Gordon Ross wrote: >> I don't see it. The old code used to _decode_ the error class into a >> human-readable string, and the new code seems to just print decimals. > > Ah, I get what you mean. Well... it used to "decode" nonsense. > There are basically no clients left in the world using the > old-style error class/code format, so there's no value in > trying to maintain that decoding. It's also more consistent > to just print the number(s) for both error formats. Both > can easily be found by searching MSDN. (Or use wireshark:) > > That's the thing: We really don't want to be investing much > in this snoop decoding now that wireshark is around, but this > particular bug has caused too much confusion to not fix.
I was confused why you were removing that from the part of the code that prints out legacy non-NT errors. Even if it would just be encountered in archived snoop files and not in real life, it's surprising to see it removed. Obviously, it shouldn't be enabled when decoding packets that are known to be in the new format. I was expecting that part. But, ok, if you feel it serves no purpose and either no user would ever see the difference or the difference would actually be a positive one (by removing confusion when looking at old traces), then I'm ok with it. -- James Carlson 42.703N 71.076W <carls...@workingcode.com> _______________________________________________ networking-discuss mailing list networking-discuss@opensolaris.org
