Conference Call

GWJ Conference Call Episode 335

Tomb Raider, SimCity, Sword of The Stars: The Pit, Jeff Cannata's Kickstarter, Games as Service, Your Emails and more!

This week Shawn, Elysium, Julian, Rob Zacny and Jeff Cannata talk SimCity, games as service, Jeff's new Kickstarter and more!

To contact us, email [email protected]! Send us your thoughts on the show, pressing issues you want to talk about or whatever else is on your mind. You can even send a 30 second audio question or comment (MP3 format please) if you're so inclined.

Chairman_Mao's Timestamps
00.01.19 Kickstarter talk
00.08.04 Tomb Raider
00.21.31 Sword of the Stars: The Pit
00.25.02 Twilight Struggle
00.31.27 Sim City
00.46.17 This week's sponsor Audible.com's book recommendation: A Memory of Light: Wheel of Time Book 14!
00.47.42 This week's topic: Gaming as a Service!
01.12.55 PAX GWJ panel updates
01.15.07 Your emails!

Jeff Cannata's Kickstarter
Tomb Raider
SimCity
Sword of The Stars: The Pit
Twilight Struggle
Raph Koster Blog

  • Subscribe with iTunes
  • Subscribe with RSS
  • Subscribe with Yahoo!
Download the official apps
  • Download the GWJ Conference Call app for Android
  • Download the GWJ Conference Call app for Android

Show credits

Music credits: 

Analytics - SimCity - http://www.simcity.com - 47:13

Population = 1 - SimCity - http://www.simcity.com - 1:12:27

Intro/Outtro Music - Ian Dorsch, Willowtree Audioworks

Comments

Can we have a moratorium on Certis using the word "verisimilitude" now? It's getting a little out of control.

AndrewA wrote:

Can we have a moratorium on Certis using the word "verisimilitude" now? It's getting a little out of control. :P

What? I CLEARLY have never said it out loud before. Mangled that one.

I want more fun adjectives life Jeff's "manipulatable."

Trepidacious!

Rambunctious

I wonder, knowing zero about the complexities of it, if there is a business in offering overflow servers to games companies for them to rent for a launch. I guess there might not be enough big launches to make worthwhile. Although that might change in the future.

You know you're in for an interesting segment when Jeff leads off with a joke about connectivity issues.

Higgledy wrote:

I wonder, knowing zero about the complexities of it, if there is a business in offering overflow servers to games companies for them to rent for a launch. I guess there might not be enough big launches to make worthwhile. Although that might change in the future.

I have to imagine that's why EA is using Amazon's cloud services. Most MMO games go through a rapid expansion of servers in the first week or two, followed by a drawn out consolidation of instances as interest levels out. I wouldn't be surprised if more companies switch to hosted VMs through cloud service providers like Amazon instead of rolling up physical servers which have a higher maintenance and deployment cost.

shoptroll wrote:
Higgledy wrote:

I wonder, knowing zero about the complexities of it, if there is a business in offering overflow servers to games companies for them to rent for a launch. I guess there might not be enough big launches to make worthwhile. Although that might change in the future.

I have to imagine that's why EA is using Amazon's cloud services. Most MMO games go through a rapid expansion of servers in the first week or two, followed by a drawn out consolidation of instances as interest levels out. I wouldn't be surprised if more companies switch to hosted VMs through cloud service providers like Amazon instead of rolling up physical servers which have a higher maintenance and deployment cost.

That's one of the biggest attractions to cloud computing, and amazon's elastic (it stretches) computing. The customer (EA in this case) doesn't have to buy/run the servers themselves, they just rent capacity and can rent a bit more when they need. Of course, their code running on those servers has to be scalable.

shoptroll wrote:

You know you're in for an interesting segment when Jeff leads off with a joke about connectivity issues.

Higgledy wrote:

I wonder, knowing zero about the complexities of it, if there is a business in offering overflow servers to games companies for them to rent for a launch. I guess there might not be enough big launches to make worthwhile. Although that might change in the future.

I have to imagine that's why EA is using Amazon's cloud services. Most MMO games go through a rapid expansion of servers in the first week or two, followed by a drawn out consolidation of instances as interest levels out. I wouldn't be surprised if more companies switch to hosted VMs through cloud service providers like Amazon instead of rolling up physical servers which have a higher maintenance and deployment cost.

There's no excuse, since EA did use Amazon's cloud servers, which are pretty easy to scale up transparently (that's why it's called the Elastic Compute Cloud). Major internet servers (Instagram, Netflix, Reddit) use it without exposing anything to the users. My view is converging to believe that EA badly misunderstood what they were building, and built servers that followed an MMO or multiplayer FPS model, instead of distributed servers that are properly scalable. There is no good technical reason why there are multiple shards exposed to the user. Does Instagram make you pick which server to upload your photos to? Does Farmville make you save your farm to a specific server? Does Reddit have shards? Does Netflix make you wait in a server queue?

My read on it is that they thought like MMO engineers when they should have been thinking like web service engineers, because they were selling a service.

I'd like some actual facts to back my opinion up, mind. But it's a pretty plausible read on the situation.

I'm so glad that Jeff's Kickstarter is doing well. He is definitely one of my favorite people to hear/watch discuss games. All the praise for Tomb Raider makes me excited to pick it up down the road. Unfortunately, I can't afford it right now with Bioshock right around the corner.

Rabbit, I'm surprised you haven't tried to sneak in some discussion about the LCS. Aren't you watching it? I'm really impressed with what Riot has done there.

Also, hey CC crew, still waiting on that promised big Dark Souls segment. Can we expect it sometime soon?

Dyni wrote:

Also, hey CC crew, still waiting on that promised big Dark Souls segment. Can we expect it sometime soon?

Yes! Just been a scheduling nightmare with everyone's day jobs being busy and lots of a travel. We haven't forgotten though!

I'm sorry, I don't understand the distinction being made between Shawn Andrich and EvilShawnAndrich?

Certis wrote:
Dyni wrote:

Also, hey CC crew, still waiting on that promised big Dark Souls segment. Can we expect it sometime soon?

Yes! Just been a scheduling nightmare with everyone's day jobs being busy and lots of a travel. We haven't forgotten though!

I totally want to write in my thoughts on playing Dark Souls.

I think this just motivated me to start a text file to put my thoughts together, no idea if it'll be worth sending / reading on the show yet, though.

But, if I do, the callers get to mock my name again, like they have every time in the past.

Elysium wrote:

I'm sorry, I don't understand the distinction being made between Shawn Andrich and EvilShawnAndrich?

One has a better goatee

AndrewA wrote:

Can we have a moratorium on Certis using the word "verisimilitude" now? It's getting a little out of control. :P

I'll add to that rabbit putting "osphere" on the end of words.

AndrewA wrote:

Can we have a moratorium on Certis using the word "verisimilitude" now? It's getting a little out of control. :P

+1

I too support discouraging others from using words having more than 3 syllables in them.

/signed

Elysium wrote:

I'm sorry, I don't understand the distinction being made between Shawn Andrich and EvilShawnAndrich?

I understand the conundrum my good fellow. It might be fairly difficult to distinguish between us since the "good" Shawn Andrich is Canadian.

But I assure you, I am evilerer, just not Canadian, as that would just be absolutely immoral, and even I have standards to keep up.

/Good day sir!
ESA

It'll be interesting to see how this Tomb Raider praise holds up when GWJCC picks its GOTYs for 2013. Did everyone play it on PC? I know Shipwreck of CAG had a really tough time getting it to not crash on his PC. He ended up Redboxing it for 360.

S0LIDARITY wrote:

It'll be interesting to see how this Tomb Raider praise holds up when GWJCC picks its GOTYs for 2013. Did everyone play it on PC? I know Shipwreck of CAG had a really tough time getting it to not crash on his PC. He ended up Redboxing it for 360.

Ran rock solid for me. I think some people were having issues with Tesselation which frankly, I can't even tell if it's off or on. I'm running an Nvidia 660ti with TressFX turned on.

Worth updating since I have finished the game now. My impressions are still largely accurate. There are definitely a few Uncharted WAVES OF ENEMIES moments, but they're still paced out with other stuff. Some excellent fights and set-pieces near the end.

Talk about service! Thanks Shawn.

I'll never forget the Half-Life 2 launch. I was so upset that I basically ragequit PC gaming for 3 1/2 years, until I finally got a high speed internet connection. Talk about a mess.

Excellent show. Really enjoyed the discussion about games as services.

Gremlin wrote:
shoptroll wrote:

You know you're in for an interesting segment when Jeff leads off with a joke about connectivity issues.

Higgledy wrote:

I wonder, knowing zero about the complexities of it, if there is a business in offering overflow servers to games companies for them to rent for a launch. I guess there might not be enough big launches to make worthwhile. Although that might change in the future.

I have to imagine that's why EA is using Amazon's cloud services. Most MMO games go through a rapid expansion of servers in the first week or two, followed by a drawn out consolidation of instances as interest levels out. I wouldn't be surprised if more companies switch to hosted VMs through cloud service providers like Amazon instead of rolling up physical servers which have a higher maintenance and deployment cost.

There's no excuse, since EA did use Amazon's cloud servers, which are pretty easy to scale up transparently (that's why it's called the Elastic Compute Cloud). Major internet servers (Instagram, Netflix, Reddit) use it without exposing anything to the users. My view is converging to believe that EA badly misunderstood what they were building, and built servers that followed an MMO or multiplayer FPS model, instead of distributed servers that are properly scalable. There is no good technical reason why there are multiple shards exposed to the user.

My guess is you have to pick discrete servers because the region files aren't replicated across the servers in a shard. Either that or each "server" is actual its own shard, but without having actually played the game I have no idea of how region play works in this version.

The level of love for the "in-depthness" of the sim during the SimCity talk is quite ironic considering all the news that came out today (traffic is completely broke and they have different houses/jobs every day)

Certis wrote:
Dyni wrote:

Also, hey CC crew, still waiting on that promised big Dark Souls segment. Can we expect it sometime soon?

Yes! Just been a scheduling nightmare with everyone's day jobs being busy and lots of a travel. We haven't forgotten though!

Sweet. Just checking

Quick question on Twilight Struggle...does the game truly take 8+ hours as mentioned in the podcast? BGG states 3 hours?

Just checking...

StewiesLoveChild wrote:

Quick question on Twilight Struggle...does the game truly take 8+ hours as mentioned in the podcast? BGG states 3 hours?

Just checking...

Not 8. 3 with 2 players who already completely know the game.

StewiesLoveChild wrote:

Quick question on Twilight Struggle...does the game truly take 8+ hours as mentioned in the podcast? BGG states 3 hours?

Just checking...

I think 8 was just a bit of dramatic license. If both players have played before and the game goes the max 10 turns then it could take 3-3.5 hours. More commonly it is done in 2-2.5.
Even when teaching my brother who had never played before it only took us 4-5 hours to play all 10 turns, with a lunch break.

Still an absolutely amazing game, and something that the podcasters only touched on a little is the amount of theme the game is dripping in. A lot of the geo-political/military systems at work in the game are based on the prevailing domino theory at the time, and how Mutually Assured Destruction and the MilOps drive players to mimic the brinkmanship of the real cold war.

Never has a game driven me to learn so much about its history either. Given I was only alive for the last 10 years of the cold war and from Europe, I really had no memory of it, but since playing Twilight Struggle I've read historical accounts and browsed wikipedia for hours, and the designers have done an incredible job of jamming almost every major military or political action into the game.

Julian went on blackout for Simcity?

Spoiler:

You build cities.

just kiddin Julian!

shoptroll wrote:

Excellent show. Really enjoyed the discussion about games as services.

Gremlin wrote:
shoptroll wrote:

You know you're in for an interesting segment when Jeff leads off with a joke about connectivity issues.

Higgledy wrote:

I wonder, knowing zero about the complexities of it, if there is a business in offering overflow servers to games companies for them to rent for a launch. I guess there might not be enough big launches to make worthwhile. Although that might change in the future.

I have to imagine that's why EA is using Amazon's cloud services. Most MMO games go through a rapid expansion of servers in the first week or two, followed by a drawn out consolidation of instances as interest levels out. I wouldn't be surprised if more companies switch to hosted VMs through cloud service providers like Amazon instead of rolling up physical servers which have a higher maintenance and deployment cost.

There's no excuse, since EA did use Amazon's cloud servers, which are pretty easy to scale up transparently (that's why it's called the Elastic Compute Cloud). Major internet servers (Instagram, Netflix, Reddit) use it without exposing anything to the users. My view is converging to believe that EA badly misunderstood what they were building, and built servers that followed an MMO or multiplayer FPS model, instead of distributed servers that are properly scalable. There is no good technical reason why there are multiple shards exposed to the user.

My guess is you have to pick discrete servers because the region files aren't replicated across the servers in a shard. Either that or each "server" is actual its own shard, but without having actually played the game I have no idea of how region play works in this version.

Shop, I think that's the point.

When you start thinking like a web services engineer, and pay attention to companies that are doing it right (NetFlix and other SaaS B2C or B2B companies), something like this should never happen.

How can management and team leads NOT think that this was going to happen? We keep seeing it happen over and over again - servers aren't set up to meet demand, and it just leaves a bad taste in everyone's mouth. But if you're able to just turn up a dial, because you've built the back-end to handle discrete elements of the service aspect, then this isn't a problem.

So I look at something like Netflix's Chaos Monkey, and wonder why EA/Maxis let this out the door without considering "what happens when we lose connectivity?" and "what happens when we don't have open server resources?". Netflix built its architecture with an understanding that they might lose resources at any time, and thus gracefully handles any exception is usage. There's been plenty of times that I've opened Netflix on my Roku, only to find that my personal recently watched history is missing - but that doesn't impact my ability to see my instant queue, do a search, or watch anything.

Instead of an "always on" connection, why not a scheme that fails gracefully? Maybe you're allowed a certain number of starts between connections, to accommodate the "airplane play" scenario? Or maybe some form of limited game-play options until you are able to connect again?

I am 100% with the CC crew - if you're going to force online play, there's got to be such an amazing benefit for me as a player. If not, then make your online requirements impact me as little as possible.

Great podcast, as per usual but I'm sorry guys, I keep calling this out for various threads and people but you absolutely cannot compare the failure of one company or product with another. They are not identical situations, they are not indicators for each other's behaviours... they're not even staffed by the same people.

I know people disagree with me but hear me out:

You never hear people saying "Oh, don't buy a Mitsubishi because Toyota had a recall." You never hear people saying, "Oh, don't buy the latest Toyota because they had a recall on a different model 5 or 10 years ago." You don't get people saying, "Don't buy [insert SF author here] because Heinlein dropped the ball in his post-troopers novels and they're published by the same company."

I mean, seriously. What is this? There's a lack of logic here.

Just because Diablo failed or EA published a game that failed 5 years ago (well, Spore was more than that I think but whatever) it's NOT an indication on how a present release will go. Every development is a microcosm in every industry. Sometimes lessons are learned and sometimes they're ignored... and it's not necessarily a progressive thing it's a project management thing and every project is different.

In summary, I think Elysium and the other guy (sorry for non-voice recognition) are correct.

I'll disagree with you Duoae.

We're not saying "don't buy from X because Y had an issue years ago", what we're saying is "why did X have the EXACT SAME ISSUE that Y had years ago?"

We know that, under the covers, there's a big difference between Diablo 3 and SimCity - but they both had problems at launch that, which to the consumer, seem to be "the same problem".

Lack of QA, insufficient servers and back-end infrastructure, and an always-on DRM requirement - those certainly seem to be "the same problem".

If a car company puts out a car with square rims, and it flops and has safety issues, we're certainly going to cry foul when another company tries triangular rims the follow year and it flops.