On 03.08.2017 10:24, Maxime Villard wrote: > Alright, thanks for the archeological digging. > > I still I'm not sure what exactly this all means: if compat_ibcs2 is in > fact > compat_svr3, and if the real iBCS2 implementation is in fact > compat_svr4, what > is compat_svr4 doing on sparc/sparc64? > Our IBCS2 is SVR3 (compat_imbcs2) and SVR4 (compat_svr4). SVR4 is also special as SunOS + SVR4 = Solaris. SVR4 is not limited to Sun's products. Solaris runs on x86 and sparc(64). This OS is still used as a reference for BSD development. > Again, iBCS2 is a standard for x86; does this standard happen to be > compatible > with the sparc-like SVR4? Sorry if I'm being dumb, but this does not seem > obvious to me at all. SVR4 is supported on x86, SPARC and certainly more hardware platforms of commercial Unices. SystemV was a competing branch of Unices with BSD (like STREAMS vs sockets etc). Today both are mostly superseded by Linux... which attempted to handle SVR4 and BSD ancestry this or that way. SVR3 and SVR4 are compatibility layers for commercial Unices, this makes it difficult for maintenance as binaries and toolchain are licensed. > > Maxime
Attachment:
signature.asc
Description: OpenPGP digital signature