On Aug 7, 2007, at 12:04 AM, Craig McClanahan wrote:
> > The Internet has a vast number of OS packages available as source, > easily compiled and installed with the typical "./config ; make ; make > install" pattern. Yes, binary packages are nice for end users, but > they are useless to most developers who want to stay on the cutting > edge of at least packages relevant to what they are working on. > > And the vast majority of such packages, IMHO, have Linux-ish > assumptions in their build scripts. > Recently I was at a customer visit where the people buy $100M of equipment at a time. All Linux. The context of the visit was a multiples of the usual purchase size. The OS requirement was summed up as "autoconf should just work" viz. if it looks and smells like linux to the extent of building software to a first approximation, they would consider it. If it didn't, it would be a huge uphill battle. I believe that the point of Indiana is to appeal to this large (both in numbers, and in financial impact) class of customer. I think this should be clearly articulated in our high order project descriptions... and should trickle down to all decisions. I am not insensitive to what this does downstream, we will have to pick technical solutions which allow a top level choice to be made (either at installation per user, per partition, etc.) whether the defaults should be traditional Solaris or Indiana. But I don't think that future pain and complexity gets us away from the reason for doing Indiana. > Keith H. Bierman [EMAIL PROTECTED] | [EMAIL PROTECTED] Strategic Engagement Team | AIM: kbiermank 5430 Nassau Circle East | 650-352-4432 voice+fax Cherry Hills Village, CO 80113 | 303-997-2749 http://blogs.sun.com/khb | <speaking for myself, not Sun*> Copyright 2007 _______________________________________________ indiana-discuss mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/indiana-discuss
