sighinggoldfish
sighinggoldfish on Hacker News
And at least the weather has picked up a bit right now haha! Time to get off the web and take advantage of it as Denmark gets so nice once the weather is good.
Sure, but I mean, at this point, I'll even take a guarantee that I can just run around on a map without spawns just so the world art is preserved. The initiative is pretty broad about what it means to be playable for this reason. Actually, that's the whole point of this initiative is to establish such a bare minimum. And if the game is truly a service, and not a good, than this wouldn't even apply! Companies would just have to be explicit about the service nature of the product which they are not unless you pay a subscription like in the WoW case, which this doesn't affect. So you could still do all this tight coupling that is apparently good for development and ship games just fine if players are properly informed on the service nature (which means some kind of expectation for how long the service will last). I agree that a testbed map isn't the same as playing offline, but I don't agree that it's preferred, or even natural, that development require that offline play be so arduous to accomplish. Especially when something as complex as WoW or PSN matchmaking + a game's specific netcode (Demons Souls PS3 emulation case) can be reverse engineered without any support from devs. Most arguments along the lines of "it's just too technically difficult" do not past the smell test to me.
> You’re moving the goalposts here. You asked about cloud tech initially. The topic of “have game budgets and teams gotten out of control” is a totally different one.
I feel like I pretty succinctly explained that this tech did not provide a huge boost in productivity in development nor did it allow for substantially more scalability of players. Not sure how that's moving the goalpost. You posted a tweet highlighting 4 people allowed for 5x the player capacity in a week. I don't know how many multiplayer engineers worked on WoW pre initial launch but considering the team was smaller than Arrowhead, I doubt it's significantly more than 4 and their concurrent player numbers around launch weren't significantly different (certainly not 5x). I get that Blizzard probably predicted bigger launch numbers than Arrowhead so there's at least some merit to being able to scale up fast.
But overall, I think, like most modern software "advances", we've been duped on their actual performance and efficiency gains, while most gains have actually come from better hardware and internet connections (hardware again). And end users really aren't feeling much of an improvement. I think more has been lost to users and players actually. Less mods, less weird community servers with new game modes, etc. And as mentioned earlier, this actually helped incumbent devs find new revenue streams! The new supposed features of this tech is more to serve all the meta stuff like progression tracking and matchmaking which requires greater centralization of online services. Ye in many cases, like Helldivers 2, the actual gameplay is P2P. So the only thing players get out of the this new "tech" is the thing designed to keep them addicted which is stuff like lootboxes and Skinner box progression mechanics. (I acknowledge my last points here on meta mechanics as goalpost moving, I do view this as part of a much larger issue)
> Systems like LaunchDarkly require active subscriptions and don’t allow for redistribution and are often deeply embedded in applications.
I mean, this is just a bad state. I view this initiative as fighting shit like this. It's bad for devs too. What happens if LaunchDarkly or some other platform holder goes down or cuts off access to a dev for some perceived breach of conduct or license agreement? That's so damaging to a developer, and if they aren't established, existentially threatening. We shouldn't encourage this kind of coupling. As for the non distributable licenses of common distributable like Java and SQL, I'm not terribly worried the economics won't catch up to make it feasible to just use distributable license. We've already seen lots of Java technologies relicensed to GPL 2 over a decade ago. It's all possible without collapsing the economics of it all.
> games aren’t developed the same way as they were “prior to this recent problem”, player behaviour and expectations have changed. You can’t just go back to the way things were here any more than you can with social media, banking, email, movie rentals, etc.
We certainly can go back, or better yet, do better than we used to! I really don't think there's a good technological arguments for why we dev this way now. It's mostly political and I think we should resist this type of excusing of rent seeking style software as "tech advancement". All those areas you mentioned are the way they are not because of technological necessity but because of political and cultural forces. We can style them in a way that's more compatible with a more free and expressive user base, and ultimately, a freer market. The more, but still limited, openness of the pass certainly contributed greatly to the grown of games!
I'm curious of your examples actually as I'd like to learn more about the Danish process specifically.
As for ECIs, the wikipedia page does show successes. Of course I have my doubts this will even reach the necessary 1 million signatures (and honestly needs more cause many will be invalidated), but if it does reach 1 million, it does have a shot of doing something!
Both should be available cause yes, you need to test like you said. But if I'm setting up skinning on a character or cloth simulation on a cape, I want my turn around time as fast as possible from DCC to engine. Maybe I just want to validate my PBR materials with a specific level's lighting rig and baked GI/reflection probes. Those specific examples are client side only, so I absolutely want that to work on an offline build. The longer a roundtrip to the engine takes for content (including the time it takes to start the game up), the harder it is to iterate and thus quality also falls. Forcing online always for devs at the very least, makes booting the game slower (or just fail) and you need to reboot the game constantly (even with hot reload). That cost adds up quick across many devs and many years. You need both lest you drive your artists and TAs insane. My last company hemorrhaged artists to literally go work next door on very similar styled art content from one live service game to another just because our tools were so exhausting to them to roundtrip, and online wasn't even forced!
>https://x.com/Pilestedt/status/1760077808146014340 - helldivers backend team was 4 people. _That's_ the point of this newfangled tech.
WoW took 4-5 years to develop with "The original World of Warcraft was created by a team of 40 people, which eventually doubled in size as the launch drew close." https://www.polygon.com/2020/1/20/21070494/world-of-warcraft...
Where as as Helldivers 2 has about 100 devs at the end of development taking 8 years according to their CEO. So not seeing the wins there. In terms of concurrent players around launch, WoW had around 0.5 million within months and Helldivers 2 seemed to have peaked at 0.75 million across PC and PS5. 19 years apart and not really seeing that 5x in practice... None of this suggests to me that this "tech" is a useful as many think.
> Handwaving away this problem doesn't make it just og away.
I mean, we have tons of precedent of these middleware licenses working just fine when redistributed as binaries to users' devices, so how is this handwavy? Like yes, there's an issue with already shipped games that make use of such licenses, but after such a law passes, the licenses will have to adapt to licenses similar to what already exist. And prior to this recent problem in games, middleware licenses worked just fine too and often shipped with server binaries. These licenses are not the norm as explained by all the middleware that does get packaged with almost every game today...
I'm really skeptical of modern online infra. I've worked on two MMOs as a Tech Artist and the way artists, designers, etc worked on the game was with a local only build! One of these companies actually barred WAN access from our dev machines so LAN only also worked for internal playtests! Trends to make dev builds online only are quite awful for developers in the trenches and are more about exec control than any usefulness added by such requirements.
Complex server infrastructure that you mention is also not needed if you ask me, at least not to the point to make it too hard to widely ship. I'd be fine if someone still had to stand up a centralized matchmaking server still, just allow any player to do it.
As an example, compare Helldivers 2 and WoW. WoW is vastly more complex but Helldivers 2 uses more modern cloud "tech" that didn't exist 20 years ago. Both had issues keeping up with player demand on launch. So what's the point of this newfangled tech? Well, it's devs convincing players they're getting a better service or good (don't know which!) and that this complexity based lock in is worth it. And it's also cloud providers like AWS and Azure convincing devs to create architectures like this which lock devs into these platform holders. I'm not convinced were seeing real improvements on the dev or player side... Scamming turtles all the way down if you ask me.
It also turns out a lot of this stuff is able to reverse engineered. I'm right now playing original Demons Souls with full multiplayer connectivity (including matchmaking from a central private server) which required PSN support of the PS3 days. WoW private servers also required a great deal of reverse engineering. If online randos can do it...
Plus WoW private servers literally revealed a new revenue stream to Blizzard in the form of WoW classic: "After a month or so of large scale protests, Blizzard invited the Nostalrius team to the Blizzard HQ to present the case for Vanilla. An eighty-page "post-mortem" document describing the development of Nostalrius, the problems that happened and some marketing strategies was presented to Blizzard, and after some time, released on the Nostalrius forums" https://en.wikipedia.org/wiki/World_of_Warcraft_Classic
Others have also mentioned middleware licenses as a problem. Those licenses would adapt just fine. A single game will ship with multiple binaries of RAD Tools, Havok, SpeedTree, etc. So now a player has potentially hundreds of copies of different versions of all these binaries and libraries. This is a completely solvable problem that won't stifle development or even require leaking of trade "secrets".
Also, this is relatively new problem in the industry. The market grew just fine allowing for players to use their goods more freely. That any such regulation would hurt the industry is dubious at best, corporate speak at worst. In fact, I'd argue that enabling player freedom an creativity over their purchased goods has helped the industry grow to where it is today! Counter Strike was a mod, TF2 was a mod, PUBG was a mod (Fortnite wouldn't exist with out it), DOTA was a mod! Minecraft in part exploded cause of mods. A healthy and innovative free market demands player and user freedom!
" The European Citizens' Initiative (ECI) is a European Union (EU) mechanism aimed at increasing direct democracy by enabling "EU citizens to participate directly in the development of EU policies",[1] introduced with the Treaty of Lisbon in 2007. This popular initiative enables one million citizens of the European Union,[2] with a minimum number of nationals from at least seven member states, to call directly on the European Commission to propose a legal act (notably a Directive or Regulation) in an area where the member states have conferred powers onto the EU level. This right to request the commission to initiate a legislative proposal puts citizens on the same footing as the European Parliament and the European Council "