Writing for Half-Life
marclaidlaw.com
marclaidlaw.com
It's a collection of innovations and feelings that few other games have been able to capture.
On an unrelated side-note, if you've played Half-Life, Freeman's Mind is a must-watch: it's an astoundingly funny bit of comedy that is just so brilliant.
Yahtzee Crowshaw's (yes, the very same guy I mentioned above) LP of the game (Let's Drown Out All Of Half-Life), while by no means a must-watch, is also reccomended.
I felt much the same way about Marathon, which ended up being under-appreciated by virtue of being Mac-only. A shame.
I've been re-playing the Metal Gear Solid series. The second and third game are better than the originals because of a control scheme update that was part of a bigger 'HD' update. But they still suck compared to the controls of the fifth iteration of the game, 'The Phantom Pain'.
MGS5 has some of the best controls I've ever used in a (third person) game. For lack of a better description, it made me feel almost like how I felt when I first played Mario64.
And I can't help but wonder why 1) it took three iterations to get here even though 5's controls could be realized even in earlier iterations of the game, 2) why other similar games don't just use the same controls instead of more dated approaches.
I mean, I get why, but it feels strange. I understand how the development of graphics, physics, AI, etc. is limited by technology but as long as you don't need a controller update for your new control scheme, such a limitation shouldn't be there. And yet it is, apparently.
OTOH, there are some old games that control fine. Megaman 2 still feels great to play, as does Quake with +mlook and WASD. But yes, control affordances matter. For the large part, Doom can only be played comfortably by modern audiences because of the work Source Port developers undertook to allow for rebindable keys, and sgtmarkiv's Brutal Doom, which allows for proper mouselook, if you really want to be that guy.
I've been doing Ludum Dare for a while now and it has only deepened my respect for games with good controls. For the last jam, I didn't spend nearly as much time on control feel as I usually do and I got a bunch of negative feedback for it.
For another example, people criticize Unity for games with poor feel, but what's really happening is that Unity has enabled a bunch of new developers to make games with less effort, and that includes spending less effort on getting the controls to work well. All Unity did was lower the bar to entry.
Take a look at Quake III, in bg_pmove.c PM_Accelerate
https://github.com/id-Software/Quake-III-Arena/blob/master/c...
This is the code that translates player input into acceleration for the physics system to handle. Notice the code blocked out with #if which fixes a bug but the fix isn't used because it "feels bad". The bug lets you accelerate to ridiculous speeds with the right pattern of movement.
That's not true, look at the code. It does something else. You can also check out the .plan from Jun 3, 1999:
> When I tried fixing the code so that it just didn't work, I thought it changed the normal running movement in an unfortunate way.
Uggh. This whole forum is full of sheep. I'll probably be downvoted by people who don't understand this gag. I get no respect.
Unfortunately, my prophecy seems to be self-fulfilling. GPP is down from +3 to +0.
Here is a playlist with the full series.
By the way, is Yahtzee still slumming it on the Escapist, or has he done the Patreon thing like Jim? He always seemed too good for that place.
Ah well, maybe he'll finally use that free time to make a proper version of Cell. Because the demo of that game was really cool.
Now, you have to ask if HL3 could ever hope to meet the expectations of the modern gamer, many of whom were probably just born when HL2 was big and have no real connection to it. What do they have to gain from it? What do they have to lose? I'd argue that they're making so much money now they don't particularly need the expense and risk of developing a game that, even if it is quite good, stands to potentially harm their reputation.
Aside from Steam, Valve (the developer) seems to be putting a lot of focus on esports right now. Many of the recent updates to Dota 2 seem to be new features and quality of life improvements for casters and spectators.
There is no way in hell they could release it as a VR game and not as regular FPS game. That would be just really bad move from Valve.
And for a long time even after it became a website, they had a weekly magazine with some really good articles. But that gradually got pushed further and further out-of-focus in favour of more videos until they basically stopped doing it entirely. And then of course they lost most of the worthwhile video series and now it's just a bland gaming press site with Yahtzee.
His early videos seemed to be interesting, but these days every video is an excuse to rant about clueless game company execs and overweight manchilds.
He does that sometimes, but certainly not every episode.
Wait, why is that? Isn't this 2016 where any game or movie that is even a little successful, must inevitably have successors?
How could valve motivate not making a game that would sell like hotcakes no matter how terrible it is? Sure, it would have enormous expectations, but that's true of all movie sequels (yet most suck). Is the value of Valve currently based on a game they'll never make, because it' can't live up to expectations? That would be an odd situation!
Even if Valve decides against ever doing HL3, could they afford not to sell the franchise? It must be worth a TON to e.g. Microsoft who could use the title it to push xboxes or VR headsets or whatever.
If they were to sell the HL franchise they risk losing a lot of customer support overnight. They would no longer be game devs, whose blunders can be excused because of their amazing games, they are money grabbers willing to whore out their franchises for short term cash grabs. The risk to the Steam platform just doesn't seem worth it.
If they were to sell it MS would be one of the worst destinations for it. If I were in their shoes I would likely explore having a smaller studio build it. A sale to MS could only be viewed as financially motivated and the quality of the game produced is almost certain to suffer.
However if I were them I would keep it as an ace up the sleeve to bring the masses to a new H/W platform Valve bring out. Such a release could seal the platform that wins its generation of consoles.
So the next generation of consoles or so they release a steam machine with VR tools and their 'next gen' games.
So still PC, but packagable for console users
I think it mostly boils down to Valve being a privately held company that is not under pressure to meet arbitrary quarterly financial targets. The article is quite clear that Gabe Newell gutted HL a few months before the original release date rather than deliver a game that wasn't quite "there" yet.
Additionally, if Valve have maintained their notoriously flat structure, HL3 will need to acquire critical mass /buy-in within Valve before it can make it out the door.
> Even if Valve decides against ever doing HL3, could they afford not to sell the franchise?
Yes, as long as Steam remains a money printing machine, they will not be desperate.
It simply delivered one of the best modding possibilities ever.
You bought HL and every week you could buy a gaming magazine with 5 new mods. Most of them crap, for sure, but it felt like paying once and getting new stuff all the time.
I never had a game installed for so long and played the "main game" for such a short time.
HL², on the other hand, I played for the story only, never installed any mods or something, but maybe I was already to old to run after this half finished games, haha.
Like I said, I got most of my stuff from gaming magazines before I had Internet, and I almost never saw any Quake mods on their CDs.
On the other hand, I was rather young when Quake was "a thing" and just started with PC gaming. I found out about these magazines a few years later.
I only worked on mods for UT and the original Jedi Knight, and I've only heard second-hand that UT and Quake (3? 2?) were easier to mod than HL1, so I might be wrong.
And compared to Jedi Knight and its only editor(s) being user-made, all these games were easy to mod!
Team Fortress, Quake Chess, AirQuake, ThreeWaveCTF, Xonotic (kinda, sort-of)
Think of three Half-Life mods, as fast as you can:
TFC, Counter-Strike, Natural Selection, Sven Co-op, Science & Industry
Think of three Quake 3 mods, as quickly as possible:
OSP, CPMA, ThreeWave (again), OpenArena
Think of three Doom mods/mappacks, fast:
Memento Mori, Brutal Doom, Batman Doom, Russian Overkill, The Sky May Be, Tech Gone Bad, and so many more
Think of three Minecraft mods, as fast as you can:
Thaumcraft, IC2, Buildcraft, Forge, FMP, Thermal Expansion, Better Than Wolves, RedPower, NEI, Ars Magica, Applied Energistics, Aether Mod, Tinker's Construct. You want more?
Think of three UT mods:
Ummm... Errrr... Ummm... If you count mappacks, somebody might have made a version of Aerowalk for it?
Now, that's just me, but I doubt others would have dissimilar responses.
True. Also I just learned that he actually finished the whole game. I used to follow this series as he was producing it, and I thought he gave up halfway through.
Likewise, if you've played Half-Life 2, you owe it to yourself to read Concerned: The Comic of Gordon Frohman
I remember reading a long time ago that John Carmack insisted that Doom and Quake needed nothing more than one interact button and the depth of level design of finding keycards to unlock doors.
Considering how ID Games have been received after Half-Life, i.e. criticized for their lowbrow gameplay and story, I wonder if he ever really understood how wrong he was.
Given the lack of ammunition, and the overall design, I find it hard to believe that there's a way to play it on which a significant amount of time would be spent shooting.
SS2 lets you use projectile weapons as your preferred way of disposing of enemies. In my mind, if you choose that path, it could be described as a shooter. That isn't all it is, but it is one word that describes a particular way you can choose to play the game.
Side note: SS2 is probably my favourite game so I certainly don't mean to "degrade" it by implying it is "just" a shooter.
It's a stealthy horror FPS, not a bulletstorm.
unlimited ammo!
Sometimes gamers have to learn to walk before they can enjoy running.
The player is supposed to be the avatar. You aren't playing Gordon Freeman, you are some space marine fighting the hordes of hell or dimensions incomprehensible.
You have an emotional investment in your character because your success or failure in combat dictates if you live or die. It is surely a simple connection, but it also certainly exists. It is implicit to almost any video game.
It makes much more sense it you view it as a target practice game instead of one with a narrative.
I think Carmack wins this one.
2) id games after Quake 2/3 really weren't id games. id wasn't even really id anymore. The guys from the gold old days had moved on except for Carmack and a few others, they became more of a technology company instead of a game company - not that there's anything wrong with that.
HL made its own legacy, and it's for a large part due to its mixture of story and gameplay. To paraphrase Yahtzee on the subject:
"The thing you have to understand about half-life is that in most shooters at the time, guns still hovered three feet off the ground, spinning gently like Barber's poles, and nothing else had ever felt so absorbingly real"
It may not be a coincedence that all but one of id's best games came before Romero left, and the other one was a pure multiplayer gameplay experience, rather than the standard story-driven affair.
It was so damn easy to mod, and so damn easy to stick mods on magazine covers. A single individual was able to turn Quake1 into a rudimentary flight simulator after all.
Later games have become more complicated to mod because of their adherence to realism in rendering, thus requiring the mod makers to be as skilled and equipped to make anything worthwhile.
OTOH, the actual language that Quake mods were written in, QuakeC, belongs in nightmares.
And you might say it's not that bad. You might say that it's okay, given its domain. And then you realize that Xonotic is still, to this day, written largely in QuakeC. And then you start to feel a bit ill, just contemplating a codebase of that size in a language that actually has less abstraction powers than C.
Combine that with the fact I grew up with Descent 2.
Disclaimer: I've been banned from Nexuiz/Xonotic servers. For winning. Using their rules. Written in QuakeC.
Botting?
That community has infinite salt, but it made my day^Wweek^Wmonth^Wyear.
There's Reflex, but that community never really seemed to form, and there's also the newly announced Quake Champions game that is basically a QuakeLive-esque game via Doom 4's multiplayer engine (whereas Doom 4's multiplayer gameplay as-is is a lot like classic Doom/Doom 2 multiplayer).
I'm looking forward to Quake Champions, but I'm afraid all the younger kids are going to bring too much salt to make it worth investing time into.
God, I hope it's not actually through Doom 4's MP engine: Doom 4 MP was pretty bad.
And there are other small scenes scattered here and there: Robot Roller-Derby Disco Dodgeball, Quake 2, Sauerbraten, Warsow, and so on.
Seriously, have you tried Warsow? That might be your sort of thing.
Wolfenstein and Doom were both great, eye-opening games because their rendering gave us a visual experience we hadn't seen before. Nobody cared if they were the gaming equivalent of dumb, loud rock. Besides Doom's art had a terrifying -- if brainless -- grandeur.
Quake also delivered a step up in rendering, a great feat that I admire as a software engineer. But as a game, it was dull.
It was no more brainless than Doom, but we had already seen that particular kind of dumb-horror before, and the full 3d really just felt like an incremental improvement.
My guess is that you only played single player. Single player Quake is indeed less exciting than Doom, largely because of lower enemy counts. The people correctly praising Quake as a classic played multiplayer. In multiplayer the 3D environments really matter. It's more than just an incremental improvement because full 3D gives you so many options for movement. Quake has very high skill ceiling and mastering movement is a big part of it.
However, in terms of an argument between story-driven vs. just-shoot-em-up, I think single-player is the most relevant.
I still love the atmosphere and total weirdness of the game.
Wrong how? Lots of players don't give a damn about story in games (me being one of them).
Doom 3 kind of forgot this, but Doom 4 has definitely brought it back.
The modern niche for lowbrow shooters is filled by pseudo-simulationists CoD and GTA.
Half-life was always more cerebral and science-y. Valve took FPS in an even more scientific direction with Portal. You could argue that that limits their audience by being highbrow.
I seem to recall the draw of CoD vs MoH was that you weren't one lone soldier vs the axis of evil. You actually had teammates! And scripted events!
The fact that Half Life 3 isn't released and there is no sign of it is almost certainly not due to some failure to produce it but due to a decision not to.
The problem is that the only "new" thing since HL2 that could take off is VR which isn't quite there yet.
At this point games are to Valve what email is to Google. An important product to be sure but not what they make their money with. Valve makes their money with Steam, their games are side projects.
Steam is popular, the games Valve makes are popular and Valve is popular. HL3 is a massive investment, a huge risk to all of that and for what? If Valve were a games company, the money would be the answer but Valve is the Steam company, that money is not all that much to them.
I thought the Portal demo on the Vive was mindblowing. If they released Half-Life 3 in that form, it would survive the anticipation.
No Man's Sky wasn't a simple case of hype, though. The developer blatantly lied about what was going to be in the release version, including videos and screenshots of content that doesn't exist[1]. The extent of this mismatch between what was advertised and what was delivered is so large that it's being investigated in the UK[2].
[1]https://www.reddit.com/r/Games/comments/4y1h9i/wheres_the_no...
[2]http://arstechnica.com/gaming/2016/09/uk-advertising-board-i...
Anyone can make a complaint with ASA, and they investigate any complaint for which they're the regulator.
They also do not uphold plenty of complaints.
To get a feel for what an ASA ruling is like people can read the archive of rulings here: https://www.asa.org.uk/Rulings/Adjudications.aspx
I think this recent ruling about a city council's campaign to stop people giving money to beggars shows there's some nuance to the rulings: https://www.asa.org.uk/Rulings/Adjudications/2016/9/Nottingh...
tl;dr An ASA investigation means nothing other than someone complained.
Valve isn't developing HL3 because a project of that magnitude requires simply more dedication and conviction than Valve's devs might even have anymore. The company is one of those flat structures where developers work on whatever they have passion for, which means anything they make is gold, but if nobody has passion for it it will never be made. And compared to the work required to make all their games since HL2 - not skill, just raw hours of design and testing - HL3 is way beyond the scope of what Valve has made since then. Portal 2 is the most significant achievement since then, and all it really was was level design and event scripting - there are no real enemies, no AI, and the game length was still quite a bit less than what a full FPS release on par with the old Half Life would expect.
It should be no coincidence the release of Steam ended the Half Life franchise. Once the company had effectively infinite cash the staff could pivot from making complex hard to develop games that would make huge returns to making whatever they wanted - Dota 2 has nothing pay to win, no payments to even play, the only in game transactions are for cosmetics. It is something no other moba, not even Blizzard's, is willing to risk because everyone else is too worried about making sure ends meet from the games revenue.
So I do not believe Valve is incapable of living up to the hype. Even making a game as good as Half Life 2 would be enough. Portal 2 was considered a success despite being inferior to its predecessor because it was still really good. The problem has nothing to do with fear of failure, but a lack of willpower in Valve staff to put millions of hours into making a worthwhile successor to Half Life 2 when they can get similar accolades for much less effort making games in other genres. It is the same reason you so rarely see expansive and good RPGs. They are much more time consuming to make than almost any other genre of game.
I think the point you're making on structure, could also explain the failure of Steam Machines. Something that would require quite a lot of people coordinating not just with others in the company but also with people outside it.
I find it hard to believe there are not loads of //notes// about it. Maybe not a "final_script.txt" but lots of data spread out over lots of files seems logical to have.
script or story notes they had for it
Have you read the fan-assembled timeline? http://members.shaw.ca/halflifestory/timeline.htm It contains some pretty juicy speculation.Half-Life was just such an experience. Even almost 20 years later, it might be the best game I have ever played.
I can't get google to give me a cache of the actual article.
Edit: server's fixed