I must’ve missed the genius. Jai just seems to add a bit of modernized syntax and features over the same old C++ semantics. And C++ is evolving to add those same modern features in any case. It’s peculiarly lacking in ambition.
The SOA/AOS split is particularly cool IMO. Tuning for performance should be as easy as flipping a switch!
Trying to solve a lot of C++'s build-time headaches is also cool. From my experience, any sufficiently complex C++ build environment is indistinguishable from punching yourself in the face. A lot of major frameworks, like Qt, have a code generator to do meta-programming that would otherwise be too frustrating to do with the language itself, because templates are the only meta-programming option and they are a bit lacking.
Relatedly, C++ templates are a nightmare once you’re trying to do anything more complex than the basics, and STL is just a mess. (do not get me started on how much I want to throw boost out the window.) There’s a lot of features like that which feel like they exist for the library developers rather than the software developers themselves.
So trying to unify meta-programming with the build time process is a really cool idea to me. It’s not the only language to try to do so, though.
The rest is kind of… I don’t know what he’s thinking, really.
1 Like
I will buy this to help jon blow buy another tesla
4 Likes
this is consistent with the “Jonathan blow isn’t a genius, he’s just a high-profile anal retentive” theory. I really am curious to see how this game turned out, though. And I wish I hadn’t bounced off of talos quite so thoroughly.
if I like this game a lot I’ll probably get talos. I almost got it during the steam winter capitalism explosion, but I didn’t because the “you’ll never play this” argument that works for me with any other game I think about buying worked with this one too.
1 Like
I think Talos starts really poorly. Part of me thinks this wouldn’t be the case if the game weren’t so damn long. It took me hours and hours to get to puzzles that I couldn’t just immediately blow through (maybe with the exception of one or two puzzles), and by that point I was kind of done with the whole thing.
I feel like I should go back to it at some point. The meta-puzzles were pretty cool and I think maybe the main puzzles get hard eventually? Any puzzles gurus have comments on Talos’s eventual difficulty?
EDIT: Hmm, and now that I think more about it, I think puzzle elements weren’t explored deeply enough. Right when the puzzles would start to have a glimmer of something complicated, they would introduce a new mechanic and they’d get easy again. I assume at some point they stop adding new ones and actually explore the ones they have?
The Witness gave me a eureka moment years ago when looking at screenshots (I think before the mazes were even shown). It struck me what they might be up to and I could not get the game out of my head, as that would totally be something worth making a game about, no matter how long it took and how much it cost. It’s probably unfair to go into a game with that kind of expectation, but it’s going to be the most uncanny thing ever if that’s what they actually did.
Glancing at some reviews it looks like the main criticism is that it’s too alienating/demanding/intense. It’s kind of exciting to think it might really be as uncompromising as they make it sound, but then again, some reviewers probably haven’t played a puzzle game since Braid or Portal 2.
The recorder is the last mechanic introduced, after that, the puzzles pretty quickly resume a higher difficulty. Nothing was really so hard that it could challenge a jaded player of Myst-alikes and IF games but that’s why difficulty of puzzles probably isn’t a good rubric for evaluating a first person puzzle game. My final playtime was 23 hours for all the endings plus messing around in some of the prototypes and trying to find easter eggs. Actual playtime was probably around 16 hours give or take.
I found that there was generally only one or two ‘introducing a game mechanic’ puzzles per mechanic before returning to the standard slowly-escalating difficulty.
portal is easily one of the worst things to happen to puzzle games
this is also a particularly game reviewer type of criticism, because those people have to plow through a game as quickly and efficiently as possible in order to get their Hot Take out onto the interwebs in time for the lifting of the holy embargo. the more a game plays itself, the better. I, on the other hand, don’t have to give a shit about any of that.
I’m curious to see how the Gamer reaction goes for similar reasons. Even players who aren’t actually Games Journalists running up against embargoes have internalized some of the same ideas about how “long” games should be, and how easy/hard they should be to “complete.” Old-style, puzzle-focused adventure games of the “you can and will spend weeks of evenings trying to solve this” are probably not any more suited to the current Games Market than they were after the novelty of Myst wore off decades ago.
[quote=“Gate88, post:47, topic:761”]
I feel like I should go back to it at some point. The meta-puzzles were pretty cool and I think maybe the main puzzles get hard eventually? Any puzzles gurus have comments on Talos’s eventual difficulty?[/quote]
One word I keep thinking of to describe Talos is ‘flat’. Although it gradually ramps up and I finally got stuck on a couple puzzles back-to-back in the third/final hub area, several of the ideas become stretched thin or repeated in later puzzles without introducing a new wrinkle (the jammer only puzzles are a good example).
The expansion pack campaign is supposed to be much harder, but if it doesn’t introduce new objects (?), I’m not sure that just upping the difficulty would be that interesting, as some of the laser puzzles already reach a breaking point in how far they can reasonably push things within the limits of your perspective.
I think if the puzzle rooms weren’t so spatially (and visually) flat, with more imaginative use of 3D space, that would have helped. The meta-puzzles do explore that a little more, and figuring out one of the early meta-puzzles with lasers was still probably my favorite moment in the game, but then a lot of the later meta-puzzles are the same idea somewhat rearranged.
Hmm, interesting. I think Myst/IFs have a different kind of difficulty than Talos has, and I wouldn’t really compare the two. I find most of the difficulty in Myst comes from not knowing what the goals are. Myst takes a lot of experimentation and careful observation to solve the puzzles, because you don’t know what the puzzles are to begin with and what information might be relevant to those puzzles.
In the Talos Principle, it’s usually pretty clear what is and isn’t part of the puzzle. You’re given all the information up front, and it’s up to you to figure out the implications of the rules and mechanics to make the optimal moves. The difficulty comes from logically deriving the end solution from the given axioms. In Myst, once you know all the information and what information is relevant, the puzzles are sort of trivial. Getting to that point is the difficult part. (I say this not having fully beaten either game, so maybe I’m missing something from those games).
OK but it was actually about the atomic bomb so you’re wrong
8 Likes
I guess I must be pretty bad because I found Talos difficult past the first major area. The meta puzzles were absolutely out of my league.
Easy puzzles rewarded more narrative more easily and hard puzzles were rewarding on their own so there was alwayd an interesting cycle going on?
It also helps that the two major tools (jammer and connector) were fundamentally alien and interesting. The big box not so much
Anyway that piss jug thing is amazing
This game has hideous trees.
3 Likes
Strongly disagree.
3 Likes
Secretly turning your lover into an atomic bomb is a pretty gross way to behave though.
12 Likes
OK, that’s pretty cool.
[quote]Trying to solve a lot of C++'s build-time headaches is also cool. From my experience, any sufficiently complex C++ build environment is indistinguishable from punching yourself in the face. […]
So trying to unify meta-programming with the build time process is a really cool idea to me. It’s not the only language to try to do so, though.[/quote]
Check out Google’s Bazel and GN+Ninja build systems, they’re clean and idempotent build systems that support C++.
The problem with trying to integrate building into the language is that as soon as you have a dependency in another language, you’re back in the same world of pain. Secondly, the primary philosophy of Bazel and even more so GN is to make the build language as weak, primitive and declarative as possible. That prevents the punching-in-the-face complexity from accumulating and allows making inferences for reusing build results across time and multiple users in a server farm. So using the full general-purpose language as a build system is the complete opposite of the trend in modern build systems. In general Jai shows an almost willful ignorance of the state-of-the-art in the areas it’s trying to revolutionize.
well I never