Mike Parker wrote:
> Bill Baxter wrote:
>> Lars Ivar Igesund wrote:
>>> Bill Baxter wrote:
>>>
>>>
>>> svn co http://svn.dsource.org/projects/cashew/trunk cashew
>>>
>>
>> Wikied!
>>
> 
> DSource has always had a page describing this at http://www.dsource.org/site/svnclient.

Quoth the above page:

"Once you have a client up and running, you may access code via the instructions on the page of the project you wish to access."

--bb

Sean Kelly wrote:
> Chris Nicholson-Sauls wrote:
> 
>>
>> Anyhow, clearly your .drop() is running very fast!  I actually expected the opposite, and am surprised/impressed.
>>
>> Maybe you should release it into Cashew?  ;)  And maybe I should put more into benchmarking the entire Cashew array module.  There might be a few more placed it could be sped up.
> 
> 
> I'd be interested in seeing a test using C's memmove as well.  Though I'm not surprised drop beats the slice version, since that will allocate a new array for the data.
> 
> 
> Sean

Well, I just benchmarked the memmove() based approach, and it is indeed notably faster still.  Here's a typical sample from the results:

<Benchmark traverse vs memmove> Baseline 3.130000
<Benchmark traverse vs memmove> Time 1.760000 & 1.778409 versus baseline

(Where the baseline was the current .drop() and the contender was a .drop() using memmove().)  So, indeed, I am migrating Cashew to use memmove().  I may also modify .dropIf() to use it as well.

-- Chris Nicholson-Sauls