On Monday, 15th March 2004, Kris Kennaway wrote: >On Sun, Mar 14, 2004 at 05:16:13PM +1000, Stephen McKay wrote: >> The change to 64-bit time is essential, of course, but I don't understand >> why it has to break backward compatibility. Surely you just allocate a >> bunch of new system call numbers (for the 64-bit variants) while keeping >> the old ones (so 32-bit time calls still work) and bump the version >> number of every library. What else is going on? (I don't have a Sparc >> or I'd join your experiment.) > >No-one donated their time to do it that way. I don't think that's relevant. The question is whether it's the right way to do it or not. If what I've suggested is technically correct (and that's what I believe) then that's how it should be done. Backward compatibility is very important and can be ignored in only a few cases (eg the switch from a.out to elf, or a port to a new architecture). Also, this is the first I've heard of this since I have no interest in sparc. If the intention is to use the sparc conversion is as the template for architectures I care about then now the first time I can contribute to improving the process. Stephen.Received on Sun Mar 14 2004 - 16:34:09 UTC
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:37:47 UTC