What the hell are those ten people left supposedly doing? Like, if Valve's "managers" are too afraid - sorry - think making Half-Life 3 may produce reactions that reflect poorly on them as a business (again I have to say, have they seen Steam Support?) to the point where they're sabotaging the product'a development, why leave 10 people working on it? What are they going to accomplish that 100 people (their figure) couldn't? Especially if Valve has no desire to release HL3 at all? Valve doesn't have a problem with canceling projects (Stars of Blood is a recent example) after all. Hell, if they're working on HL3 that means Ep3 was essentially canceled as well.
Like, if Valve decided the direction they were going wasn't where they wanted, and as a result wanted to go back to a more conceptual portion of development, I'd totally believe that. But the idea that Valve's "managers" wanted to kill the project, but instead of just killing it, they slowly removed team members until a 10-man crew remained doesn't really make sense to me. What purpose does it serve if they have no intention of releasing anything?
Well, if we take what they have said about their design policy at face value, the reason to have that 10-man crew on would be to keep looking for the thing that's going to make the game worth developing. So basically looking at story concept, gameplay concepts, whatever pieces they can find that could eventually be slotted into a perfect whole. The long-term purpose of that team would be to (potentially) create a pitch strong enough to convince Valve that the game should be made.
So I was watching this Funhaus podcast from after the Half-Life 3 news, where they go into more detail about it, what happened, and their opinions on the whole thing. The first 10-11 minutes is them taking the piss out of a dumb letter, so I've tried to start it around where Bruce starts talking more about the situation they reported on: https://youtu.be/FeXWNETnxw8?t=10m40s
I figure those that think they're assholes for reporting what they did will still think the same, but I appreciate them talking about it like this. As someone who watches a lot of their stuff, they're being pretty honest and open here, not defensive or trying to hide behind jokes.
edit: okay so the embed doesn't start when I wanted it to, so I guess skip to about 10m40-ish.
darleysam on
0
Options
Dark Raven XLaugh hard, run fast,be kindRegistered Userregular
Hey did y'all know Portal is appearing in that new Lego toys to life thang Dimensions? Chell, Turret and a Companion Cube!
...my question is, what is that stuff under the cube? The clear panels could be like, flip pad things, but it looks like it's holding a heart shaped potion? Any idea what that's supposed to be?
SteamDB has typically been on the level. Though we know they've been working on HL3, this doesn't really tell us much about how far along it is, or much about the game period. Could all be old info.
SteamDB has typically been on the level. Though we know they've been working on HL3, this doesn't really tell us much about how far along it is, or much about the game period. Could all be old info.
The file is definitely legit imo. It is also probably recent for it to find its way into a Dota 2 update that it wasn't already in before.
Not much solid info in the file, but here's all you can kinda deduce
-it has stuff for procedurally spawning groups of NPCs based on stuff like whether the player is nearby or not, what's visible to player, etc
-it has "quest citizens" who have different behaviors from other citizens. If a quest citizen is a "quest member" it's "neutral to combine" and doesn't take damage. My first thought is "neutral to combine" means this citizen won't open fire on combine on sight but might still do so if attacked by combine, which is probably a useful thing for a lot of types of quests? This is supposition (though I do have knowledge of HL's relationship systems, etc etc).
-quest citizen has "beckon player when player is close enough" functionality but I'm not sure if this is just inherited from npc_citizen and was present in hl2/episodic
-"prop_zipline" exists so that's a level-placeable zipline of some sort, which doesn't really tell anyone anything
-there are a number of quest-specific info entities and vars. Of particular interest is "height of marker beam", "radius of bottom disc", "trace up to place top of beam", etc. Sounds like vertical-beam quest markers in an open-ish world map.
-"Combine Pulse Ceiling Turret"? Were those a thing in earlier games?
-there's an npc_hunter_invincible which is interesting. I'd have thought you could just make a hunter invincible rather than making a new class called invincible hunter. So that's a thing
-generic pickup has a "prevents respawning" var and a "associated playercounter name" var which are both marked as "HL3 only". The respawning thing only really makes sense to me in the context of an open world game
JUST SPECULATIN'.
edit: some of this stuff could also have slipped in there from other projects/games, or it could be references to shit from years ago
I'm guessing neutral to combine would mean the combine wouldn't attack the Quest NPC, because if your quest NPC got killed before you got or turned in the quest, that'd kinda suck.
Depending on how the game is yeah that could be it. These are just named vars you would set on an entity from hammer, so yeah, it could be anything. The reason I suggested the opposite is in the HL2 ai releationship system, a citizen being neutral to combine would usually be taken as meaning he doesn't care about the combine, but the combine might still care about him. Every npc type likes, hates, or is neutral towards every other type so you can set up an ai relationship where dude A defends dude B while dude B tries to kill dude A or whatever (awkward)
Also the Black Mesa guys have a patch planned for december
Greetings all!
We've been a bit quiet on the forum front for a while, but we have been steadily toiling away on Content Update 2! Content Update 2 is going to be a huge, awesome update, and we're super excited for the coming weeks. Over the next few weeks, I will be periodically showing off snippets of what's to come in Content Update 2 to you guys. There's a lot!
First up, in today's post: CU2 will feature 3 new weapon models. The Crossbow, the Shotgun, and the Gluon. All are completely new models/textures. The Shotgun and Gluon still currently (and probably will continue to) use the old animations, the Crossbow has also been completely reanimated. Stylistically, the Crossbow and Shotgun are both quite different to the models which you can see in retail at the moment (which are the same as the mod release), and are now much closer to their HL1 counterparts, design wise. This is part of our effort to bring our viewmodels all up to the same high standard and ethic for design, and this marks an end to our weapon remodels. The Gluon is pretty much the same looking as it was before, just with a higher polycount and a cleaner texture.
You may have saw the crossbow update that Chon posted. We originally had two game updates planed, but decided it was better to merge them into one big update for December. We historically have been all about surprise updates, but now we want to be more open to show people that we are making progress, and to hopefully build some excitement for the December update. So throughout this month I'll be posting previews of the coming features.
Weapon Drag
First up we added weapon drag. It's a seemingly small feature, but I think it adds a lot. You can set how much sway you have. So you can blow it out if you love it, or turn it to zero if you hate it. There also is an editable amount of weapon bob when you walk.
New Egon Model
We made a new Egon model. It is very similar to the last one, but more polished and without the deleted back faces.
Gman Arms
We created some Gman arms, so when you select him in MP the arms are accurate. He also no longer explodes into HEV gibs :P
Bug Fixes
-Fixed Steam Cloud
-Flashlight is now brighter and has a wider radius!
-Fixed satchel detonation to be faster and not switch mouse buttons on last satchel
-Fixed the MP5 grenade spawning from then center of the screen when fired
-Score persist on leaving and rejoining a MP match
That's just a quick sample. I'll be back through out the month to show off some new stuff that you guys/gals will hopefully find more and more interesting!
Welcome to December Update, Part 2.
Important Point: This update WILL NOT contain Xen. We are hard at work on Xen and are preparing media to show to you guys. But there will not be a release of Xen soon and we still do not have a timeframe for it.
Here are some more interesting details for what's to come in December's Content Update 2. This list is not exhaustive. The update is huge and there's going to be A LOT of stuff.
Godrays
Godrays (also known as Crepuscular Rays if you want to be more scientific) are the first of the new graphical features we have written with Xen in mind. We're testing them by deploying them publicly on our existing singleplayer and multiplayer maps. Very pretty, and highly customisable per level - useful for modders and level designers.
Shell Ejection
Marines and other players in multiplayer now eject weapon brass when firing. Seems like a small touch, but in a big firefight with multiple Marines it can actually add quite a lot to the atmosphere!
AI Tweaks
We've started our first round of AI tweaks. There will be more to come. One big one was tweaking the way Marines react to Alien Grunts. They used to use grenade avoidance behaviours to try and escape the Alien Grunt's Hornets, which led to some very odd behaviours in fights between them. They now no longer do that. They would also try and avoid their own grenades which led to some odd behaviour too. They don't do that anymore, either.
Multiplayer Tweaks and Changes
Multiplayer has seen quite a bit of love for this update, though there's plenty to come for you singleplayers too!
Firstly, all of the base multiplayer code was completely rewritten from scratch. This is HUGE. MP should now be far more responsive, better performing, and less buggy/crashy.
Fixed prediction/lag compensation on the Crossbow. Zoomed fire should now be extremely responsive and reliable.
Added subtle spawn effects/sounds for weapon and player spawns.
Added a "walk" function.
Unlocked all movement related cvars to allow us to better tweak movement and feel to match HL1.
Removed some Taus and ammo from most maps, to better match the fact that the Tau is really, really powerful now.
Better colours on the scoreboard/coloured killfeed and coloured chat for better clarity.
And many more! We may reveal more later!
Bugfixes
Fixed .357 Guard being unable to fire.
Fixed Hivehand falling through the elevator on C2A5E.
Headcrabs now properly inherit momentum when shot off Zombies instead of just falling to the ground.
Crossbow's lowered sensitivity when zoomed in now works properly. No more weird feeling.
And many more! We may reveal more later here too!
Crossfire, and a Singleplayer Update
Next part, we will talk about the main MP focus of the update: Crossfire! That's right, Crossfire is coming in the December Update, better than ever! And we also have a Singleplayer Update lined up for Content Update 2, which we suspect a lot of you have been anticipating and excited for (reiteration: it is NOT Xen!)
HL3 showed up in some Steam databases. These may be faked (as explained below), so nothing to get worked up about. I figured someone would post about them eventually though. Here's the skinny from SteamDB https://steamdb.info/blog/recent-package-name-additions/
Today we updated SteamDB with package names from Valve's new help website after Nephrite found a method to get them.
Valve recently implemented functionality to automate some frequently asked Steam support questions such as permanently hiding a game from your Steam account and these package names being exposed are a result of them not hiding them when they added this. A similar issue occurred with app names as well.
Due to the way Steam works, we are now also able to guess some app names that were previously unknown based on what packages they are in. An example of this can be found here, where the 3 packages it is connected to clearly indicate what the name of the app would be. We ran a script that detected these kinds of apps and added their name to the "Last Known Name" field in the table on the top. This can also be seen in the app history with the source being "Best guess based on help website disclosure".
While this information comes straight from Valve and many of these kinds of leaks turned out to be real in the past, it is important to note that third-party Steam developers can create and name their own apps/packages, a noteworthy mention of developers doing this are these packages, in which they are trying to XSS Steam. (EDIT: It looks like the XSS packages might have been by Valve themselves. Thanks for letting us know, aiusepsi!)
Many things could also be old entries and as such no longer be in development/coming to Steam. An example of a previous leak that ended up being nothing (yet) is Halo 3.
Dude, it totally checks out! I mean, I'd be a bit more skeptical, but there are THREE entries related to Half Life 3! And, the SUBID for Half Life 3 is itself divisible by 3! PLUS, 66300 DIVIDED BY 3 IS 22100, WHICH EVERYONE KNOWS IS THE FLYWHEEL LOCKING TOOL FOR THE 6.6L DURAMAX ENGINE FOR ALLISON AUTOMATIC TRANSMISSIONS, THE VERY SAME TRANSMISSION USED IN THE HALF LIFE 2 SCOUT CAR!!!
I for one am totally convinced, and will be picking up Hallmark "I was wrong, whoops!" cards to all those people I informed about considering HL3 a rather remote possibility of 2012 or so.
It's been confirmed by PCGames and a few other sources, he's retired and no longer at Valve. People are already spinning it a number of ways; he finished his writing job and retired, HL3 soon or he's just plain done and HL3 written by Marc Laidlaw is never going to happen.
Nosf on
0
Options
mojojoeoA block off the park, living the dream.Registered Userregular
Probably, but as he's mentioned in the past, Valve isn't a "here's the script, let's go make the game from it" kind of company. (Very few game developers are.). They figure out what the games is going to be, create a story outline, they build off that, the story changes based on the development cycle. Bits get cut, changed or moved based on the design. Stuff will get thrown out wholesale before the game even makes alpha. Story will change to service game mechanics, new ideas that come along, or ebcause stuff became too difficult or unwieldy to add to the game.
HL2 has a lot of shit that got cut out of the game or changed. (the borealis opening, the atmosphere processor, child laborers)
Obviously I don't mean to say that he's written a tightly-plotted script for Half-Life 3 that goes all the way from the opening train ride to the final, reality and time shattering finale.
I mean that he's detailed character arcs for Alex and others to take, designed plenty of background material for various planned settings, and left behind an enormous Half-Life bible with all kinds of information for people to incorporate into the game.
It's not really Valve's style to tie a product to a specific platform's release. Even HL2 came out over a year after Steam did.
Note that November said, "In the first few years,"
Also note that HL2 was well slated as an incoming product well before it's release.
I don't know what more needs to be said. The lead writer who was a fundamental part of making the magic happen is now gone, Valve has never actually come forward and announced plans for HL3 and the only bits of leaked material the Internet has been able to find to keep the 'imminent HL3!' myth alive for all this time have been scattered .txt files and old domain registrations (it was rather funny when the Internet was totally convinced that the Pronunciation Channel project was related to HL3).
Half-Life is a valuable enough IP that no doubt someone somewhere down the line is going to make something with it. I doubt that Valve sees that kind of effort as worthwhile, given the success of their non-game products and the risks currently involved with AAA publishing.
HL3 development probably spun it's wheels for a while - long enough to create a bunch of refuse that eventually found it's way to people looking for anything to latch onto - and then quietly died for any number of reasons.
Probably, but as he's mentioned in the past, Valve isn't a "here's the script, let's go make the game from it" kind of company. (Very few game developers are.). They figure out what the games is going to be, create a story outline, they build off that, the story changes based on the development cycle. Bits get cut, changed or moved based on the design. Stuff will get thrown out wholesale before the game even makes alpha. Story will change to service game mechanics, new ideas that come along, or ebcause stuff became too difficult or unwieldy to add to the game.
Yes, this is pretty much how it goes with writing: you sit down and create a bunch of semi-related, semi-coherent plot points and background exposition, then you just shrug and walk away from the project and someone else just starts fitting the pieces together until you have the Great Gatsby.
It's not really Valve's style to tie a product to a specific platform's release. Even HL2 came out over a year after Steam did.
Note that November said, "In the first few years,"
Fair enough, I totally see how my comment was super misleading by mentioning a timeframe. My bad. My point there was supposed to be that its not really Valve's style to plan content around platform availability. They'll release a game when its done. I mean, hell, they released an Xbox version of HL2 a only a year before the 360 came out.
Also note that HL2 was well slated as an incoming product well before it's release.
Yeah, but given it wasn't finished until well after Steam's launch, it doesn't strike me as them waiting for Steam to be out before they released it, which is my point.
I don't know what more needs to be said. The lead writer who was a fundamental part of making the magic happen is now gone, Valve has never actually come forward and announced plans for HL3 and the only bits of leaked material the Internet has been able to find to keep the 'imminent HL3!' myth alive for all this time have been scattered .txt files and old domain registrations (it was rather funny when the Internet was totally convinced that the Pronunciation Channel project was related to HL3).
The project tracker leak in 2013 was the bigger piece of news; not only revealing HL3, but also giving an idea of the size of the development team at the time, which includes a lot of people who aren't Marc Laidlaw. Valve has other writers on stuff, by the by. It's a bummer to lose Laidlaw, but its nowhere near the end of the world. HL3 may never release, but it's not going to be because one man is gone.
Half-Life is a valuable enough IP that no doubt someone somewhere down the line is going to make something with it. I doubt that Valve sees that kind of effort as worthwhile, given the success of their non-game products and the risks currently involved with AAA publishing.
HL3 development probably spun it's wheels for a while - long enough to create a bunch of refuse that eventually found it's way to people looking for anything to latch onto - and then quietly died for any number of reasons.
Valve literally just hired Brad Muir (of Massive Chalice fame), who confirmed he's working on videogames, so he's doing something, even if its not HL3. And I doubt its making hats for DOTA, or just maintaining an already released game.
Probably, but as he's mentioned in the past, Valve isn't a "here's the script, let's go make the game from it" kind of company. (Very few game developers are.). They figure out what the games is going to be, create a story outline, they build off that, the story changes based on the development cycle. Bits get cut, changed or moved based on the design. Stuff will get thrown out wholesale before the game even makes alpha. Story will change to service game mechanics, new ideas that come along, or ebcause stuff became too difficult or unwieldy to add to the game.
Yes, this is pretty much how it goes with writing: you sit down and create a bunch of semi-related, semi-coherent plot points and background exposition, then you just shrug and walk away from the project and someone else just starts fitting the pieces together until you have the Great Gatsby.
I didn't mention anything about people walking away (in fact, I made no distinction from the writing and production parts of Valve in that description.) Also, I don't know about Great Gatsby, but this is literally how HL2 was made. (As described in Raising the Bar). Laidlaw even confirmed it during the kerfluffle with The Know back in July. The story creation with Videogames is an ongoing process, greatly affected by the games development. It's very rare that a writer will just pop out a complete story and that specific story lives all the way to the end of development. There's some especially crazy tales of how HL2's story changed shape over the games development.
Yeah, no doubt it's been an incredible long wait with no signs of ending anytime soon. Or at all, really.
As I've talked about before, it really seems like at some point they decided to not do more of the same (Episode 3) and rather try to make something that would have a big impact like HL1 or HL2; but trying to figure that out in this day in age is a goddamn Sisyphean task. I can't even remember when the last truly innovative FPS came out. Titanfall, maybe? But even that didn't light the world on fire. DNF may be an apt comparison.
I do get the sense that valve doesn't want to release another Half-Life unless they're sure they can "wow" everyone with some new technology or mode of storytelling.
Which would be fine and all if they hadn't ended Episode 2 on such a goddamn cliffhanger.
I do get the sense that valve doesn't want to release another Half-Life unless they're sure they can "wow" everyone with some new technology or mode of storytelling.
Which would be fine and all if they hadn't ended Episode 2 on such a goddamn cliffhanger.
This is basically how I feel. Valve doesn't "owe" anybody Half Life 3, but it feels pretty shitty that they spent three games doing their best to make the players form an emotional attachment to their characters only to end with that brick to the face at the end of Episode 2 and then not provide any closure for eight years and counting.
In a lot of ways I think Valve and Epic are in the same boat. They haven't released a game in their flagship franchise in almost a decade (the last Unreal game was UT3 in 2007) because they've been focusing on a side thing that eventually went on to be their primary income source (for Valve this was Steam, for Epic this was Unreal Engine licensing). If Steam had never been released (or had fizzled out) I'm sure we would have seen at least two more Half Life games by now.
Posts
Well, if we take what they have said about their design policy at face value, the reason to have that 10-man crew on would be to keep looking for the thing that's going to make the game worth developing. So basically looking at story concept, gameplay concepts, whatever pieces they can find that could eventually be slotted into a perfect whole. The long-term purpose of that team would be to (potentially) create a pitch strong enough to convince Valve that the game should be made.
Or not, if their source is someone who they believe works at Valve.
NSFW
Made me realize how much I really want to see the story continue. I miss the story happening around you thing.
I figure those that think they're assholes for reporting what they did will still think the same, but I appreciate them talking about it like this. As someone who watches a lot of their stuff, they're being pretty honest and open here, not defensive or trying to hide behind jokes.
edit: okay so the embed doesn't start when I wanted it to, so I guess skip to about 10m40-ish.
...my question is, what is that stuff under the cube? The clear panels could be like, flip pad things, but it looks like it's holding a heart shaped potion? Any idea what that's supposed to be?
SteamDB has typically been on the level. Though we know they've been working on HL3, this doesn't really tell us much about how far along it is, or much about the game period. Could all be old info.
Find/Replace "Harrison Ford" with "Gabe Newell"
Find/Replace "Star Wars" with "Half Life 3"
The file is definitely legit imo. It is also probably recent for it to find its way into a Dota 2 update that it wasn't already in before.
Not much solid info in the file, but here's all you can kinda deduce
-it has stuff for procedurally spawning groups of NPCs based on stuff like whether the player is nearby or not, what's visible to player, etc
-it has "quest citizens" who have different behaviors from other citizens. If a quest citizen is a "quest member" it's "neutral to combine" and doesn't take damage. My first thought is "neutral to combine" means this citizen won't open fire on combine on sight but might still do so if attacked by combine, which is probably a useful thing for a lot of types of quests? This is supposition (though I do have knowledge of HL's relationship systems, etc etc).
-quest citizen has "beckon player when player is close enough" functionality but I'm not sure if this is just inherited from npc_citizen and was present in hl2/episodic
-"prop_zipline" exists so that's a level-placeable zipline of some sort, which doesn't really tell anyone anything
-there are a number of quest-specific info entities and vars. Of particular interest is "height of marker beam", "radius of bottom disc", "trace up to place top of beam", etc. Sounds like vertical-beam quest markers in an open-ish world map.
-"Combine Pulse Ceiling Turret"? Were those a thing in earlier games?
-there's an npc_hunter_invincible which is interesting. I'd have thought you could just make a hunter invincible rather than making a new class called invincible hunter. So that's a thing
-generic pickup has a "prevents respawning" var and a "associated playercounter name" var which are both marked as "HL3 only". The respawning thing only really makes sense to me in the context of an open world game
JUST SPECULATIN'.
edit: some of this stuff could also have slipped in there from other projects/games, or it could be references to shit from years ago
Don't tease me like this valve...
We've gotten like a dozen of these leaks over the years; all it means is they were, at some point, working on it.
For a while I've been wondering how Valve could manage to make a linear FPS mind-blowing in the same way as HL1 and 2 in today's video gaming market.
This is kind of an obvious answer to that, in retrospect.
Also the Black Mesa guys have a patch planned for december
https://steamdb.info/blog/recent-package-name-additions/
Here are the HL3 entries.
Half Life 3 Developer Comp
Half Life 3 for Beta Testing
Half Life 3
http://forums.blackmesasource.com/index.php/Thread/28683-The-Crossfire-Update-Content-Update-0-2-0/
http://www.gameinformer.com/b/news/archive/2016/01/08/report-half-life-sole-writer-marc-laidlaw-has-departed-valve.aspx
I don't see why Laidlaw would leave if Half-Life 3 was in serious production.
EDIT: Okay, this article makes it sound like it's more of a age thing than a "there wasn't anything for me to work on" thing.
http://www.gamasutra.com/view/news/263256/HalfLife_writer_Marc_Laidlaw_leaves_Valve_after_18_years.php
its real. he retired. He also confirmed its real.
Rock Band DLC | GW:OttW - arrcd | WLD - Thortar
HL2 has a lot of shit that got cut out of the game or changed. (the borealis opening, the atmosphere processor, child laborers)
I mean that he's detailed character arcs for Alex and others to take, designed plenty of background material for various planned settings, and left behind an enormous Half-Life bible with all kinds of information for people to incorporate into the game.
Rock Band DLC | GW:OttW - arrcd | WLD - Thortar
I wrote him a sad fanmail
Really thought they would release a HL title and maybe an Orange Box type product in the first few years of the new console gen.
Time to wake up and smell the ashes.
Note that November said, "In the first few years,"
Also note that HL2 was well slated as an incoming product well before it's release.
I don't know what more needs to be said. The lead writer who was a fundamental part of making the magic happen is now gone, Valve has never actually come forward and announced plans for HL3 and the only bits of leaked material the Internet has been able to find to keep the 'imminent HL3!' myth alive for all this time have been scattered .txt files and old domain registrations (it was rather funny when the Internet was totally convinced that the Pronunciation Channel project was related to HL3).
Half-Life is a valuable enough IP that no doubt someone somewhere down the line is going to make something with it. I doubt that Valve sees that kind of effort as worthwhile, given the success of their non-game products and the risks currently involved with AAA publishing.
HL3 development probably spun it's wheels for a while - long enough to create a bunch of refuse that eventually found it's way to people looking for anything to latch onto - and then quietly died for any number of reasons.
Yes, this is pretty much how it goes with writing: you sit down and create a bunch of semi-related, semi-coherent plot points and background exposition, then you just shrug and walk away from the project and someone else just starts fitting the pieces together until you have the Great Gatsby.
Fair enough, I totally see how my comment was super misleading by mentioning a timeframe. My bad. My point there was supposed to be that its not really Valve's style to plan content around platform availability. They'll release a game when its done. I mean, hell, they released an Xbox version of HL2 a only a year before the 360 came out.
Yeah, but given it wasn't finished until well after Steam's launch, it doesn't strike me as them waiting for Steam to be out before they released it, which is my point.
The project tracker leak in 2013 was the bigger piece of news; not only revealing HL3, but also giving an idea of the size of the development team at the time, which includes a lot of people who aren't Marc Laidlaw. Valve has other writers on stuff, by the by. It's a bummer to lose Laidlaw, but its nowhere near the end of the world. HL3 may never release, but it's not going to be because one man is gone.
Valve literally just hired Brad Muir (of Massive Chalice fame), who confirmed he's working on videogames, so he's doing something, even if its not HL3. And I doubt its making hats for DOTA, or just maintaining an already released game.
I didn't mention anything about people walking away (in fact, I made no distinction from the writing and production parts of Valve in that description.) Also, I don't know about Great Gatsby, but this is literally how HL2 was made. (As described in Raising the Bar). Laidlaw even confirmed it during the kerfluffle with The Know back in July. The story creation with Videogames is an ongoing process, greatly affected by the games development. It's very rare that a writer will just pop out a complete story and that specific story lives all the way to the end of development. There's some especially crazy tales of how HL2's story changed shape over the games development.
Might be funny to go hunting for a "things that have happened since Duke Nukem Forever was announced" style list for HL3/Ep3 at some point.
Not like we have to rush, of course...
As I've talked about before, it really seems like at some point they decided to not do more of the same (Episode 3) and rather try to make something that would have a big impact like HL1 or HL2; but trying to figure that out in this day in age is a goddamn Sisyphean task. I can't even remember when the last truly innovative FPS came out. Titanfall, maybe? But even that didn't light the world on fire. DNF may be an apt comparison.
Which would be fine and all if they hadn't ended Episode 2 on such a goddamn cliffhanger.
This is basically how I feel. Valve doesn't "owe" anybody Half Life 3, but it feels pretty shitty that they spent three games doing their best to make the players form an emotional attachment to their characters only to end with that brick to the face at the end of Episode 2 and then not provide any closure for eight years and counting.
In a lot of ways I think Valve and Epic are in the same boat. They haven't released a game in their flagship franchise in almost a decade (the last Unreal game was UT3 in 2007) because they've been focusing on a side thing that eventually went on to be their primary income source (for Valve this was Steam, for Epic this was Unreal Engine licensing). If Steam had never been released (or had fizzled out) I'm sure we would have seen at least two more Half Life games by now.