What sort of QA process do you use for web development

Hmmm, haven’t heard much(anything?) from Raph Koster since the KS. Would be neat to check in with him, ask some questions and find out what he has been up to in relation with CF. I guess it would be more interesting if he has than if he hasn’t. It was always a cool thing during KS that he was involved with the project.

What sort of QA process do you use for web development? Do you have a mirrored site? I’ve seen a lot of new features be implemented on the website, and while I understand the challenges of building on top of IP Board, I can’t recall one new feature come online without major functional bugs.

I don’t recall specifically who, but I seem to remember Todd saying that one of the ACE developers brought along the work they had done on an MMO server framework. I’d love to hear a little from the server guys. What considerations go into building the back end of Crowfall?


Remembering the launch of GW2 there was a thing called culling. It meant that as soon as there where too many players on a screen, the game would only display x amount of them, because it was too heavy for a lot of computers to handle. Also they implemented an attack hit cap to prevent servers from crashing because of too much calculation. How will Crowfall servers handle HUGE Player Masses (with huge, I mean 80+% of the players on a map on one point fighting – even if this will only happen rarely / if the community as a whole does this on purpose)?

For now the ‘exotics’ are Guineceans, Centaurs, Stalkers and Minotaurs. Is it more that you go away from your typical biped (for example a Medusa or a Naga, both without legs) the less fitting it is for CF? I’m not asking from budget/animation pov. but more from your personal taste and maybe narrative reasons.

If that is the case, do you still see a place for such exotics, modified in a biped form – like we see Stalker’s stag head or judging by the pictures an Aracoix with bird elements on a human body? I know it’s not easy to put a beholder head on a human and make it work, but you know magic and i don’t. I already have much love for you for the archetypes that are in the game so far.

The only thing I have been a little disappointed in regarding the archetype system is that promotion can’t come from more than one base the way it did in SB. Was that design decision based on some of the logistic concerns about animations and art that have been discussed previously, or was there something about the way that design works that the team decided was wrong for this particular game?

It was Thomas ‘Dreadflame’ Sitch, and he was one of the core server engineers on Shadowbane (I can remember long nights where he and I were ripping out and redesigning major chunks of the city siege system)… I’ll mention this thread to him, see if I can get him to jump on. I might have to bribe him with a breakfast taco to take a break from his code, but I’m not above doing that.

Even if it was legal to use diku code (which of course it isn’t) the code is a complete mess. It uses binary flat files on the backend, rather than a persistent database, so every time you change the character record you risk corrupting your player database. It doesn’t scale. and, of course, it doesn’t have a concept of 3D space. You’d also have to pull a bunch of systems (like rent, and item spawn limits) out completely.

SB had a small budget, just like Crowfall — but it had a huge advantage, that being a much (much) lower quality bar in terms of animations and character art. The game looked crappy, but we couldn’t afford to do anything else (and didn’t know better) so we just decided to run with it.

We made some decisions that were effective, but acceptable (just barely) given the expectation that players had for graphics of the time. For example, “hmm, we can’t afford actual dwarves… I guess let’s just squish the human model down to make it shorter, and slap a beard on him. that’ll work”). That would never fly today, with players comparing us to games like WOW and SWTOR.

No Comments, Be The First!

Your email address will not be published.