Robert Bradshaw wrote: > On Oct 22, 2009, at 4:45 PM, Dr. David Kirkby wrote: > >> Peter Jeremy wrote: >>> On 2009-Oct-16 13:05:02 +0100, "Dr. David Kirkby" <david.kir...@onetel.net >>>> wrote: >>>> I'm updating a configure script which will warn users if their >>>> operating system >>>> is too old. In the case of a Solaris 9 machine it will say: >>> Whilst I agree in principal, I have some concerns about the wording. >>> >>>> configure: WARNING: You are using Solaris 9, which is a very old >>>> version >>>> configure: WARNING: last released in 2004. Solaris 10 was first >>>> released in >>>> configure: WARNING: 2005, and there have been many updates since >>>> then. You >>>> configure: WARNING: are advised to upgrade to Solaris 10 or later, >>> This is all very nice in principal but may not be possible in >>> practice >>> due to requirements for other applications. Eg, we still have a >>> Solaris >>> 2.6 server to support an old application. Whilst the application >>> might >>> work with Solaris 10, that would require effort to upgrade the OS and >>> then regression test the application - for which there is no funding. >> I had changed the wording somewhat anyway, but I would admit I had not >> considered other applications. >> >> I also added a check for the old sun4m and sun4c systems, as they >> can not be >> upgraded to Solaris 10. >> >> I'm well there may be reasons for sticking to an older OS, though >> Solaris has >> excellent binary compatibility. Obviously there are exceptions, but >> generally it >> it runs on version N, it will run on N+1, N+2 etc. >> >>> I suspect you will find that Solaris 9 is still in fairly common use. >>> Whether it's used widely enough to justify a Sage release, I'm not >>> certain. (And I'm not certain how much effort would be needed to >>> build Sage on Solaris 9). >> I would agree Solaris 9 is in quite common use. If Wikipedia is to >> be trusted, >> Solaris 9 will be supported by Sun until 2014, and even Solaris 8 to >> 2012. So >> neither should be written off. >> >> The current wording I have is: >> >> if test x$build_os = 'xsolaris5.0 || test x$build_os = >> 'xsolaris5.1' \ >> || test x$build_os = 'xsolaris5.2' || test x$build_os = >> 'xsolaris5.3' \ >> || test x$build_os = 'xsolaris5.4' || test x$build_os = >> 'xsolaris5.5' \ >> || test x$build_os = 'xsolaris5.5.1' || test x$build_os = >> 'xsolaris5.6' \ >> || test x$build_os = 'xsolaris5.7' || test x$build_os = >> 'xsolaris5.8' \ >> || test x$build_os = 'xsolaris5.9' > > Is there not a way to say \le solaris6.x? Listing all blacklisted > systems explicitly seems both brittle and cumbersome.
Cumbersome I would agree. Brittle I do not. It's complicated by the fact that some are x.y releases, and some are x.y.z. Solaris 10 (latest version) is SunOS 5.10, so 6 does not come into it. There are currently no 6.x versions, and I do not see anything to suggest that will change in the near future. SunOs 5.0 (Solaris 2) was bought out in 1992, and every Solaris release since then has been a SunOS 5.y or a 5.y.z If this was a simple number, a comparison would be somewhat easier, but for example how does one differentiate in autoconf between 5.1 (Solaris 2.1) and 5.10 (Solaris 10) and determine that 5.10 is newer than 5.1? In theory if one could extract the second part of this (and sometimes there are two, and sometimes 3 parts), then it may be possible. I think trying to write a macro to do this would probably introduce more problems than it aims to solve. I will in any case test this configure script on Linux, OS X, AIX, HP-UX, Solaris and perhaps someone else can test on FreeBSD, as I do not have access to a FreeBSD machine. Dave --~--~---------~--~----~------------~-------~--~----~ To post to this group, send an email to sage-devel@googlegroups.com To unsubscribe from this group, send an email to sage-devel-unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URL: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---
