The Hub of All Blame: A Postmortem

[Trippy]

One question occurred to me reading that - has McQuaid never heard of a college course entitled "Software Development"?  They actually teach people how to keep these metrics and complicated things.  It sounds like whatever process they were using at Sigil was not the SOP used by modern software businesses. 

It sounds like what Sigil was using could be sort of descrbed as agile development by someone kindly, or "bunch of people hacking away in groups" by everyone else. You had a Vision substituting for design documents, you had Management Fiat substituting for actual deliverables and milestones, and once MS figured out what was going on they dropped Sigil like a hot potato because you get consistent levels of pure shit when you use that as a design process in the real world.

Even kindly described that's not what agile programming is.

[sam, an eggplant]

Actually, it's not. 

You should have kept reading. Sigil's mistake wasn't initially relying on MS, it was failing to make arrangements during the last 9 months because they "simply didn't have room to grow".

Looking at the interview without context, some of Schild's questions sound unprofessional or prurient. Butler's affair is really none of our business, that's just gossip. But he was the executive producer, married to the lead GM, screwing the PR director, and my guess is that a lot of the emails flooding in from anonymous ex-sigilites mentioned it as a real problem. Same deal with the Christian references. Would be nice to get a comment on that by the way, Schild, although I can see not wanting to further push unsubstantiated gossip.

[Merusk]

Xanthippe - your opinion is very much the minority, proven by rapidly falling subscription numbers and ghostly servers on the VG client side. Look on virtually every MMO fan or information site and youll see thousands of threads talking about the piss poor game design and ideas that went into making this POS.

Your proof could go either way.  Falling subscription numbers and ghostly servers could just as easily be a symptom of the functional completeness of the game as they are of unsuccessful design.  Pretty sure the two together are a death knell in any event.

Or to frame your argument another way,  Reign's;  "Shadowbane's failure proved PvP games won't ever sell!*"

*(Just ignore Warcraft BGs, EvE, and all the nostolgia about 'old school' UO)

There's tons and tons of threads examining this subject in the MMO discussion & Development forums, please peruse & stay a while.

Anywho, Brad's delusional world of "it wasn't me!" shouldn't be surprising, and really it's not after seeing it for the 2nd time.   Anything else I had to say has been said better, or more profanity-laden already.

[Engels]

Some of those questions were out of line and I'm surprised Brad answered them.  Speaks to his character to be so forthright - regardless of what you think of Sigil, Vanguard or whatnot.

I'm sure he'll bounce back and learn from the Sigil experience.  He has had more success and given more to this industry than any of the snarky douchebags taking potshots on this forum, that is for-fucking-sure.

Good heavens.

[damijin]

Some of those questions were out of line and I'm surprised Brad answered them.  Speaks to his character to be so forthright - regardless of what you think of Sigil, Vanguard or whatnot.

I'm sure he'll bounce back and learn from the Sigil experience.  He has had more success and given more to this industry than any of the snarky douchebags taking potshots on this forum, that is for-fucking-sure.

 :-(

[wraith808]

You know, seeing the big boss break down in tears might have been classier than some dude making sarcastic jokes about buying a house (shades of needing an upgrade on a ferrari, eh?).

At least if he had joked about buying a Ferrari, you could look forward to the day he cracks up and gets arrested.

[Sjofn]

Nah, man.  We Xboxers aren't to blame, it those damn Zoo Tycooners!  Hang 'em high, I say!

It's my fault. I bought the first one and the second one! And every single expansion. Bwahahaha!

[Merusk]

Some of those questions were out of line and I'm surprised Brad answered them.  Speaks to his character to be so forthright - regardless of what you think of Sigil, Vanguard or whatnot.

I'm sure he'll bounce back and learn from the Sigil experience.  He has had more success and given more to this industry than any of the snarky douchebags taking potshots on this forum, that is for-fucking-sure.

You drank too much of the Kool-Aid.  This isn't the first time for this type of behavior on his part, and what he gave to the industry vs what he took from games now 20 years-old is debatable.

[Morat20]

Tears or not, Boss Man should be there when you cut loose your people.

Having the Ferrari and all the cool toys and the PC Gamer covers and shit is nice, but there's an actual job part to the "Boss Man" title. I think Brad might have forgotten.

You know why I've avoided management oppurtunities? I'd hate them. It'd spoil my pleasure in my work. Technical lead? I'm all for it. Project lead -- as long as it was in a leadership, and not a management roll -- bring it on. Put me at the level where I'm doing performance reviews, sketching out project documentation, attending boring ass meetings about budget, and hiring and firing people?

Fuck, you better have a sweet offer on the table, because while I can do it -- I sure as hell want some serious compensation for doing all the shit work.

On the other hand, if I got Brad's salary and his car (which I'd sell for something useful) -- I'd feel like shit if I wasn't doing the real work. Which included being there to deliver the news in person if my performance was shitty enough to lead to layoffs.

Wraith: Regarding QA, I can see Sigil's point --- but personally I would have wanted one or two full-time internal QA folks for peace of mind. Someone to basically backup the coders and designers and check for blindspots. No sense waiting to the end, when you can have QA slogging through early prototypes looking for sticky points in the design. Plus, they'd also fufill a useful roll in the "Is this fun? Is this intuitive? Does this really work for a player?" perspective.

Sauced: Dude, it pretty much IS pure shit. Yeah, it's got all the teamwork and communications buzzwards, but in practice it tends to suck donkey balls. Look, I'm sure there are managers -- and teams -- that can make Extreme Programming work. There are people who can turn street trash into fucking works of art. However, it's really fucking unlikely that any sizeable team is going to consist of geniuses, and it's really goddamn unlikely that any sizeable team will be able to actually manage the sort of teamwork and communication that Extreme Programming calls for -- you've got too many people, doing too many things, and it quickly gets too much for anyone to keep track of and falls apart.

If you're a 5 person shop -- Extreme it away. If you're employing 20, I suggest you look elsewhere. There's no way in fucking hell one person is going to be able to focus on his job, keep an eye on what 19 other people are doing so he doesn't break them, attend meetings with every other fucking group that might break him (or he might break), and get anything done. Fucking managers are there for a reason. Team leads, project leads, technical leads -- whatever you want to call them -- are there for a reason. If I am engine design, I do not give a flying fuck what the art people are doing 80% of the time. I do not need to keep track of it. I have more than enough shit on my plate. I DO need someone to be keeping tabs on them for me, so that if they have some suggestions about the toolsets they're using I can show up -- or if they're talking about pushing more polys and need an expert opinion on whether the engine can take it.

But everything else is a waste of my time. That's what managers are for. That's what actual processes are for. Extreme Programming more or less requires me to attend those meetings, in case the art people decide on something stupid (like insanely high poly models) OR I delegate a team member to go keep an eye on the idiots -- in which case I'm already instituting a management structure, complete with team interfaces, so I should fucking formalize it so the art people know who to pester when they hold their damn meetings.

Can you tell I've sat through more than one meeting I didn't need to be in, when I could have been doing something productive like downloading porn?

In any case -- communication and teamwork are a vital part of ANY process. Iterative/Incremental development is a much better choice -- and frankly the point is you tailor the process to fit your design needs, which means you can tailor it down to Extreme Programming if you have a tiny shop that doesn't require all those processes just to keep track of workflow and who is doing what.

[wraith808]

Actually, it's not. 

You should have kept reading. Sigil's mistake wasn't initially relying on MS, it was failing to make arrangements during the last 9 months because they "simply didn't have room to grow".

You should have read my whole post.  Then you would have seen that I said the same thing.  What part of

"What he can be blamed for in this case is not bringing it up as a risk to SOE and getting them to foot for QA if Sigil couldn't afford it."

did you not understand?

[Trippy]

Worrying that what you doing might break somebody else's work is a sign of bad software design, regardless of what software development methodology you are using. Also there's nothing about agile programming that says you have to worry about what everybody else is doing.

[Sauced]

Worrying that what you doing might break somebody else's work is a sign of bad software design, regardless of what software development methodology you are using. Also there's nothing about agile programming that says you have to worry about what everybody else is doing.

Discipline was mentioned before, and it is an important aspect of the process.  It's a different enough approach that it can take a while to change habits.  I want to say "bad habits", but that's just a personal bias.  I've been working in agile environments since 2000, at fairly large companies and teams.  Good testing habits (I'm a test-first guy, myself) and use of tools like Continuum are important.

[almagill]

It just seems shallow.

It is shallow. It's basically malicious gossip. And it's entertaining, and for the most part newsworthy. Nothing wrong with that.

Works for me.

I had been going to go do some RL stuff but this car crash is way more interesting...   Will there be a made for TV dranatisation one day?

[Reign]

Some of those questions were out of line and I'm surprised Brad answered them.  Speaks to his character to be so forthright - regardless of what you think of Sigil, Vanguard or whatnot.

I'm sure he'll bounce back and learn from the Sigil experience.  He has had more success and given more to this industry than any of the snarky douchebags taking potshots on this forum, that is for-fucking-sure.

Brad, quit posting with your alternate handle, get out here, and take your lickin' like a man..Dont bitch out of the process twice....

Thats about the worst wishful thinking I've seen in my lifetime.....

[Morat20]

It sounds like what Sigil was using could be sort of descrbed as agile development by someone kindly, or "bunch of people hacking away in groups" by everyone else. You had a Vision substituting for design documents, you had Management Fiat substituting for actual deliverables and milestones, and once MS figured out what was going on they dropped Sigil like a hot potato because you get consistent levels of pure shit when you use that as a design process in the real world.

Even kindly described that's not what agile programming is.

Yeah, that's the gist I got too. I thought I would be a bit more generous, because it really looks like "A bunch of people hacking away". I'm not sure what process they used -- if any. I kind of got the impression everyone was sort of doing their own thing, and probably the only real process was "Management decides and you do" and whatever the coders had to set up just to actually get work done.

I'd love to see their personnel records over time. Who they hired and when, and what their jobs were. I've seen a lot of projects go south because people were brought in too early or too late. DBA's brought in to design a DB schema when the interface guys had already done so much work the DBA was more or less stuck tweaking whatever the guy with the 1 DB class had put together. GUI folks brought in to try to assess a GUI that was so far into development that changing it was more work than it was worth, that sort of thing.

I'm sure they had basic stuff in place -- version control, that sort of thing. The interviews suggested they had some sort of communication channels in place that looked more or less right -- but management ignored them. Brad's comments about MS requiring them to have some formal documentation of what they were doing and when they planned to get there -- over what sounded the short term -- implied that either Brad was ignorent of management processes, or there were no management processes. You really can't ad hoc something that big, with that many people. You don't need 50 reams of paper, but you need some structure and some discipline.

Trippy: My understanding -- albiet limited -- of Extreme programming is that you better damn well be keeping tabs on everyone else, because they're expecting you to speak up if what's on the menu for today has some impact on you. I wasn't speaking to agile -- which I'm kind of viewing as a stripped down and faster version of iterative (is that right?) -- but with that sort of process, I'd be more worried about "Breaking someone else's work" in the sense that if I'm evolving core components, other programmers might be better off waiting awhile to use the better functionality I'm bringing online (as opposed to having to do rework), or that I might be adding features someone else is anticipating but not in the manner they're designing for.

Interface issues aren't a sign of bad software design, but general of poor communication. I'm not really talking about stuff that's avoidable by proper black-boxing and the like, but of problems where interfacing groups have slightly different ideas of what's going on.

[wraith808]

Wraith: Regarding QA, I can see Sigil's point --- but personally I would have wanted one or two full-time internal QA folks for peace of mind. Someone to basically backup the coders and designers and check for blindspots. No sense waiting to the end, when you can have QA slogging through early prototypes looking for sticky points in the design. Plus, they'd also fufill a useful roll in the "Is this fun? Is this intuitive? Does this really work for a player?" perspective.

Oh, no... I agree with you.  I was just saying pretty much the same thing... I can see Sigil's point.  If we were developing software for an external audience rather than a client, we'd have internal QA, no doubt.  And I agree with your point about management... I own my company, but I'm not a manager... just an architect and a tech lead.  I hired a manager, because I'm no good at that.  It's a matter of knowing your own limitations, and resisting the urge to be promoted past your competency, which so many people tend to do.

[Ironwood]

Some of those questions were out of line and I'm surprised Brad answered them.  Speaks to his character to be so forthright - regardless of what you think of Sigil, Vanguard or whatnot.

I'm sure he'll bounce back and learn from the Sigil experience.  He has had more success and given more to this industry than any of the snarky douchebags taking potshots on this forum, that is for-fucking-sure.

You're Special.

Will you be my friend ?

[Morat20]

Also there's nothing about agile programming that says you have to worry about what everybody else is doing.

And for the record -- the stuff I've been personally working uses an agile programming process. The stuff the rest of my group works on uses an iterative process. We ended up at agile by stripping away all the iterative crap my stuff didn't need, and speeding it up. :)

[Mage]

Great work on the interview...

So I don't know that much about game development, but I do know a little something about management. I manage 120+ people (IT Industry).

No wonder the company failed. After reading this article I would not let this guy manage a popcorn stand for me.

I've decided to take a break from Vanguard. I cancelled my account. It will expire next week. I may or may not come back to it.

Obviously it still needs work, and I am tired of playing the potential game.

I am a long time player: MUDs, UO, EQ1, AO, EQ2, WoW, EvE, Vanguard and others...

Kill with Grace.

[egoslicer]

http://www.fohguild.org/forums/mmorpg-general-discussion/28686-vanguard-f13s-interview-brad-mcquaid-16.html#post734523

Brad's reply to some of the interview/forum posts.

[Endie]

Not for McQuaid. He has now had 2 chances, and he's fucked them both directly in the clownass. He's not listened repeatedly when myself and folks less sweary than me have told him how misguided his ideas of game design are, and how his game would fail partly because of it. If I had been WRONG about any of the Vanguard predictions, then maybe I might have some kindness left for him, but I haven't been at all wrong.

His world fell apart long ago, he was just too much of an egotistical douche to see it. And he brought 100 people down with him because of said ego. There are 50 people out there without fucking jobs anymore because of his myopic ego AND he didn't even have the sack to go and face them when they got shitcanned. How much kindness should I have for someone like that?

Yeah, I just feel sorry for him.  of course, I feel sorry for all of them.  And yes, his decisions mean they're out of work now.  But "entrepreneurs" try stuff all the time.  People say it won't work to them, and the stats show that those people are usually right.  And lots of people that took a punt on them lose their jobs.

I do think that he should have had the guts to be there at the end.  I totally understand why he wasn't, sure.  But he had responsibilities.  Literally: he was responsible.  Yes, it would have been horrible, yes he might have cried.  Suck it up.

So I suppose I kinda agree with you after all.  Ooops.

[Merusk]

http://www.fohguild.org/forums/mmorpg-general-discussion/28686-vanguard-f13s-interview-brad-mcquaid-16.html#post734523

Brad's reply to some of the interview/forum posts.

Pure Spin.  He's Cut & Pasted it to Silky Venom as well.

  It's a press release from the press master.

[Morat20]

http://www.fohguild.org/forums/mmorpg-general-discussion/28686-vanguard-f13s-interview-brad-mcquaid-16.html#post734523

Brad's reply to some of the interview/forum posts.

Sweet! Minor details!

We tried very hard to work with the new team but their approach to game development simply wasn't compatible with Sigil's, did not adhere to how we had worked with them previously for several years, and IMHO not compatible with how one (or at least Sigil) in our experience working on a LOT of MMOGs, some failures, one a huge success, some cancelled, etc.

How? Okay, sounds like a new MS liason. What did they want? What was incompatable?

I can read that in a lot of ways. Two that spring to mind:

1) The new MS guys come in, start telling Sigil they want a far more concrete and less flexible process that works for shit on the sort of development an MMORPG goes through, holds Sigils feet to the fire, screws everything up.
2) The new MS guys come in, take a look around, realize Sigil is fucking jerking their chain and starts demanding a less flexible process because someone is taking advantage of that process to scrap and rework constantly, never making real progress, and dictating real results and not -- just making up an example here -- mocked up 'demos' strictly for MS consumption.

I would lean towards the second based on timing. It seemed like MS kept Sigil on a very short leash for at least a year or two, all while funneling in cash. While I can see a new liason or MS oversight team being a real hardass and an idiot, I can see Sigil having a very solid response with "Look at what we've done and how far we've come" -- unless they haven't done much or come very far.

MS was on the hook for real cash. If they were happy with what Sigil had done, a new liason wouldn't have the weight to demand changing something so drastically. Ergo, it appears they weren't happy. As events turned out, it appears obvious why.

[Trippy]

Trippy: My understanding -- albiet limited -- of Extreme programming is that you better damn well be keeping tabs on everyone else, because they're expecting you to speak up if what's on the menu for today has some impact on you. I wasn't speaking to agile -- which I'm kind of viewing as a stripped down and faster version of iterative (is that right?) -- but with that sort of process, I'd be more worried about "Breaking someone else's work" in the sense that if I'm evolving core components, other programmers might be better off waiting awhile to use the better functionality I'm bringing online (as opposed to having to do rework), or that I might be adding features someone else is anticipating but not in the manner they're designing for.

Communication is always an important thing in any software project where the number of programmers is > 1 unless they are working on completely orthogonal systems and even then they should be talking to each other. There's nothing about XP or Agile that says you have to keep track of what everybody else is doing. Regular meetings is a part of Agile methodologies (e.g. Scrum has very short daily "standup" meetings) but again that doesn't mean everybody has to keep track of what everybody else is doing. Fred Brooks proved long ago that that sort of thing is what actually slows down software development rather than speeding it up.

Interface issues aren't a sign of bad software design, but general of poor communication. I'm not really talking about stuff that's avoidable by proper black-boxing and the like, but of problems where interfacing groups have slightly different ideas of what's going on.

Modules should be loosely coupled, generally speaking. If a programmer A working on module A has to worry about the specific code programmer B is writing in module B then that's a sign those two modules are not perhaps as loosely coupled as they could be. Of course not all modules can be loosely coupled so this sort of thing doesn't always apply.

[Aradune Mithara]

http://www.fohguild.org/forums/mmorpg-general-discussion/28686-vanguard-f13s-interview-brad-mcquaid-16.html#post734523

Brad's reply to some of the interview/forum posts.

You got to it before I did.  Here it is:

------


I'd like to address something I've seen come up several times on different boards re: the interview:  The original MS team DID require milestones and a schedule.  By NO means did we ask for or expect free reigns.  In fact, we met every milestone and even exceeded the expectations and requirements for many of them.

I am all for schedules.  I am all for planning.  I am al for Milestones.  

It was the degree and detail level and how detailed they needed to be going way out into the future that changed and the assertion that MMOGs should be handled the same way as a single player game in terms of development and an apparent desire to not want to work collaboratively but rather dictate development (which was not part of the original spirit of the agreement, where we were hired on because of our experience) that became a problem.  We tried very hard to work with the new team but their approach to game development simply wasn't compatible with Sigil's, did not adhere to how we had worked with them previously for several years, and IMHO not compatible with how one (or at least Sigil) in our experience working on a LOT of MMOGs, some failures, one a huge success, some cancelled, etc.  

We were all for organization, scheduling, milestones, accountability, etc.  Up until the change I described we excelled in this process and were praised for our performance and progress.  When all of this changed, we tried to make things worked (as did the other side), but it simply turned out that the two teams were not compatible.  Again, I have nothing against MSFT or those people, etc.  They meant well and I'm sure were doing their best given how they were told to deal with an external developer.  Changes like this happen fairly often I would think.  Incompatibilities and disagreements on development (in this case MMOG development) happen.  

Thus it was time to move on.  I will also say that working with SOE later on worked out fairly well (not perfect, but perfection happens rarely).  My point is that we worked very well with the first group at MSFT and then we worked very well with SOE after we broke things off.  

That's it.  Again, while I am disappointed, I understand how things happen and that they are not all that uncommon.  A developer and a publisher must mesh and be able to work together as a solid team towards one solid goal (making a great game).  When that does not occur, or something changes such that it is no longer occurring, something needs to change.

Microsoft agreed with this and allowed us to switch to SOE in a very amicable way.  There are no hard feelings.  All the MSFT people are in the Vanguard credits in fact.  If anything my respect level for MSFT management for letting go of Vanguard to increase to the chance that it would become a great game increased greatly.  They are truly a class act.  They put Vanguard above politics and team issues and different development philosophies and did what was best for the game.  That is huge and I'm not sure if very many other publishers would have done the same.  So I remain eternally grateful for them allowing us to make the switch to SOE and getting the deal done so quickly so as to allow us to take advantage of E3 and to interrupt the development process of the game as minimally as possible.

I wish this part of the picture had been included in the interview and hope you all spread it around as it's an important piece that was missed.

thanks,

ps. I don't blame the interviewer for missing this -- we went over a lot of stuff.

[Simond]

Kieth Parkinson's contribution is probably one of the only reasons to even play VG.

It's a shame that Vanguard was his last project, in much the same way that it was a shame that Streetfighter: the Movie was Raul Julia's last film.

[Hoax]

Some of those questions were out of line and I'm surprised Brad answered them.  Speaks to his character to be so forthright - regardless of what you think of Sigil, Vanguard or whatnot.

I'm sure he'll bounce back and learn from the Sigil experience.  He has had more success and given more to this industry than any of the snarky douchebags taking potshots on this forum, that is for-fucking-sure.

I think more then enough people have given Brad credit in this thread for actually talking to f13 and being honest in the interview.  Obviously nobody from M$ would talk to a site like this other then to tow some party line bullshit.  Ditto most likely for SOE.

Also most people here aren't taking potshots, we've been calling the completely whack parts of "The Vision" what they are, which is cockblocking at its worst since '99.  Nobody needs to be snarky about it anymore.

Thanks for trying to learn us though, perhaps next time you'll do better?

[Margalis]

1. Schild deserves for sort of award for putting major websites to shame and showing people what actual game journalism looks like.

2. It is clear that Brad had emotionally and mentally checked out a long time ago.

3. The fact that Brad thinks posting on a message board is something he should be doing is quite frightening. His posts are absurd and at some point you have to think they are so long they are detracting from the actual work he should be doing.

4. CEO not showing up for major events like firings is just plain gutless.

[Aradune Mithara]

http://www.fohguild.org/forums/mmorpg-general-discussion/28686-vanguard-f13s-interview-brad-mcquaid-16.html#post734523

Brad's reply to some of the interview/forum posts.

Sweet! Minor details!

We tried very hard to work with the new team but their approach to game development simply wasn't compatible with Sigil's, did not adhere to how we had worked with them previously for several years, and IMHO not compatible with how one (or at least Sigil) in our experience working on a LOT of MMOGs, some failures, one a huge success, some cancelled, etc.

How? Okay, sounds like a new MS liason. What did they want? What was incompatable?

I can read that in a lot of ways. Two that spring to mind:

1) The new MS guys come in, start telling Sigil they want a far more concrete and less flexible process that works for shit on the sort of development an MMORPG goes through, holds Sigils feet to the fire, screws everything up.
2) The new MS guys come in, take a look around, realize Sigil is fucking jerking their chain and starts demanding a less flexible process because someone is taking advantage of that process to scrap and rework constantly, never making real progress, and dictating real results and not -- just making up an example here -- mocked up 'demos' strictly for MS consumption.

I would lean towards the second based on timing. It seemed like MS kept Sigil on a very short leash for at least a year or two, all while funneling in cash. While I can see a new liason or MS oversight team being a real hardass and an idiot, I can see Sigil having a very solid response with "Look at what we've done and how far we've come" -- unless they haven't done much or come very far.

MS was on the hook for real cash. If they were happy with what Sigil had done, a new liason wouldn't have the weight to demand changing something so drastically. Ergo, it appears they weren't happy. As events turned out, it appears obvious why.

I'm sure our perspective differes from others who were involved.  Like I siaid, in the end everything was resolved amicably and I have great respect for the Manager who did what he thought was best for the game.

As for how things turned out, I guess that's your opinion.  You may not like Vanguard, but it's a great game and many people agree.  Under SOE now I think it will continue to improve and evolve and get better and better, but I will always stand up and say, that even though we could have used a few more months, and even though it would have been preferable to not ship so close to TBC, etc. (see my post I made a few weeks back about where things screwed up and it's pretty clear that I don't blame any one entity over another and admit to many mistakes we made ourselves).  All that said, despite the game's issues which have been discussed in detail, we did launch an MMOG, a very compex and expensive one, a very ambitious one, and a very fun one for many people, and for that  I will always be proud, especially of the team, and then also thankful to both MSFT and SOE for giving us the opportunity to finish the game and realize a great many of our dreams.

[wraith808]

Kieth Parkinson's contribution is probably one of the only reasons to even play VG.

It's a shame that Vanguard was his last project, in much the same way that it was a shame that Streetfighter: the Movie was Raul Julia's last film.

Ouch.  Really.  Ouch.

[Morat20]

Communication is always an important thing in any software project where the number of programmers is > 1 unless they are working on completely orthogonal systems and even then they should be talking to each other. There's nothing about XP or Agile that says you have to keep track of what everybody else is doing. Regular meetings is a part of Agile methodologies (e.g. Scrum has very short daily "standup" meetings) but again that doesn't mean everybody has to keep track of what everybody else is doing. Fred Brooks proved long ago that that sort of thing is what actually slows down software development rather than speeding it up.

I see what you mean. Even so, I think there's a natural team size that limits Agile or XP as an overall design methodology. But not for components -- Agile is an excellent way to develop subcomponents (or primary modules) of a large scale system. Care must be taken at interfaces, but there's absolutely nothing wrong (and a hell of a lot right) about using a very agile process down at the bottom where the real work is done.

And that applies to artists and designers as well as coders.

But you've got to have the large-scale processes -- specifically the management set -- to handle that for the project as a whole. One of the hardest problems I had with the whole "project management" concept was in trying to allow maximum flexibily for the individual aspects of a project, while keeping the bloody thing unified as a whole. Doing it with five people it's just a matter of keeping tabs on people -- doing it with 100, and I really need those papers, those Gannt charts, those milestones, those work breakdown structures.

Someone's got to be keeping a big enough eye on the picture to go down to the engine guys and tell them what they're doing is all awesome and shit, but the designers really need that upgraded toolset about two weeks ago and it's delaying the artists ability to do all the awesome stuff, and it's going to slip the whole bloody project so can they tweak the lighting code or whatever next week and please for the love of God get those tools done so the artist shut up and the money people stop screaming at me?

Modules should be loosely coupled, generally speaking. If a programmer A working on module A has to worry about the specific code programmer B is writing in module B then that's a sign those two modules are not perhaps as loosely coupled as they could be. Of course not all modules can be loosely coupled so this sort of thing doesn't always apply.

I'm more used to very large modules that often have highly coupled interfaces -- we try to do it through open APIs and the like, but more often than we want we're forced to go sit down with what is -- in effect -- an entirely different project's team lead and patiently explain that yes, indeed, we love their API but our respective bosses and payers of our salary really WANT that connectivity whether they want to open things up that much or not and we're about to start Pointing the Finger of Blame. :)

A good process, in the end, allows for maximum flexibility for individual components, a large scale of flexibilty for the overall system, but protects against feature creep and -- and this is a must -- makes sure the end result actually does what the thing was designed for. Whether it's control the Space Shuttle or makes a fun game.

[Morat20]

I'm sure our perspective differes from others who were involved.  Like I siaid, in the end everything was resolved amicably and I have great respect for the Manager who did what he thought was best for the game.

As for how things turned out, I guess that's your opinion.  You may not like Vanguard, but it's a great game and many people agree.  Under SOE now I think it will continue to improve and evolve and get better and better, but I will always stand up and say, that even though we could have used a few more months, and even though it would have been preferable to not ship so close to TBC, etc. (see my post I made a few weeks back about where things screwed up and it's pretty clear that I don't blame any one entity over another and admit to many mistakes we made ourselves).  All that said, despite the game's issues which have been discussed in detail, we did launch an MMOG, a very compex and expensive one, a very ambitious one, and a very fun one for many people, and for that  I will always be proud, especially of the team, and then also thankful to both MSFT and SOE for giving us the opportunity to finish the game and realize a great many of our dreams.

I never actually expressed an opinion on Vanguard, just the way it's development appeared to the public. If it makes you feel better, I'm pretty much consistent in complaining that the games industry is the bastard child of software industry when it comes to professionalism, and that the software industry spent a few decades trying to pretend all that 'project management crap' the engineers did didn't matter for software.

If you did have a poor process -- you've got lots of company. LOTS of company. Games -- especially MMORPGs -- are just too big to get away with the sort of things that they could even a decade ago.

And hey -- your baby got out the door. I won't take that from you, seeing as how most people's don't.

But if you're still kicking around the industry, you owe it to yourself (at the very least) to do a very thorough and brutally honest personal post-mortem of the project. And it needs to focus -- first and foremost -- on the management processes you and Sigil implemented. Everything else -- from overly ambitious design, bad foresight on hardware advancements, architectural mistakes -- all that could and should have been caught by a mature process.

If you plan to oversee another game -- whether it's you and a few hobbists just doing it for kicks, or another MMORPG -- you need to think about where management screwed the pooch on this baby, and either figure out how to fix it on your own or identify the sort of people that can fix it for you. You owe that to your customers, to your players.

[AaronC]

You drank too much of the Kool-Aid.

Oh teh irony.

I think more then enough people have given Brad credit in this thread for actually talking to f13 and being honest in the interview.

Yeah and for the most part people have been surprisingly cool, considering... (http://www.penny-arcade.com/images/2004/20041103h.jpg)  (http://www.penny-arcade.com/images/2004/20040319h.jpg)

[Sauced]

At the "first annual XP Conference" in 2001, I managed to get most everyone but Kent Beck (so Ward Cunningham, who is great, Martin Fowler, who is a savant, and Ron Jeffries, who is very "GET OFF MY LAWN") that XP only worked in large-scale corporate environments if teams were allowed to modify or abandon any of the 12 rules that did not or could not apply to the situation.  Beck's response was a hard line "you have to make everyone do everything, or it isn't XP".  The word "Extreme" was also a barrier for us, as it flat out scared the crap out of management.

Anyways - to me, an "Agile" methodology comes from this way of thinking.  XP is a great theoretical idea, but is not practical in it's full incarnation.  But neither of them would preclude the ability to effectively communicate schedules, deadlines, shifts in scope, etc. 

[Arthur_Parker]

Yeah and for the most part people have been surprisingly cool, considering... (http://www.penny-arcade.com/images/2004/20041103h.jpg)  (http://www.penny-arcade.com/images/2004/20040319h.jpg)

Did you bookmark those in 2004?