Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!osborn From: osborn@cs.utexas.edu (John Howard Osborn) Newsgroups: comp.sys.next Subject: Re: 88110 rumors Message-ID: <20692@cs.utexas.edu> Date: 17 Jun 91 21:03:40 GMT References: <1991Jun12.200656.10243@linus.mitre.org> <1991Jun14.172316.10315@math.ucla.edu> Organization: U Texas Dept of Computer Sciences, Austin TX Lines: 20 Concerning the possibility of NeXT shipping an 88110 based computer, I'd like to note something that Apollo did to help solve the problem of incompatible machines. Apollo had a 68xxx-based series and an 88xxx-based (modified, called PRISM) workstation. Obviously, these machines were not binary compatible. But, the operating system let you build compound executables. The resulting file had a 68xxx code segment, an 88xxx code segment, and a single data segment. As a result, your code grew by about 33% in size (roughly), but a single program would run on either machine. Now, suppose NeXT ships an 88xxx based machine along with NeXTstep 3.0. Using this technology, developers could ship a single executable that would run on either the old 68xxx machines or the newer machines. Thus, the transition would be much eased. - -John H. Osborn -osborn@cs.utexas.edu