[MUD-Dev] [NEWS] Warhammer Online Cancelled -- Why?

Adam adam at grexengine.com
Wed Jun 23 12:25:21 CEST 2004


David Kennerly wrote:

> There's a lot of articles on the net.  None of them explain the
> internal reasoning.  Their one-sided mention of projection of
> costs leaves it up to the reader to assume a not so hot projection
> of revenues.  Here's one that was above average:

Just random conjecture, but I thought they were going to kill
themselves with the costs...

I'd always been suspicious they might not make it - that they might
do an Elite 3 and belatedly realise "ohmygod this is hard and
expensive to try and hack together from nothing" - from our talks
with them going back a couple of years. They had a brief look at
using the grexengine, and then apparently just chose to do it
in-house, despite the obvious problems. Having spoken to others
since, it seems that quite a few people made it clear to them the
enormity of the problems, but they still chose to roll their own.

I kind of got worried when Jim posted here asking for help solving
really basic scalability problems that suggested they'd wasted many
months trying to build a large scale server with little or no idea
what they were doing:

  Currently we've got a send and receive thread per-connection and
  need to support ~2500 connections per box, so we're going to end
  up with ~5000 threads.

  http://www.kanga.nu/archives/MUD-Dev-L/2003Q3/msg00466.php

Obviously, Jim was trying to fix such things (I'm not trying to put
him down), but I got the impression he was in a minority (possibly
of as little as one) of people who had a clue about the major
technical issues. Arriving in the middle of a project at a games
company as a single individual to try and fix a from-scratch mmog
engine sounds to me like a pretty impossible mountain to climb :(.

Another notch on the "I told you so" for the middleware companies?
(tongue in cheek!). Although we did tell them so. And, as I said, I
believe we weren't the only ones to do so.

But, as David said, it could easily have been a revenues issue
rather than costs.

Adam M
_______________________________________________
MUD-Dev mailing list
MUD-Dev at kanga.nu
https://www.kanga.nu/lists/listinfo/mud-dev



More information about the mud-dev-archive mailing list