Eric Blake <ebl...@redhat.com> writes:
> On 11/23/2015 07:27 AM, Paolo Bonzini wrote:
>>
>>
>> On 20/11/2015 18:32, Eric Blake wrote:
>>>> static void qlist_size_iter(QObject *obj, void *opaque)
>>>> {
>>>> size_t *count = opaque;
>>>> (*count)++;
>>>> }
>>>
>>> Yuck - we don't track size independently? Seems like it might make a
>>> worthwhile addition,
>>
>> Would you change your mind, if I told you that qlist_size is unused? :)
>
> Deleting dead code is a perfectly acceptable alternative to advertising
> what should normally be an O(1) operation with an O(n) implementation. :)
Well, "length of list" certainly isn't O(1) everywhere. The Common Lisp
Hyperspec, for instance, gives an O(n) example implementation[*].
Generally just fine as long as callers are aware.
[*]
http://www.lispworks.com/documentation/HyperSpec/Body/f_list_l.htm#list-length
