Slashdot Mirror
Half-Life for Macintosh Cancelled
Bobbi Style wrote to us with the word from insidemacgames that the Macinoth version of Half-Life has been cancelled. Valve Software sent out a letter explaining that they had found that the version of Half-Life that would be released for the Macintosh was not up to what they wanted the game to be. Rather then treat Mac users as 2nd class citizens, they have decided to simply not release it at all. Disappointing news, however, which also casts doubts on the viability of projects like Tribes 2 for the Mac.
This could happen to any Linux game port in a heartbeat. As far as 'coexistence with Windows', it's a very bad sign. It might be that Linux can't trust the commercial world any more than the Mac can. What a rotten day for gaming (also today iEn, makers of Warbirds and Dawn Of Aces, laid off a big big chunk of their staff)
another thing you might want to think about is that maybe mac users don't _want_ games which are a year and a half old.
Game companies frequently release their mac versions just as the game is ending its cycle of life; frequently the release of the mac version is simultanious with the release of the windows version of the sequel, or the windows version of the next generation of the same kind of game. and then they just assume the mac users won't mind, because hey, they're mac users.
Like mac tomb raider I and II.. they came out at the same windows tomb tomb raider III was released. Mac Starcraft came out at about the same time that the PC Broodwars expansion pack became really popular.
These are bad examples, because tomb raider I and II for mac had extra levels that the windows version didn't, and tomb raider III looked pretty bad from the PSX-based demo i got from Pizza Hut, and Starcraft still has a very healthy online gaming community based around it even so long after the original release. But the point is the same; why would mac users want the cold, leftover crumbs of the windows world?
Even if they HAD released halflife, why would we want it NOW? Quake3 (like halflife only better) is just around the corner, and some form of Tribes (an ORIGINAL game, instead of just more derivitive mindless point-and-shoot..) may be released. Having Halflife would have been cool at the time it was released, but now I for one am no longer very interested in it.
If the game companies really want any serious showing from the mac users, they should develop all versions of the game at the same time.. or exersize *gasp* good programming, and have a bit of an abstraction layer and write the programs in such a way that instead of calling things in MS-propeitary functions directly, you call functions of your own that call the MS API up for you.. so that later instead of doing the mac/linux version ground-up you can just replace the abstraction layer functions with Sprockets/X (i'm totally BSing this, by the way, i've never written a game and i don't know whether you could really do this without taking a massive performance hit. Maybe if you inlined/#defined the functions it would..? never mind)
Anyway, the point is, if you release the mac version a year later, when all the mac users have been playing it on friend's PCs for over a year, the mac users are _much_ less likely to actually buy the game. Meaning you'll make less money.
Oh, and then a funny thinghappens; the game companies just scoff at the idea of porting to the mac in the first place, saying "mac users don't buy games". And ignoring _why_ mac users don't buy games, and if there's anything they could do to get the mac users to buy games..
[ramble] PS.. whatever happened to ambrosia? they used to be the bomb. Avara was so far ahead of its time it was rediculous. Ah well. Anyway, i want tribes 2 for mac. and a mac capable of running tribes 2. Instead of this 7200/75 with no hardware acceleration. WORMS ARMAGGEDDON ROCKS!! [/ramble]
Irritable, left-wing and possibly humorous bumper stickers and t-shirts
John Carmack said that only about 15k of Quake 3 is platform-specific, when the first test came out. Bungie has certainly done well, and they don't seem to think their focus on cross-platform simultaneous releases has been a waste of resources. It might take a little more to start the project off, but once you have the engine firmly cross-platform, you end up releasing for two platforms (with an easy way to port to other platforms if desired) for less effort than it would have taken to write for one specific platform at the outset and port to another one later.
One thing to bear in mind is that while the Mac market is smaller, there is less competition in that part of the marketplace. Releasing for Windows means advertising in a lot of magazines and fighting for recognition against a slew of other games. Releasing for Mac involves fewer platform-specific magazines to advertise in, not to mention that most of the hype for the Windows side would spill over to the Mac side, and it also means that you're releasing to a market starved for current titles.
Valve may have done the right thing - it is true that the Mac community might have held a grudge if Half-Life had been a bad port. If it had been a good port, though, they might have been able to make up any losses in future sales, since they would have established themselves in the Mac market as a provider of good games.
Naked.
Sure it's possible to work around problems, but the question is at what time/expense? The game is based on the Quake2 engine, IIRC, which is starting to get a little on the old side. Not onyl that, but there are new cool games coming out (Q3, Unreal Tournament, etc). It's kinda hard to sell a game that has been around for around a year, and by the time the problems were fixed would have been over 2 years old, when your competition is the likes of Q3, UT, and similar titles. Maybe they just decided that they wouldn't be able to get Mac Half-Life out in time to succeed. Let's not get too paranoid about the great anti-Mac conspiracy :)
.plan file for details). Since Half-Life is based on Q2, there are I'm sure many issues with porting to the Mac (or anything else for that matter). The port in this case is NOT easy.
Another issue to consider in your comparisons with Quake, Q3, etc is that those games were designed from the beginning to be relatively platform-independent. This makes porting to another OS somewhat trivial. Quake 2 on the other hand, was not (read John Carmack's
It seems to me that you're getting upset at the company because after they made an honest effort (as best we know) to make the Mac port, they found it can't be done in a reasonable time schedule (gee, they only have to do a little hacking of the engine, reverse engineer some networking protocols, etc etc.) Give them a break. Things don't always work out how one might want them to. I'm sure they would have loved to ship the Mac version, any company likes to make money. But if they can't do it, they can't. I don't see much point in boycotting Sierra for this, I really don't think it was an attempt to screw over the Mac community. And anyway, I doubt Mac users boycotting the company will have any significant effect, except perhaps to make the PHBs think future Mac ports aren't worth bothering with.
The state of Mac video cards right now is:
The original iMac (Rev. A) used an ATI RAGE IIc with 2 MB VRAM, expandable to 6.
The Rev. B, C, and D (233, 266, 333 MHz - 266 was the beginning of "flavored" iMacs) all use the ATI RAGE Pro, with 6MB VRAM. All A-D iMacs are PCI-based.
The Blue & White (Yosemite) G3 Macs use ATI RAGE Pro video, but run on a "special" 66 MHz PCI slot.
The Wall Street and Lombard PowerBooks use the ATI RAGE Pro LT. Lombard may use AGP, I'm not sure. I know that Wall Street is PCI.
iBook uses the RAGE Mobility video chipset - a tweaked version of the RAGE Pro LT that's supposed to be pin-compatible. It has 4 MB of VRAM (embedded), and the iBook uses a 2X AGP bus for video.
Finally, the G4 macs all use the ATI RAGE 128 chipset on an AGP 2X bus.
The ATI RAGE 128-based cards are available as retail cards for the Yosemite G3 Macs, as well as for earlier G3 and PCI PowerMacs. But they're constrained by PCI.
Virtually all earlier PCI Macs (clones included) use the older ATI Mach64 chipset and it's cousins. The PowerBook 3400 and the original PowerBook G3 (also known as the 3500 or Kanga) use Cirrus Logic video chipsets, with no 3-D acceleration on board.
I do think that there's OpenGL support for the RAGE IIc and Mach64, but I'm not sure. It wouldn't be too fast, though, given the constraints of both PCI and the earlier ATI chipsets. Fortunately, the Rev. A iMac is the only one that sold in high volume, and it was only on the market for about 2 months (it was announced in May, but it wasn't available until August) before they up-revved it to the B and switched to the RAGE Pro.
- -Josh Turiel
-- Josh Turiel
"2. Do not eat iPod Shuffle."
Andrew Meggs was the same guy who walked away from the RC5 project back when Steve Jobs cancelled MacOS clones. He made this big political statement by taking his source code for the MacOS RC5 client and saying he was protesting this action by Steve Jobs. Christ! I thought after that stunt, no one would trust that guy enough to involve him in a project. Now I read in the first paragraph of the letter (http://www.insidemacgames.com/news/index.shtml)fr om the guy at Valve, "Andrew Meggs at Logicware has been doing a good job on the port..." What the hell?!?! So he got tasked with the responsibility of a port of another piece of software to the MacOS and this time he screwed up the project by running it over budget. I really wish people would blacklist that guy until he gives up his grudge against Apple. What a jackass. sjohnson|at|smart.net
However, as we got closer to shipping the product and reality set in, it was increasingly obvious that in order for us to break even
on the Mac version, much less be profitable, we were going to have to cut some corners.
Essentialy, its a profit - cost over market_share equation. I'm sure the modern busines model promptly said 'cut' and there goes HL. The next step is the 'spin.' Here's another little snippet from Valve.
They are happy because we do our best for them, and that's what they expect from us in the future. -SNIP- I would much rather we just eat the money we've spent so far than take money from Mac customers and short-change them.
Now they're suggesting that they're a responsible company and trying to keep up a decent reputation by 'eating up' the cost. When in reality the could not sell such a shabby product, if they wanted to or not, without a larger loss of revenue.
Don't quote me as either good or bad here, this is just business as usual.
At least spin is portable and multi-platform.
I find it annoying that other users besides Mac dudes don't find this incident disturbing. Game developers are increasingly ignoring markets like Macintosh and Linux because they're "small." It's very insulting to think that countless millions of users are ignored, and their potential revenue dismissed as petty. Some people think that the API that Apple uses is pathetic and we should all bow down to Microsoft's DirecctPlay. What is wrong with Apple's API acceptance of OpenGL? Plainly put, it is a portable, scalable, and industry standard gaming API that runs on any platform. With Half-Life being ditched from the Mac, it sends a heartless message to consumers and developers: You aren't anybody unless you use Windows(TM)... no matter how crappy, proprietary, and ugly the code.. it just makes more money for developers.. period. Bottom line is everything and consumer choice is irrelevant. Things need to change and I'm hoping that you Linux/Unix/whoever guys will realize that when Mac users get shafted by software developers it will eventually or indirectly affect you too! In some ways Mac users and Linux/Unix/other users share the same disadvantage of being underdogs in the marketing/corporate standpoint. Only by resisting such developer prejudice can we make the Windows-diluted press realize that there are "other" people on this planet. It's obvious that the mainstream press doesn't give a damn about Macs/unix.. CNN's online site list of top 10 budget pcs don't even mention the TOP SELLING iMac nor do does their top power PCs include any PowerPCs! They regard Unix and its variants as rogue software only used by hobbyist tinkers and not by people of productivity. We need to gain more public attention... we can't get ahead by just always watching for Microsoft's "daily bug and security flaw" (however it's so invigorating to constantly see this in the news). We need to contact developers and let them know what we think and tell them that dammit... WE DON'T USE WINDOWS AND WE ARE REAL PEOPLE.. AND WE DESERVE SOFTWARE!!
OpenGL at Apple's site
-----
Linux user: if (nt == unstable) { switchTo.linux() }
Those who laugh at you for you having a Mac.. are the people who constantly call you to fix their PC.
I'm happy to see someone willing to stand up and take action when a project is all over the floor.
Hey Gabe! Throw the port source under GPL and let a fresh set of eyes hack a bit.
Half-Life for the Mac is not being cancelled because of DirectX, or poor performance on a crappy OS, or because it's based on the LithTech engine, or because it's impossible to make a decent first-person shooter on the Mac.
Half-Life is based on the Q2 engine. Folks, Q2 is available for the Mac right now, and it plays well. Unreal runs well, and you can expect Unreal Tournament and Q3A to run well, too.
Half-Life for the Mac used OpenGL, and was working great. DirectX is not part of the equation.
Valve (the developer) made the decision to cancel the port, not Sierra, the publisher. Note that this is a little unusual as the publisher generally pays a fee to the developer in exchange for the right to license their material to a different firm to actually handle the port, so ordinarily Sierra would be the ones cancelling it.
The reasons that Half-Life for the Mac was cancelled had mainly to do with networking issues. I'm going to take a wild, uninformed leap and suggest that the reason for this is that Half-Life for the PC was written to use DirectPlay, Microsoft's extremely proprietary networking protocol. DirectPlay cannot and will not be ported to any platform other than Windows, as Microsoft won't license it, and most likely sees DirectPlay as a competitive advantage to keep the best games on Windows. Half-Life for the Mac may indeed have been a victim of this strategy.
I'm not that worried about game support on the Mac. It's getting better:
Apple did have a proprietary gaming API (gamesprockets). And a proprietary 3-D api (qd3d/rave). And a proprietary game networking api(netsprockets). And according to developers who have used them, every one of those api's was vastly superior to the then-current standards. And guess what? Not many games. Now that Apple's pretty much killed all those in favor of cross-platform standards like OpenGL, we're getting some pretty good games. And the current generation of Mac hardware, while a little behind current Wintel standards, is still pretty darn cross-platform as well: PCI, AGP, USB, etc. (You can take a 3dfx card out of your PC and plug it right into a Mac. Not too shabby. And there are two independent programmers working on TNT2 drivers, too, so Mac users won't be stuck with *cough* ATI forever.)
I still would like to see Valve change their minds, and release Half-Life for the Mac even if it is crippled. Sure, there's a vibrant online community surrounding Half-Life, but the single-player game is what really turned the PC gaming world on its head last year. Here's hoping Valve revisits their decision soon (and here's hoping the Mac zealots don't knee-jerk themselves into doing something stupid like mailbombing Valve).
--
Anonymous cowards are working on a massively multiplayer shared first-person persistent virtual reality that's cross-platform, open source and hosted on thousands of independent servers.
Having experience with MacOs, Windows, Linux (Suse) and BeOs, I believe I should set the record straight here. MacOs is NOT the worst system out there. Granted, it has plenty of flaws, and granted again, you shouldn't buy it if you want to play games. However saying that MacOs can only run Adobe software is idiotic. There are a number of things that Macs do better than other platforms. One is recording multitrack direct-to-disk. With its better multitasking, Linux could do that even better, but it lacks the software (VST effects, etc) and support for high-end sound acquisition hardware. Between Mac fans touting the mac as the panacea and those poking fun at it without really knowing much about it, it's amazing how many myths float about this platform. The truth is the Mac is still relevant for a limited market of people with very specific creative needs (Publishing, music recording, etc). This relevance may dwindle if support for BeOs increases, but this battle is not won yet. Also MacOs X, due early next year, may bring considerable improvement to the platform, as it should be Unix technology dressed up with the Mac's superior interface. That is, if Apple doesn't screw up again, something they seem to be as good at as designing good GUIs...
My goof. I thought it was Rage Pro. I'm sure I've got the PCI-66 part right, though.
- -Josh Turiel
-- Josh Turiel
"2. Do not eat iPod Shuffle."
>What do you think...is Sierra dying? Will they
>be dead and gone by a couple years from now?
As far as I'm concerned, when they laid off the entire Yosemite Entertainment group (most of the classic hard working Sierra folk), that was when Sierra slit it's collective wrists.
Then to see whats left of the company butcher it's product line, (like you said, with some of them EAGERLY awaited and almost completely ready to ship), is like watching watching an old friend on his death bed writhe in agnoy.
This bit about the Mac Half-Life being cancelled really sucks. I'm not a Mac guy, but I sympathize with them. I had high hopes for a Linux port of the client, also denied.
If it's as difficult to port as it sounds, then damn. The actual game itself is nice enough, but some of their design decisions leave me a bit mystified: being forced into 640x480x16-bit to access the game menu (even while playing), hiding access to the 'console' behind a command line switch, etc. If that sort of mentality carried though into the source, I'm not suprised.
(But hey, the game itself is some of the best FPS action I've ever experienced so far. A shame they inadvertently limited that experience to only the Windows 95 crowd.)
-- www.bteg.com | bleh.n3.net | hac47.dhs.org
There WAS a superior cross-platform 3d API than DirectX. It was called QuickDraw3D.
It was "Steved" in favor of OpenGL.
Ask anyone who used it, it was a great high-level 3d API. It lives on in an Open Source project called QUESA (they're trying to make it live on top of the Mesa OpenGL implimentation).
"The number of suckers born each minute doubles every 18 months."
These are my friends, See how they glisten. See this one shine, how he smiles in the light.
Actually, testing on a Mac is a b*tch. One bad pointer and your machine is toast, due to no memory protection. Forget to relinquish control of the program and the OS never saves you, you have to reboot. Best to do lots of debugging on the Windows machine and then struggle through the macintosh development afterwards. But if Macintosh had some level of security in the operating system, you could probably get done testing a lot faster than on a PC. Then again, I've never tried to collect 'one of everything' for PCs before I released a program. If you are releasing an OpenGL game, you are pretty much limited to three machines - G3, G4, and third rev iMac (they changed the video card when they started making them in other colors). Now the iBook and the new version of the iMac. That is still a lot of hardware to test on. I usually just test it on the PC by swapping out video cards. Hardware is abstracted for most things besides 3d video, so if there is a problem it is more likely a system issue than an appication issue. So if issues like that come up after release, I just work with the person (sometimes even having them send me their computer for testing!) to get it fixed for them. So far, it has always been the user messing with the registry settings or bad hardware/ram.
...here I'd been telling all my Macophile friends about how good Half-Life was, and some were even looking forward to playing it. I'd been hoping I would be able to gib them on my favorite 3D-shooter game, but alas...and if Gabe's note is to be believed, I wouldn't have been able to anyway.
I posted about this to the Half-Life newsgroup, and thus far every one of the 10 or so responses has been resoundingly antiMac ("Great! Now they need to cancel the Macintosh!"). Losers
When I mentioned this to some of my chatroom friends, they wondered if Gabe's note might not be the whole story. Sierra, it seems, is being gutted from the inside out, with even some of the more popular games being pulled for no apparent reason (such as the B5 flightsim, which was essentially ready for release when it was canned...what sense does that make?). They seem, my friend suggested, to be staking their life on "Deer Hunter".
What do you think...is Sierra dying? Will they be dead and gone by a couple years from now?
I'm glad that Half-Life: Opposing Force is being made by another company...means there's much more chance we'll actually get to see it.
In related news, has anyone heard about the Christian first-person-shooter game that's hitting stores this week? There was a story in the NY Times about it the other day. It sounds like a dumb idea, but then, so did "Deer Hunter" and look at how well that's done.
Editor Emeritus and Senior Writer, TeleRead.org
An engine isn't 'Multiplayer only' by design.
Anyway, the q3a is already being used for Star Trek : Voyager 3l33t SomethingOrOther and Heavy Metal : FAKK2, neither of which look like intense multiplayer games.
That's what this is. Logicware could easily have made Half-Life for the Mac have total feature-parity with the PC version, or even have added features (as evidence, I give you Westlake Interactive, who has done this with Quake, Unreal, the Tomb Raider series, Shadow Warrior, Unreal Tournament, Total Annhilation, Railroad Tycoon 2, Alpha Centauri, Madden NFL 2000, and others). The fact is, they chose not to.
Why? I don't know for certain, though I have a few suspicions. I do know that Logicware's Andrew Meggs, head of the Half-life project, is very anti-Mac, and has been since his shareware company fell flat after releasing one game. He even says as much if you read the original press release; he's happy to not be working on a Mac project anymore. But hey, why was Sierra dumb enough to put a well-known Mac-loather on the Half-Life team, anyway? That would be like putting John Dvorak on the design team for the next generation of iMac.
Look. None of the reasons Sierra cancelled the project were actually valid, because all of them could have been worked around. Quake/MacOS is compatible with PC mods, so Half-Life (which, last I checked, was based on Quake) has no excuse not to be. Quake/MacOS's networking code works perfectly with its PC counterparts, so Half-life has no excuse there, even with this DirectPlay stuff (if worse comes to worst, networking protocols aren't that tough to decode, especially when you're paid to do it, and since reverse-engineering to achieve compatibility is known to fall under the fair use clause there's no problem there).
Whatever. I won't be buying Sierra or Logicware stuff for a long time because of this. Even though it probably won't happen, I hope that Sierra takes their code to Westlake and pays them to finish it; they'll do it right.
I know what I'm going as for Halloween! (hint: it's UNCLEAN!!!!)
Hey guys. I'm just attaching this here since it's near the top. There's a petition going on to save Half Life at Demand Mac.
The problem with game programming, is that much of the code is related to grpahics and sound. This is not possible to do universally. There are two ways to do this.
1. Program to the metal
This way, you write code the directly manipulates the hardware and provides the fastest possible speed. The drawback is that you now have code that is specific to one video card/sound card. Basically, you write your own procedures for sending the assembly code via interupts. This is the old way. Remember DOS games where you would select your sound card from a list? The list was REALLY small until manufacturers released useful APIs for programming to their cards.
2. Use an API that encapsulates the hardware
This is the DirectX/OpenGL/Glide approach. (Yes Glide is 3dfx specific, but it still abstracts the hardware, and you could write a glide driver for anything, and I believe that there is one for the TNT2).
This requires that manufacturers provide the implementation for the API for their hardware. i.e. you decide what to do, the implementation deals with the hardware side.
This is how modern games are made.
The problem with a unidersal code approach to games is the API. Without a standard API, you need to write all the graphics and sound code specific to the machine. Because of different capabilities between platforms, attempting to use universal approaches won't yield optimum performance.
In Computer Science, we worry about the Theta or Big-O running time, and ignore the constants. This is fine for scientific algorithms, but not for games. In games, a Theta(n) game will be much better than a Theta(2n) game, and therefore, the hardware must be used to the max. In a scientific algorithm, we don't care. We worry if processing more elements increases exponentially (i.e. 1 year for X, 10 years for X+1!!!), not if it increases the time from 1 year to 2, because we just buy multiple machines and divide up the task until we are satisfied with the time requirements.
Games are a different breed of applications, with a different set of rules than traditional programming.
Alex M. Hochberger
Computer Science and Engineering
MIT '01
With my physic abilities I look into the Valve's meetings with the dev team
OoooOoooOoOOoooo OoooOoooOoOOoooo OoooOoooOoOOoooo
Marketing guy: we need the Macintosh market, if we can get Half Life on those cute little iMac's well sell millions
Dev dude: ya, but iMac's only have a rage 128 (I think). It's no game card. We might be able to get the performance out of a G4, with a snazzy new card
Marketing guy: video what?
Dev dude: also we build the game on top of Monolith's engine, which is build on DirectX.
Marketing guy: and.....
Dev dude: well how are we going to re-write Monolith's engine to work with OpenGL or a 3D API that works with mac's
Marketing guy: ya... sounds great. Write up a spec and when you think we can start shipping beta' we need this in time for Christmas.
Dev dude: excuse me.... BUT ARE YOU NUTS? This will take at least a year! We don't have anything to work with!
Marketing guys: [blank stare] ok well, who feels like lunch?
OoooOoooOoOOoooo OoooOoooOoOOoooo OoooOoooOoOOoooo OoooOoooOoOOoooo
-Jon
this is my sig.