Looks like Half-Life 2 has been delayed – no specific date now.

Half-Life 2‘s existence was a heavily guarded secret – the game has been in development for close to five years now (shortly after Half-Life hit shelves, in other words) – until late May or so. Valve took the odd step of announcing the game to print magazines under NDA restriction until their June issues – places like Blue’s News once again had to get their dish from people whose subscriptions came in early. Part of the bold claim they made – in addition to showing it off to wowed audiences at E3 – was that the game would be in stores on September 30th.

The entire affair was significant since the usual course of events is to announce a game as soon as you have an idea for it – or at least a screenshot or two. The development goes on for a while – oftentimes years – and the game oftentimes suffers delay after delay. This is assuming it’s not cancelled outright – I’ve seen print ads for games that never saw the light of day. The publishing marketing machine coordinates the hype – which is why it’s so disasterous when games get delayed, since it throws off their timing considerably. Gamers are so jaded to delayed games that it’s hard to get excited about a game anymore. So what would be nice is if someone made an amazing kickass game, didn’t tell anyone about it, then came out with it with a minimum of hype but a maximum of sales.

And that’s what looked to happen with Half-Life 2 – the original Half-Life is like a textbook on how to do a game right. The single player portion raised the bar on what electronic entertainment is capable of, without being too ambitious for its own good (see Shenmue). The multiplayer portion was helped considerably by the fact that, by accident or design, Half-Life became the platform to modify. Counter-Strike is the most popular mod of all time, with some 50,000 players playing it at any given point in time (almost as many people as live in Texarkana, my home town). Couple this seemingly perfect game to make a sequel to with the fact that the sequel actually looked pretty good, and the fact that the game was coming out on time and early for the holidays and it seemed to be a match made in heaven.

Valve contracted the team that made Team Fortress for the original Quake to make their planned Team Fortress 2 modification an expansion pack for Half-Life. Then, some time after Team Fortress Classic was released to fill the gap, TF2 was to be a full game. Its current status is unknown, though it’s thought that more will be revealed after Half-Life 2 ships (one theory states that it might be included in HL2). Other than additional patches, additional Counter-Strike updates (by then Valve had hired that team full time) and the occasional boxed release of Counter-Strike and Day of Defeat (standalone HL mods – the 2 HL expansion packs were done by Gearbox), all was quiet from Valve – which irritated gamers. Where was TF2? Had they even thought about HL2? Everyone knew that one of the most successful start up game developers in the world had to be up to something – but what?

Valve was more or less hearlded once HL2 was announced, but shortly thereafter problems started to arise. For one, there’s been something going on I haven’t really followed where cards from Nvidia didn’t do some FSAA right, but apparently that was resolved. Then they announced the game was, confusingly, in three versions – one cheap version with single player only that wouldn’t run mods, “aimed at the Costcos of the world”, a second version complete with single and multiplayer that could run mods, and a third that also had limited edition crap. The single-player version is baffling – how many people could have a system capable of running HL2 but not want multiplayer? And shop at Costco for their games?

Then there was Steam. Steam is this sort of nebulous client Valve devised. It’s a way for them (and you) to update the Valve content on your hard drive – so that way when a new patch comes out you can get it automatically, so you don’t have to worry about waiting until most of the servers are patched to play on. Also it’s a server browser, chat interface, and even a way to purchase games – a fourth Half-Life 2 option is apparently to pay $9.99 a month to play it – which will also apparently guarantee access to future commercial expansion products. Also, it seems it is destined to become the authientication method for Half-Life and Half-Life 2 – currently when you want to play Half-Life online, Half-Life sends your CD key to an authentication server (currently operated by WON) to make sure it’s not already in use. Essentially Valve pioneered the CD Key system for games. In about a year or so that WON system is going away, and you’ll have to play with the latest version of Half-Life and the latest version of Steam.

This irks gamers because it means that they’ll suddenly have to have more crap installed in order to continue to play a game they already own. Even this wouldn’t be so bad, except that Steam has problems. For one thing, when thousands of gamers suddenly need an update, the thing is darn slow. Recently they took Steam out of beta and tons of users needed to update their Half-Life installations. So thousands of users were downloading hundreds of megabytes of data. Steam was/is supposed to migrate your existing content over and update only what is neccessary, but this didn’t work so well. It got to the point where Valve released standalone Steam installers, more or less negating the point of Steam.

There’s talk that you’ll have to have Steam in order to play Half-Life 2, either at the outset or in the future. This bugs gamers to no end for various reasons. Steam’s auto update feature is a nice idea, but imagine if it was designed such that you couldn’t play unless it had updated. So you sit down to play a quick game but instead you have to sit and wait for the client to download an update along with a million other users. It’ll be worse if it checks between levels. I don’t know for sure that either of these will happen, but it goes back to one belief a lot of serious gamers have – they want the control, and they like having things separated. They don’t want to use the server browser Steam provides – they like theirs just fine (which won’t be a problem, assuming it doesn’t purposely lock out other server browsers). They don’t want or need the “Friends” system – having four IM paradigms is too many already – plus this is something their server browser does already. And most people – myself included – would rather a traditional boxed product.

Now personally I don’t mind the idea of Steam, provided they can work out the bandwidth issues (and also provided that it’s even possible to do so). I have little use for the server browsing or chat features, but having the equivalent of a critical update notification for my games isn’t half bad. But I won’t buy games from Steam. It’s funny – years back when it was predicted that they would or could do away with the boxed product notion I was resistant, and I thought perhaps I was the only one. However, it seems I’m not alone. People on the whole want something tangible for their money.

But then after all the fiascos with Steam, Valve delays their game from its etched in stone date. Personally, I don’t mind since it’s not like it’s Duke Nukem Forever delays (yet), but people who just upgraded their systems are livid. I guess this buys me some time to do some research and see what I want to get (or have to get) to run Half-Life 2.

Still, it would have been really cool if Valve could have delivered on their crazy claim of September 30.

The process of game software porting fascinates me. Porting (I believe from the “port” in transport or portable) is taking software from one platform and getting it to run on another platform. Every company does it differently, some do it better than others, and some do it horribly.

A recent and upcoming example is Halo. Bungie was primarily a Macintosh developer who thrived on the dearth of entertainment titles on that platform. Back in 1994 or so when everyone in the world was playing DOOM on the PC, Bungie was making a killing selling their FPS Marathon to the Mac users who wanted in on the fun. At some point they decided to start working on Halo. Originally a 3D RTS, they changed it to a third person action game. The graphics were amazing and the big question was: would it be PC or Mac? They confirmed they were working with both platforms, but refused commentary on which would get the game, or if they both would. Then in late 2000 or so it was announced that Microsoft was buying Bungie. Gamers smelled what was coming. At the insistence of Microsoft, Halo was transformed into a FPS and was released in November 2001 as an Xbox launch title. Some time later Bungie announced that they had not “started” on the PC port, and had handed it off to Gearbox Software to do. It’s about to be released (nearly two years after the game hit Xbox) and soon a Mac port will “start”. The maneuver paid off for Microsoft – they sold millions of Xboxes and almost every single one has a copy of Halo to match.

Then there’s Electronic Arts. EA is in the lucky position of having a few franchises which translate well onto multiple platforms. Their crown jewel is their EA Sports line, which contains their flagship title, their Madden NFL franchise. The game Madden NFL 2004 was released fairly recently for PS2, Xbox, GameCube, PC, GBA and PSOne (BTW, the traditional moniker for the PlayStation (1) was always PSX, which was short for PlayStation X, the original code name for the retargeted SNES CD-ROM drive Sony tried to do with Nintendo – but now that they’ve announced a new PS2 with DVD writing and TiVo capabilities named PSX, most people – myself included – now refer to the system as the PSOne to avoid confusion. Too bad it looks stupid). The “next gen” (or is it “current gen” now?) consoles have more or less identical versions, and I think the PC version is the same as these. I can’t vouch for the GBA version, but as far as I can tell, the PSOne version is simply Madden 2003 with new rosters and updated jerseys. Since there’s obviously no code shared between the PSOne/GBA/”next gen” versions, by some definitions the GBA/PSOne versions aren’t “ports”, but rather “versions”. No one says the “GBA Port” of Madden 2004 since they’re clearly have different source code.

But then there’s the GBA “port” of DOOM, which I don’t believe actually uses the DOOM source code (I’d be willing to bet it’s able to achieve what it does on the GBA due to assembly language) but it still gets considered a “port” since it’s essentially trying to be the same game.

EA also finds themselves in charge of the Harry Potter games. Their second title, Harry Potter and the Chamber of Secrets, released in conjunction with the movie, was on seven platforms – PC, Xbox, PS2, GC, GBA PSOne and even GBC. For the most part they were completely different games. I think it also has the distinction of being the (so far) last GBC game.

Recently the aforementioned Madden NFL 2004 was released and the PS2 version outsold the GameCube version 35 to 1, and the Xbox version 5 to 1. However another high profile game also recently saw a simultaneous console release – Soul Calibur II. The original Soul Calibur was only released on the Dreamcast console (and in arcades, running on PSOne hardware), and many people bought a Dreamcast to play it. Soul Calibur II was released on each console with one major difference (other than improved graphics to compliment the capabilities of each platform) – an exclusive playable character. The PS2 version has a character from Tekken, the Xbox has Spawn from the MacFarlane comics, and the GameCube has Link, from the Zelda series. So far the GameCube version is outselling the PS2 version 5 to 1, and is even outselling the marginally graphically superior Xbox version. This says that people will buy cross-platform games for the GameCube when there is compelling reason to do so (and it also says that there are considerably fewer casual gamers buying SCII.)

Of course SCII and Madden NFL 2004 benefit from simultaneous release – it’s not uncommon for a game to come out on a platform, sell a ton, and then sell considerably less when released for other platforms. Part of this lies with the perception that the game was done best on the original platform (not always the correct assumption), the fact that most people have already purchased it on the original platform, and the fact that some games just don’t hold people’s fascination as long as others.

The one exception to this seems to be when one platform is a console and the other is the PC. These platforms are sufficiently different that people don’t have a problem buying or re-buying the game multiple times. A good test of this will be whether or not the PC port of Halo will sell. Knights of the Old Republic is Xbox only now, but many people are holding out for the later PC port. Another reason is that often the PC is immune to exclusivity agreements (like with the Grand Theft Auto games).

And it’s not always taking advantages of strengths that works either. The game Splinter Cell was released on the Xbox and its developed with the intention of exploting its abilities to the fullest, including shadows and such. As a result, it simply couldn’t be recreated on the PS2 accurately, so to make up for it that version contained extra levels. It fared better on the GameCube, but to make up for the differences they programmed the game to link up with a Game Boy Advance to serve as a radar – a nifty feature at least. Neither sold all that well.

Sometimes ports don’t make sense. Sim City was ported to the SNES and sold badly – one magazine said titles like this were “too computery” to work on a console. Conversely, Star Wars: Rogue Squadron was lambasted on the PC but hailed on the Nintendo 64. My favorite conundrum was the port of Quake 3: Arena to the Sega Dreamcast – despite the best efforts of the porting team to make the game work for a controller, they also included the ability to use the Dreamcast Mouse and Keyboard for players used to playing on the PC. Why these gamers would then not just play it on the PC is beyond me. But then again the PSOne port of Civilization II sold pretty well despite the technical limitations of the PSOne – speculation as to why mostly centers on the fact that no other PSOne games were quite like it.

To some, Gearbox is a porting house. I suppose that’s unfair since they have done some original titles – albeit expansion packs. But most of their games are ports of other games. Some places do their own ports, others outsource them. id Software used to do all of their own console ports, but they haven’t done one of their own in years – though that may change as they say they are “commited to delivering” DOOM 3 on the Xbox. Any ports to other consoles will have to be done by other people.

In so far as ports amongst Windows, Mac and Linux are concerned, most modern development is layered – when properly done the layer that interfaces the game from the platform itself is the only thing that has to be rewritten. To an extreme, all development can be done this way – this is how those games like The Hulk and Spider-Man can hit all three consoles and the PC on the first day. The trick is that a lot of important decisions have to be made from day one – witness Neverwinter Nights. The plan/promise was to deliver on all the major platforms on day one – Linux, Windows, MacOS and even BeOS. The BeOS port was scrapped when Be, Inc. got out of the OS business (they got bought by Palm and more or less dismantled), but the other three were a go. Seems the developer Bioware got fed up with their Interplay contract so they got Infogrames to get them out. Consequently, when the Windows port of NWN was done first, Infogrames ensured it was shipped first. The Linux and MacOS ports were in unfinished states. Though the initial plan was a few weeks at best, the Linux port was out over a year after the Windows version and the Mac version was later after that – and due to licensing issues, the movie cut scenes couldn’t be in the Linux version (not sure about Mac though). In addition, the rich toolset only exists on Windows.

Another factor in porting is the costs involved. As we’ve seen, when planned ahead of time actual development costs can be minimized with forward thinking, but sometimes the costs in manufactuiring and distributions aren’t worth it. The Dreamcast port of Half-Life was cancelled even after the porting development efforts were finished, since the dwindling market share in the wake of the console’s cancellation meant the cost of manufacuring copies wasn’t worth it. When it became apparent that modifications and multiplayer would be impossible, the Macintosh port of Half-Life was abandoned as well. Quake 3: Arena was released on Windows, Macintosh and Linux simultaneously in 1999 in separate boxes. 95% of the sales were to Windows, over 4% were to Macintosh, and less than 1% were to Linux, mostly because of Loki’s lesser distribution power and the fact that most Linux gamers dual boot with Windows anyway. Simply put, often times porting to a niche platform simply isn’t worth it.

In any event, I have a good PC, a GameCube, an Xbox and a PS2, so I’m pretty much set. Of course since most people have at best one, maybe two consoles, it makes porting possible – if everyone had all the consoles then ports to multiple consoles systems wouldn’t be neccessary – just make the one version and everyone will buy it. The only way I’m screwed is if the game is Macintosh or Linux only.

But that never happens.