calumb: > Sun's plan has always been to base the next long term support release > of Solaris on a suitable snapshot of OpenSolaris. So, modulo anything > Oracle has in mind, S10 won't be the last Solaris release.
Well, correct me if I'm wrong but I don't think it's possible yet to install things like Oracle RDBMS or Oracle Tuxedo on Indiana yet, is it? It's only possible to install them on Solaris Express, am I correct? So, to me one of the many, many problems with discontinuing SXCE is that you are creating a divorce or maybe a "temporary separation" in the happy (and profitable) marriage that has existed between Solaris and the Oracle RDBMS for many years now. This has the potential to create more confusion around the future of Solaris that can be exploited by Red Hat, IBM, HP, Microsoft, Zemlin + the Linux foundation, etc. to create F.U.D. (Fear Uncertainty Doubt) around the future Solaris and to try to use the F.U.D. to move Sun customers away from Solaris and on to other platforms. This in turn is not going to be good for Solaris revenue or for Oracle revenue. Case in point: what if you are a sysadmin or a Fortune 500 company that is a prospective Sun / Oracle customer? Say that maybe you are a Oracle RDBMS user who is running your Oracle instances on Red Hat Enterprise Linux or HP-UX. But you have heard good things about this "OpenSolaris" project and how it is supposed to be better than Red Hat and you want to switch from Red Hat Enterprise Linux to OpenSolaris / Solaris Express, the first thing you would want to do is download the latest build of Solaris Express (or Indiana), try to install Oracle on it, and run some tests and benchmarks and do a cost / comparison analysis study, right? Well, how are you supposed to do this with Indiana if Oracle is only compatible with Solaris Express? If the sysadmin downloaded OpenSolaris Indiana 2009.06 and got enraged that he had problems assigning a static IP address to the server, IPS ran dog slow, had problems with being unable to deploy zones when he couldn't connect to the IPS repository, etc. etc. and on top of it all, couldn't get Oracle to install and work properly on Indiana (and this is supposed to be the preview of the "next" version of Solaris, and Solaris is supposedly the "best" OS to run Oracle on) then this would give the prospective Sun customer a poor opinion of Solaris / OpenSolaris and they would be inclined not to come back to Solaris / OpenSolaris for their Oracle needs in the future (which, like I said before, is a tragedy, because there are a lot of really good things in both Solaris Express and OpenSolaris Indiana). This is yet another case where having a stable build of Solaris Express that does well in benchmark tests, etc. and works well with Oracle products would be good for both Sun and Oracle's marketing image and revenue streams. Also, how do you think the Oracle engineers are going to react when they download Indiana for the first time and try to install Oracle on it? Any links to blogs or other evidence so far that shows a positive reaction to it? This is the kind of stuff we have to think about before discontuing SXCE and throwing all resources into Indiana (which has yet to prove itself as a viable server OS and as a worthy successor to "The King of the Data Center" - a.k.a. Solaris 10). Obviously I think Indiana is a good thing or I wouldn't be running it as my main desktop OS at work, and I think it was something necessary that had to be done to gain developer mindshare from the Linux community, but I don't think it's ready to take over from SXCE in enterprise environments and I'm skeptical that all the problems with it will somehow be magically resolved in two months time given that, for the most part, there hasn't been a whole lot of progress made towards resolving them in the last year. -- This message posted from opensolaris.org _______________________________________________ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
