at least four different people installed a new SUNWonbld and experienced webrev failures. I introduced a new incompatability: up until today, folks were able to use webrev on S10 systems. Now they're not. In spite of having broken no big rules, I have still changed the minimum system requirements to use webrev. Obviously, "we" (I discussed this with the current ONNV core team) decided that this was acceptable. But it's still appropriate to tell people when you're knowingly breaking tools that they use regularly. --Mark
