ATI - AMD’s “Advances in Real‐Time Rendering in 3D Graphics and Games Course“, during the highly technical SIGGRAPH 2008 conference, included a whole chapter dedicated solely to the implementation of the shader framework and the rendering techniques used in StarCraft 2.
As the name implies, the paper, which won’t be an easy read even for seasoned programmers, is a part of a course with the following prerequisites:

Prerequisites
This course assumes working knowledge of a modern real‐time graphics API like OpenGL or Direct3D, as well as a solid basis in commonly used graphics algorithms. The participants are also assumed to be familiar with the concepts of programmable shading and shading languages.

The StarCraft 2 engine definitely has a split personality, and for a good reason - there is a fully interactive, playable “Story Mode”, which requires a completely different form of real-time rendering. Rendering four high-resolution characters and a detailed static environment is quite unlike rendering two hundred units with hundreds of constantly changing lighting sources, but both of the above scenarios must be supported by the same engine.

Starcraft II is supported by an engine that in many ways has a split personality; during normal game play we typically render scenes from a relatively far away distance, with high batch counts, and a focus on action rather than details. At the same time, we really wanted to push our storytelling forward with Starcraft II, and this is where the game’s
“Story Mode” comes in. In this mode, the player generally sits back to take in the game’s rich story, lore and visuals, interacting with other characters through dialogues and watching actions unfold….


…story mode generally boasts lower batch counts, close‐up shots, and a somewhat more contemplative feel – all things more typical of a first person shooter.

All technical details aside, Chapter 5 includes one over-quoted statement which presumably reveales the GPU system requirements for StarCraft 2.

Yet we also wanted to utilize the full potential of any available hardware to ensure the game’s looks were competitive. This meant supporting a widerange of hardware, from ATI Radeon 9800/NVIDIA GeForce FX’s to the ATI Radeon HD 4800s and NVIDIA GeForce G200s, targeting maximum utilization on each different GPU platform.

The above statement effectively puts the minimum requirement bar at an ATI Radeon 9800 or at Nvidia’s GeForce 5 family - cards which have been introduced more than 5 years ago. Blizzard’s developers are staying loyal to their tradition of aiming for the mainstream and putting an emphasis on gameplay instead of intense system punishment.

Reports from the Game Convention in Leipzig are streaming in, but considering that the presented build is actually identical to the World Wide Invitational one, nothing major is expected to be either announced or discovered during the event.
IGN’s brief interview with Frank Pearce, Blizzard’s senior vice president of product development, hasn’t revealed much, but it did confirm that there is no super-secret  StarCraft 2 closed beta testing group, and that the beta is months away at best.

Currently, the testing for StarCraft II doesn’t include that many people. “It’s like a limited internal alpha phase,” says Pearce. “The development team that’s working on it is playing it and we’ve released some alpha builds to some other development teams internally and then the quality assurance department is also playing it.”

…We asked when the field of testers might expand to beyond Blizzard’s walls. “That’s months away,” said Pearce. “We still have to expand our internal alpha to include the rest of the organization, iterate on that feedback, and then we have to evaluate what kind of beta we need, whether we need one at all. There’s a pretty high level of certainty that we’re going to need a beta and then we have to decide if we want to do a host beta or a closed beta or both. So, months.”

Another interview was given to VideoGamer, with Frank’s line remaining consistent - things are going well, but the game is far from being being ready to be released.

VideoGamer.com: Which leads me nicely on to the next question which is when will they get the chance to play the game? Do you have even a rough ballpark for when the game’s gong to be released?


FP: Not really. We still have a lot of work to do. Anyone that’s had the opportunity to play it at any of the shows we’ve shown it at might think wow, this feels pretty good, this feels pretty complete, so why don’t they just release it? But we still have a lot of work to do on the Battle.net side and we still have a lot of work to do on the single-player campaign side. And until that stuff is done, the total package isn’t there.

The second part of the interview, however, is interesting, and sheds quite a lot of light on the development and decision making processes inside Blizzard. When the interview touches WoW and the possibility of a StarCraft MMO, Frank puts things in perspective by reminding that World of WarCraft has been in development since 1999, and says that bringing the StarCraft Universe to the same level won’t be easy.

VideoGamer.com You’ve made it difficult for yourself!


FP: Yeah, we’ve made it very difficult for ourselves and it’s an overwhelming thought! I’ve been there at Blizzard since we started on WoW and it’s been a very long and arduous journey. And to think about taking that journey again with the StarCraft franchise is a little scary! It makes an RTS like Starcraft 2 seem very achievable [laughs].


VideoGamer.com: Fans say a StarCraft MMO would be great, but they don’t see your side of things do they?


FP: The World of Warcraft team is 135 people. The StarCraft 2 team is 40 people.

During the interview, Frank also confirms that unlike StarCraft (which was ported to Nintendo 64), StarCraft 2 is not planned to be ported to any sort of console, mainly due to the game’s fast pace.

FP: It’s not something we’re specifically planning for right now. The development team is working on the game and the interface with the PC and the PC interface peripherals in mind.


BC: One barrier to entry there is just the speed of the game. If you look at how the RTS genre has evolved a lot of it has slowed things down, make things a little bit more, not a plodding place, but a little bit more of giving players more of a chance to evaluate. StarCraft is really action packed.

All in all, It looks like once again, StarCraft 2 fans are going to have to wait for BlizzCon for real news and information about the game.

Blizzard will be introducing a massive cross-game Achievement System, which will be featured in all three announced Blizzard projects. If things go well, the system is also likely to be applied retroactively to existing Blizzard titles.

World of WarCraft’s lead designer, Jeff Kaplan, has told MTV Multiplayer that the Wrath of the Lich King’s Achievement System is just the first step…

…he revealed that eventually players will have a Blizzard Account that shows Achievements from other Blizzard games they’ve played, including the upcoming titles “Diablo III” and “StarCraft II.”

n

“Eventually, our plans are for the Achievement system to become an account-based system,” he explained.

The Achievement System will undoubtedly introduce a long term, life-time spanning reward system for loyal Blizzard fans. It might even include retroactive achievements which won’t be available to newcomers, as promised by Blizzard on the Wrath of the Lich King achievements web page:

Feats of Strength will be awarded retroactively (that is, they’ll be immediately granted to your character if you qualify) since many of them will be impossible to earn in Wrath of the Lich King.

Blizzard intends to unify all the player’s achievements under a single Blizzard account which will contain a permanent record of achievements. Having to make an educated guess, based on the interview and the somewhat similar systems implemented on other platforms, Blizzard will probably keep record of your highest WarCraft 3 ranks, Diablo 2 levels and official StarCraft tournaments, as well as other multiplayer and single player in-game achievements.

You’ll have this Blizzard identity, and you’ll be able to see things like ‘Oh, this guy was great at Diablo III, but he never played Starcraft and he was mediocre in WoW.’

It’s been more than a month since the last official Q&A batch. This does not come as a surprise, of course, since the entire Blizzard crew was busy dealing with the WorldWide Invitational event in Paris.

This Q&A batch is composed of a long, comprehensive “Chat with Devs” introduction, dealing with the new Vespene Gas mechanics, and six new answers to fan-submitted questions.

Chat with Devs: Since the Worldwide Invitational in Paris, the topic of the new Vespene Gas mechanic has come up a lot across many different fansites and message boards. Thus far, this is one of the biggest changes which will affect the macro management of bases in StarCraft II. To shed some more light on this new mechanic, I have gotten a chance to talk to Dustin Browder, our Lead Designer for StarCraft II, about the progress thus far of the new mechanic, as well as the objectives this new mechanic is designed to achieve.

  

To start, the new Vespene Gas mechanic is to further distinguish the play style in which players gather minerals versus gathering gas. In the original StarCraft, the gathering of gas was very linear in the rate in which gas is gathered. Often, players would put 3-4 workers on the gas, and the players would forget about it until the geyser was depleted. Minerals on the other hand, were much more exponential in the rate of growth and were also often played differently amongst different races. Zerg would likely expand rapidly with less drones in each expansion and Protoss/Terran could sustain a sizeable force with higher numbers of workers on a smaller number of expansions.

 

How the New Vespene Gas Mechanic Works
For StarCraft II, with the new Vespene Gas mechanic, players will have 2 gas geysers at their starting position. These geysers will start with X amount of gas (currently 600 and subject to balance) and at any time players can purchase additional gas in their geysers for X minerals (currently 100 and subject to balance). With each purchase of additional gas for your geyser, the geyser increases with X gas (currently 600 and subject to balance) and the geyser shuts down for 45 seconds. When a geyser is depleted, workers will still be able to gather gas at a low rate of 2 per round (subject to balance).

 

How the New Vespene Gas Mechanic Plays
With this new gas mechanic, players have a wider variety of strategies in developing and maintaining their refineries, as well as additional attention needed to make sure they are collecting gas at the most efficient rate. On the production side, players now also have to decide between sticking to Tier 1 units longer, or to play it balanced with one geyser, or even max out on gas to invest heavily on teching and higher tech units. Additional, the relationship between minerals and gas have an added layer of depth since investing in additional gas will actually cost the player minerals as well. How often a player invests in gas will also not necessarily be consistent through the game too and will depend upon what units that player is currently choosing to mass. Scouting too has an added layer of depth as well, as a players gas collecting play style may determine if the player is teching to a higher tier mineral heavy unit (like a Dark Templar) or a higher tech gas heavy unit (like a High Templar).

 

Overall, players will have to build the appropriate buildings as well as gather resources in a particular method in order to execute a certain strategy at a professional level. It is the hope of the development team that this new mechanic will not only make gas collecting more interesting, but also increase the amount of macro management skill needed to compete in StarCraft II at the top levels while at the same time making the game playable for mid level players without using some of these more advanced techniques.

Aside from an adjustment to the numbers involved, the system described is the same as the one revealed during the WWI event, which we have already covered:

All these major changes have been implemented to StarCraft 2 for one purpose: complicate the “macro” portion of the game, which has been downsized severely with the introduction of new and improved user interface aspects, and mainly, Multiple Building Selection (MBS) - which allows players to select all their similar unit producing buildings together and deliver a single command to construct units out of all of them.

These changes will steal some of the added focus to the micromanagement portion of StarCraft 2, which pro-gamers, who’ve had a chance to play the game extensively, have reported about, and divert it back to base and economy management. Other such changes, meant to give players more macromanagement decisions to play with without dumbing down the UI, are being considered by Blizzard.

Check out our post about the subject here.

Moving on, the Q&A portion has interesting new tidbits and clarifications:

1. How exactly does the Corruptors attack work? Is it a stacking debuff that takes effect after a certain number of stacks? And if so can the debuff time out? Would you be able to hit and run kill for example Battlecruisers with a few Corruptors over a couple of minutes?

 

The attack is technically a debuff, but does not do damage over time. Every time the Corruptor attacks a unit, it’ll leave a debuff on it for a couple seconds. If the unit dies within those couple seconds, the unit will be corrupted.

This description of the Corruptor is incomplete, as evident by the Zerg reveal video, where a group of Corruptors corrupted a few Terran fliers on their own. The Corruptors likely cause damage themselves, or at least damage that goes towards Corrupting a unit which does not actually lower its physical hitpoints. Aside from this, the Corruptor also leaves a “Corrupting debuff” on the attacked unit, which causes it to become Corrupt should it die while under the effect.

2. What are the current stats and build times for the Queen defensive buildings?

n

To catch us up, Zerg defensive buildings arent built by the queen anymore. Instead, they are built from the drone once again.

 

Spine crawler:
- Only hits ground
- Health is 300 (uprooted health 100)
- Movement speed is 2.25 (1 off of creep)
- Damage is 20 +20 armored
- Range is 7
- Attack speed is 1.5 sec

 

Spore Crawler:
- Only hits air
- Health is 300 (uprooted health 100)
- Movement speed is 2.25 (1 off of creep)
- Damage is 15
- Range is 7
- Attack speed is .8608

The developers have decided that the Zerg deserve means to create defenses that don’t rely on the Queen, which can’t be everywhere at once. The new defensive colonies, aside from being granted mobility, are almost identical to their StarCraft 1 versions. However, the Queen likely still possesses her “Swarm Infestation” ability, allowing her to turn any building into a defensive structure temporarily.

3. The Queen seems to be a very potent unit, although its tasks are more defensive ones, it can be used in crazy rush strategies, according to Karunes experiences. Well, if the Nydus Worm was able to transport even queens, she would get even more potent. You could easily think of crazy rushes using your opponents creep to just overwhelm him with your units AND your defensive structures. So here is the simple question: Can the Nydusworm transport queens among all the other units?

n

Defensive structures will not be able to enter the Nydus tunnel network, but the Queen will be able to. Furthermore, the Queen will no longer be able to build defensive structures. Drones will morph into defensive structures, similar to the original StarCraft.

 

4. When the Zerg Sunken Colony is uprooted and on the move is it more vulnerable to enemy attack?

 

Yes, the Spine and Spore Crawlers will have less hit points while they are uprooted. The actual number of hit points will be determined through balance testing.

 

5. In many cases, the micromanagement of units in StarCraft revolved around gameplay mechanics (Dragoon dance, Mutalisk stacking, Reaver/Shuttle micro, etc.), rather than special abilities with cooldown/charges (Stalker’s Blink, Phoenix’s Overload, etc.).

 

Is the amount of this kind of special abilities in SC2 a concern of Blizzard, and how would this affect the overall gameplay?

 

Players will still have dependence on both gameplay mechanics as well as special abilities. For instance, Stalkers will have the basic dancing mechanic as Dragoons had in the original StarCraft. Marauders are another unit highly dependent on micromanagement to get the most effectiveness out of the unit, making sure you use their attacks slowing effect at opportune times.

 

Though for StarCraft II, we are introducing much more positional micromanagement, which will amplify units damage significantly. A Colossus will fire in a line and lining up that radius with the enemy units will be crucial in battles. Flanking Jackals from multiple angles will surely add to its potency as well.

 

Overall, we definitely want to balance the game with both plenty of gameplay mechanics as well as special abilities that create opportunities for the players to initiate clever strategies as well as innovative maneuvers on the battlefield.

Unlike what many people may think, Blizzard’s developers know what made StarCraft 1 the great game that it is, and are not likely to forget to implement one of the most important and fun aspects of battle in their new game.

6. Terrans currently appear to be at a disadvantage in terms of troop mobility, (as compared to ‘Warp-in’ and ‘Nydus Worm’) are there any plans to bring back the, ‘drop-pod’ or other new transport mechanic?

 

Actually, we consider the Terran side to be quite mobile. Let us first look at the Reapers. This unit is the fastest ground unit in the game which traverses terrain without even having a spotter. In addition, the Medivac Dropships, allow added mobility to all Terran ground units. With the addition of the Dropship being able to heal, it has become even more of a staple in Terran strategies, giving even more increased incentive for players to build Dropships than the original StarCraft. On top of this, Vikings providing both ground and air support at a click of a button, gives that added support of mobility and options when moving a Terran army around.

 

The method in which Terran will be mobile is indeed different than Warp-in and Nydus Worm, though they are not considered less mobile than the other races.

Also worth mentioning is the Terran’s ability to relocate their buildings by lifting them into the air and repositioning them on the ground. Aside from that, the Salvage ability makes it easy for Terrans to create impromptu production bases, knowing that the investment will not be lost.

In accordance with the StarCraft tradition, one race lacks a distinct ability that is available to the other two races, while still managing to function in a balanced way. While the Terran race may lack a point-and-click transportation method which the Protoss and Zerg now have, they will not be at a disadvantage on the whole.

Dustin Browder, StarCraft 2’s lead game designer, has given an interview to IncGamers’s brand new StarCraftWire.net. He discusses everything from the map editor to super units in length and provides some insight into the gameplay design process.

Highlights:

So when will the Starcraft II editor become available to modders?

 

I don’t know yet, it’ll certainly be shipped with the game but I don’t know the answer to whether it will be earlier. That would be cool though.

Could the StarCraft 2 map editor be released before the game? Blizzard has not done this before, but some other companies have turned to this move to increase interest in their game before it is released, especially if it has been baking in the development oven for a long while.

Concerning 3D graphics, what’s tools do you need? In StarCraft it was just using paint and opening it and just adding new units. What 3D do you consider for modders? Will you be giving your own editor for the graphics?

 

I think we may ship the art tools but you still have to have a 3D program, there’s no way around that.

 

Well, StarCraft seems very… “conservative” would be a good word I suppose. Were there any other concepts that you discussed at the beginning of the development process that you ended up scrapping because you felt; “no, we want it to be like it was”?

 

Some… By the time I started working on the project three years ago the team really wanted to make a game that was true to the legacy of the original StarCraft. So in terms of “are we being too conservative or not?” We talked about that every day. It was a constant discussion.

It’s obvious that the same debate the fans are always engaged in - is StarCraft 2 enough of a “new game”? - is shared by the developers as well. Currently, it looks like Blizzard has managed to stay true to StarCraft 1 while enhancing it in almost every possible way and while introducing new ideas that further differentiate the races, creating a new experience if not a completely new game.

When asked about the Rock-Paper-Scissor relationship between units, Dustin provides a good explanation for how unit interactions in StarCraft work:

It’s not even just that, of the relationships in StarCraft; lot of them aren’t even Rock Paper Scissors, which is one of the things that makes the game so exciting. One of the standard RTS paradigm is that we use the Rock Paper Scissors but a lot of the relationships, especially in the early tech tree of StarCraft, are positional based. It’s not so much *whether I beat you* its *where do we fight*.

 

Zerglings will crush Zealots in the open field. They’ll just overrun them completely, and these are both the core units. Whereas the Zealots, at the choke, will just kill hundreds of Zerglings based on the Zerglings getting all trapped up behind. So in addition to where you fight there are also the questions of micro that are really interesting.

 

What we are really worried about are overlapping roles, it’s a constant struggle for us, but if you go there and play it now you can find a couple:

 

Player: “What about these guys”
Dustin: “Yes, I know!! They overlap, oh my god”.

Many RTS games completely rely on unit statistics and their predictable interactions to determine the outcome of battle. It is indeed one of StarCraft’s greatest strengths that so many other factors affect the result of an engagement between units, and produces equal match ups between even the lowliest of units and the most powerful ones.

Which particular unit has been the most problematic one when working, balance wise, not only for the Zerg but the other races as well?

 

The Mothership - hands down.

 

What is the problem with it?

 

Well, by its size it doesn’t look like a super unit from other RTS games. The problem here with super units is we want every unit to be a decision. There’s no point in shipping a unit if the player says: “I have to build that, it’s powerful, I should always build that. If I get to this tech level, I build it.”

 

We don’t have that in StarCraft II, the games meatier then that; you have to work for everything. So we don’t like the “super units”… The Mothership visually seems to suggest that, but at the same time we don’t want that to be part of the core gameplay experience so we’re continually balancing the ship, we’re continually looking for a new spell kit for this unit to make sure there are reasons to build it, really solid reasons, and really solid reasons not to build it.

 

Today the really solid reasons not to build it, and it’s been this way for many months, is the Mothership can’t really, cost for cost, defend itself effectively in the air. This means you can’t have air superiority dumped on the Mothership. If you already have your superiority, go for it, the Mothership is a good addition, but like I say; “it’s been a big pain.” It’s not helpful that it’s located at the end of the tech tree against how fast StarCraft games can be and how brutal it can be. There is a great chance that you can die within the first three or four minutes. So even in our play test process we don’t get to see this unit as often as we like:

 

Designer: “Did you get to play with the Mothership”
Tester: “Well, no, we didn’t, we topped out at the Immortals and Stalkers/Colossus”
Designer: “Ok ok ok, play again, play real hard”
Designer: “Did you get to play with the Mothership”
Tester: “No, no, we didn’t really get that high in the tech tree”

 

So once in a while it happens and then we’ll get some data but it just makes it a lot more complicated. We get a lot more data obviously on play testing Stalkers, Zerglings, Marines and all these guys because you see them all the time.

The Mothership is one of the units which has gotten the most attention - both from fans and from the developers - and has changed a lot accordingly. The Monthly Topic initiated for its sake generated a lot of comments and ideas, and Blizzard is still trying to nail down the right formula for making this unit worthy of the StarCraft 2 Protoss.

How did you decide to change the Queen that dramatically? Like, from a flying unit to a ground unit, from a mass unit to one-only unit?

 

It started conceptionally as kind of a story hit, we felt like the Queen was an opportunity to create a creature that owns the base, that lives inside and that she somehow lays eggs maybe, she’s monstrous and evil - like a Queen that you may see in an ant hive or a film.

 

We wanted that Queen, because we felt like the old Queen didn’t really hit that vibe. For example; that Queen could be called any other name and you’d be fine with that. It wasn’t a Queen in the classic sense, so that was the core idea.

 

How could we make this unit into something that feels like the Queen and we tried it a bunch of different ways and for awhile she was laying eggs to create weak organisms, so for example, there’d be different types of Hydralisks running around. That was kind of fun but then it got confusing, like which Hydralisk could attack me in the air, and what do I counter that with, I don’t remember.

 

So it was kind of a problem, so we’ve sort of settled down on this base defender which seems pretty successful, I don’t know if it’s really good enough for StarCraft II [but] that’s what we’re going to find out. It seems pretty fun.

The Zerg Queen is another unit that has gone through many changes, but this one is more set in its role due to story considerations, unlike the numerous incarnations of the Terran Thor and the Protoss Mothership. While the current “base mother” role is not likely to change, there is still plenty of room for flexibility left in terms of implementation.