Stephen P Potter <[EMAIL PROTECTED]> writes: > What stops us from imposing order on this chaos? If they are currently > defined as not having any specific order, why can't we say they always > return in numeric || alphabetic || ASCII || whatever order we want? Because the fewer guarantees you make, the more efficiency you can get. The above would prevent a hypothetical future smart Perl interpretor from reordering your hash behind the scenes in another thread while your program is using it to optimize for the usage pattern that it's seeing, for example. If you have to guarantee a sorted traversal of the hash keys, your choices of data structures are *far* more limited. -- Russ Allbery ([EMAIL PROTECTED]) <http://www.eyrie.org/~eagle/>
- Re: RFC 84 (v1) Replace => (stringifying comma) wit... Piers Cawley
- Re: RFC 84 (v1) Replace => (stringifying comma... Damien Neil
- Re: RFC 84 (v1) Replace => (stringifying c... Russ Allbery
- Re: RFC 84 (v1) Replace => (stringifyi... Stephen P. Potter
- Re: RFC 84 (v1) Replace => (string... Ariel Scolnicov
- Re: RFC 84 (v1) Replace => (string... Russ Allbery
- Re: RFC 84 (v1) Replace => (stringifyi... Uri Guttman
- Re: RFC 84 (v1) Replace => (stringifying comma... Dan Sugalski
- Re: RFC 84 (v1) Replace => (stringifying comma... John Porter
- Devils advocacy (Re: RFC 84 (v1) Replace =>... Nathan Torkington
- Re: Devils advocacy (Re: RFC 84 (v1) Repl... John Porter
- Re: Devils advocacy (Re: RFC 84 (v1) ... iain truskett
- Re: RFC 84 (v1) Replace => (stringifying c... Dan Sugalski
- Re: RFC 84 (v1) Replace => (stringifyi... Jonathan Scott Duff
- Re: RFC 84 (v1) Replace => (string... Chaim Frenkel
- Re: RFC 84 (v1) Replace => (s... John Porter