Re: PHP telnet server
| From: | Robert Cummings | Date: | Wed, 31 Dec 2008 17:10:33 +0000 |
| Subject: | Re: PHP telnet server | ||
| References: | 1 2 | Groups: | php.general |
| Request: | Send a blank email to php-general+get-285298@lists.php.net to get a copy of this message | ||
On Wed, 2008-12-31 at 17:35 +0100, Per Jessen wrote:
> ceo@l-i-e.com wrote:
>
> >
> >>> I often thought PHP would be a nice language for a MUD, if one could
> >>> get the performance out of it...
> >>
> >> Design your code such that you can just throw more hardware at it
> >> whenever you need more performance.
> >
> > That's easily said, but a MUD means all the users have to share a
> > significant portion of your data model.
>
> That's fine - there are many well-known schemes for distributing and
> updating such data.
>
> > Though I suspect the bandwidth issue would be the main bottleneck most
> > of the time...
> > I don't want to get into this in much detail, mainly because I've
> > spent all of 5 minutes seriously thinking about it, and may just need
> > more bake-time...
> > But it's not ALWAYS that easy to architect something to be
> > "shared-nothing" even with PHP.
>
> It's not easy in any language, but if your key concern is the
> performance of PHP (as a language), hardware is what you need. You can
> design your software to run on a single box with lots of CPU cores, or
> you can go for a distributed (and more easily scalable) approach. If
> you don't need/want straight scalability, go for the 32 cores all
> ticking at 3GHz. Once that is saturated, buy another one.
If you go multi core then you need to go with a threaded approach...
which makes the development a bit complex for newbies to MUD
development. I don't think I'd go distributed since people whine about
lag that takes a 1/4 second... distributed would inherently require more
time while messages are passed to and fro.
Cheers,
Rob.
--
http://www.interjinn.com
Application and Templating Framework for PHP