amd64 was very company specific and as far as I knew, we were favoring x64, but I have no objection to x86_64. But in some places, like the jre/lib/ARCH name, we are kind of stuck with amd64, but now that I think about it, I'm not exactly sure why we would be stuck with any directory name like this.
I think where we can be consistent and avoid adding new names for the same thing, we should. Just not sure this will be floating up as any high priority effort any time soon. -kto On Aug 15, 2012, at 1:52 PM, Magnus Ihse Bursie wrote: > On 2012-08-15 10:38, Dmitry Samersoff wrote: >> Hi Everyone, >> >> Currently we uses i386, i486 and i586 within x86 build. > Not only the 32-bit Intel architecture, but the 64-bit as well has naming > confusions. > > Unfortunately, it is not that simple to just replace everything with x86. > > 1) Some things have external requirements for their name, and cannot be > easily changed (or at all). For instance, Java programs expect a specific > os.arch property. Library names must be named amd64 on Solaris (as far as I > understand). Marketing is likely to have their own view on what to name the > packages. Etc. > > 2) Changing names on files or directories (unfortunately) makes version > control harder. If files are to be moved around all over the place anyway > with Jigsaw, we might as well fix some names. It is not clear if it's worth > the effort to rename files before such a major restructuring. > > 3) If we should change all instances that we can, we still have to agree on > *what* to call it. Not everyone thinks that "x86" is the obvious choice. And > even if this is the most common view (I think it is), what about the 64-bit > platform? x86_64? x64? amd64? I've heard people defend all three names. > > I actually posted a proposal for naming cpu's in the new build infra some > time ago (see > http://mail.openjdk.java.net/pipermail/build-infra-dev/2012-June/001019.html). > It boiled down to: > 32-bit Intel: x86 > 64-bit Intel: x86_64 > Intel arctitecture: x86 > > (As reference: > 32-bit Sparc: sparc > 64-bit Sparc: sparcv9 > Sparc architecture: sparc) > > There was some follow-up discussion, mostly on how to identify what different > kinds of names we need. > > /Magnus
