nobu.nokada / softhome.net wrote:
> Hi,
> 
> At Thu, 2 Dec 2004 15:51:46 +0900,
> Yukihiro Matsumoto wrote in [ruby-talk:122141]:
> 
>>Nobu himself made a patch to preserve hash order.  I have not decided
>>yet to merge it.  The only concern is performance.
> 
> 
> The performance wouldn't increase for insertion, iteration and
> lookup, but do for direct deleting (except for st_delete_safe).
> And the memory usage increases 2 pointers for each hash
> entries.
> 

In my opinion, it would be worth it.

I am curious: How often do people use "large" hashes? Nearly all
of mine are under 50 keys, I think.

It becomes more of an issue if the hash has 1000000 keys, of
course.


Hal