[PD] Efficient FIFO in Pd: How?

[Roman Haefeli]

On Sun, 2015-11-22 at 00:47 +0100, Christof Ressi wrote:
> Just in case you don't know: There's the [fifop] object in zexy which works with lists and also lets you set the priority. I'm sure it's also faster than anything one can do with Pd objects. 

Thanks, somehow I forgot that zexy has that. It is indeed fast (as in
linear fast). I'm going to use it.

Just out of curiosity, I wonder if there is a O(n) vanilla
implementation, too. 

Roman


> > Gesendet: Sonntag, 22. November 2015 um 00:24 Uhr
> > Von: "Roman Haefeli" <reduzent at gmail.com>
> > An: Pd-List <pd-list at lists.iem.at>
> > Betreff: [PD] Efficient FIFO in Pd: How?
> >
> > Hi all
> > 
> > I'm looking for a way to implement a FIFO buffer in Pd. I'd like it to
> > be able enqueue messages with an arbitrary number of elements in order
> > to retrieve them in the same order later. I came up with three different
> > methods, but they are all either slow at enqueuing or at dequeuing or
> > even both. It seems for a quick implementation you need something like a
> > ring buffer, but I can't think of any method that would work for
> > symbolic elements, too (as opposed to numbers). 
> > 
> > Attached patch comparison.pd compares the methods. Isn't there really a
> > O(n) implementation as opposed to O(n^2)?
> > 
> > Roman  
> > _______________________________________________
> > Pd-list at lists.iem.at mailing list
> > UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
> > 

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 473 bytes
Desc: This is a digitally signed message part
URL: <http://lists.puredata.info/pipermail/pd-list/attachments/20151122/ae0f766b/attachment.sig>