Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!usc!aero-c!gumby.dsd.trw.com!deneva!news From: thomsen@spf.trw.com (Mark R. Thomsen) Newsgroups: comp.sys.next Subject: A laptop NeXT + FSF NeXTstep Message-ID: <282AE226.3250@deneva.sdd.trw.com> Date: 10 May 91 18:10:46 GMT References: Sender: news@deneva.sdd.trw.com Organization: TRW Inc., Redondo Beach, CA Lines: 25 Chris Petrilli writes I might be able to address IBM and NextStep. It'll most likely never happen. IBM has said they are going to stay with the AIX kernel (ack!), and not go to Mach3 under OSF/1. NextStep 2.0 is VERY VERY reliant on the Mach kernel for speed, and simulating it on an AIX machine is VERY SLOW. It is bad on a RS/6000, imagine it on a 386. The best bet is Next being nice enough to give NextStep to the FSF, and let us run it on GNU (based on Mach3). One of the problems we have had with RIOS is that it is slow, when measured by us, for context switching. Putting software for UI (e.g., NeXTstep) that relies on context switches on top of a full-bloated UNIX kernal (AIX) and the problem is 'bad'. The '386 context switches a bit faster. But without commitment from IBM to make a wonderful NeXTstep on their platforms, I guess the portable- NeXT-as-PS/2 is unlikely to fly. NeXT give NeXTstep to FSF? This would be a wonderful, weird thing. How likely is this? NeXTstep is one of the technologies that NeXT has invested in that is superior to competition. It gives them a difference that is valuable to users. Giving such away would be a complex decision, I imagine, for NeXT. Mark R. Thomsen