| To: | yyyyyyyyyyyyyy@xxxxxxxxxxxxx |
|---|---|
| Subject: | Re: Exact types and non-8-bit-bytes Position Statement |
| From: | Tom Lord <yyyy@xxxxxxx> |
| Date: | Fri, 16 Feb 2001 01:03:45 -0800 (PST) |
| Cc: | yyyyyyy@xxxxxxxx, yyyyy@xxxxxxxxx, yyyyy@xxxxxxxxxxx |
>>> My basic reaction is that I don't want us to suffer through the >>> decade of >>> pain that changing POSIX bytes to a length other than 8 bits will >>> entail. > >> How much pain do you think there is ? Apart from anything else, >> I can't see >> anything outside networking that's even affected. > Just the char = byte hassle. Uh...no. Sort of. An unbounded number of high-performance algorithms depend on the assumption that the C type `char' is an eight bit type, and that other integer types are power-of-2 multiples of that. I hope (presume, even) that that's something HW designers think about. You could argue that the OS layer shouldn't care about such algorithms, though. with apologies to Knuth and MIX, -t |
| Previous by Date: | RE: Exact types and non-8-bit-bytes Position Statement, Joe Gwinn |
|---|---|
| Next by Date: | Re: Bug in xbdd5 Integer types/networking, Andrew Josey |
| Previous by Thread: | Re: Exact types and non-8-bit-bytes Position Statement, Eric Vought |
| Next by Thread: | Re: Exact types and non-8-bit-bytes Position Statement, Clive D.W. Feather |
| Indexes: | [Date] [Thread] [All Lists] |