MudMan

I don’t know, some of these guys have acccess to a LOT of code, and even more debate about what those good codebases entail.

I think the other issue is more relevant. Even 128K tokens is not enough for something really big, and the memory and processing costs for that do skyrocket. People are trying to work around it with draft models and summarization models, so they try to pick out the relevant parts of a codebase in one pass and then base their code generation just on that, and… I don’t think that’s going to work reliably at scale. The more chances you give a language model to lose their goddamn mind and start making crap up unsupervised the more work it’s going to be to take what they spit out and shape it into something reasonable.

I think the effects of it are… a bit more nuanced than that, perhaps?

I can definitely tell there are places where I’m plugging knowledge gaps fast. I just didn’t know how to do a thing, I did it AI-assisted once or twice and I don’t need to be AI assisted anymore because I understood how it works now. Cool, that. And I wouldn’t have learned it from traditional sources because asking in public support areas would have led to being told I suck and should read the documentation and/or to a 10 video series on Youtube where you can watch some guy type for seven hours.

But there are also places where AI assistance is never going to fill the blanks for me, you know? Larger trends, good habits, technical details or best practices that just aren’t going to come up from keeping a smart autocorrect that can explain why something was wrong.

Honestly, in those spaces the biggest barrier is still what it was: I don’t necessarily want to “progress” on those areas because I don’t need it and it’s not my job. I can automate a couple things I didn’t know how to automate before, and that’s alright. For the rest, I will probably live with the software someone else has made when it exists.

The problem is hubris, right? I know what I don’t know and which parts I care to learn. That’s fine. Coding assistant LLMs are a valid tool for someone like that to slightly expand their reach and I presume there’s a lot of people like that. It’s the random entrepeneurs who have been sold by big corpos that they don’t need a real programmer to build their billion-dollar app anymore that are going to crash and burn and may take some of the software industry down with them.

No, you were fishing for a bug to dismiss the point the guy was making, but the issue is deeper than that.

I play a bunch of fighting games and I can tell you I can feel combos not working going from PC to PlayStation or from a DS4 to a DS5 or from my PC gaming monitor to my TV.

That didn’t use to be the case. Timing could be different in consoles or arcades because those were entirely different games, but if you could do a thing on one TV with one controller you could do it on any TV with any controller on that game. That’s not the case anymore.

And yet “react to a visual cue in something the enemy does and press the button in a small frame window” is a much, much more frequent core mechanic now than it was then.

It is super important, but it’s not a timing problem, it’s a knowledge check.

The thing you want to know is whether you can buffer a button during your recovery and get it off before the other guy. If you can, the timing is often very loose, you mash on that thing and you probably get your punish out frame 1.

And if the game is good you don’t learn it by spending a ton of time with a wiki or even with the training mode, you learn it by playing the game and looking at the animations.

If you are a bad game, like the first few Tekkens (yeah, I said it), then you won’t get that from the animation, but you can still learn it by trial and error. And crucially, once you learn it it’s always the same. In most decent games with consistent base mechanics, anyway.

So no, you shouldn’t have to learn whether your jab is four or five frames. That’s how the game is put together, but it should be good enough at communicating how fast your button came out that you can intuit when it’s safe or effective to put it out after blocking. At least after you try it once.

There are a ton of fighting games that are still grandfathered into the notion of putting complexity in this part of the design. You know, the ones where your fastest attack is different per character so you need to know this particular guy’s fastest opener is a crouching medium kick, but only when you’re close enough or whatever. I’d argue those games are less elegant without adding anything to the skill ceiling when compared to newer games like my previous DBFZ example, where everybody’s jab is probably the same speed and the basic flowchart of what to do after you block an attack doesn’t require a textbook and complementary materials.

But the point is you built a timing mechanic in a turn based RPG that is dependent on latency in a world where you are adding arbitrary amounts of latency in the display, the rendering, the controller polling and the platform itself.

That’s a bad fit between design and hardware. I’d argue that’s always the case and modern gaming SUCKS for timing-based mechanics compared to old ride-the-scanline CRT gaming. And yet they are extremely popular and people seem to assume timing challenges are the only way to make fun mechanics.

You can trivialize the parrying in Clair Obscur with the difficulty levels (and automate the offensive timing of it altogether).

But it sucks that you can’t disable it and have to stare at QTE prompts over the beautiful combat animation.

If you need to ask that question the mechanic is fundamentally flawed.

I call bullshit on fighting games and frame data.

I remember ages ago when SF4 happened I was friends with a pretty solid tournament competitor and talking about this he mentioned that frame counting is for nerds because the only thing you need to know is if you can push a button or not and that happens from intuition, which is true. At least it’s true in a good game that has good animation. I’ve always had a kick of beating sweaties with their “this is minus three” obsession by having solid fundamentals. Some of them got pretty annoyed.

And to this day I will claim that Dragon Ball FighterZ is the best fighting game of the last decade specifically because a) every basic combo route is built out of repeatable, simplistic, easy to remember chunks, and b) you can mash the CRAP out of 90% of that game’s links and they work perfectly fine unless you’re trying to do a rejump or time an assist extension.

I will die on this hill, except I won’t because I’m right. I will survive on this hill. Make a little cottage on it and play fighting games inside it.

I will take it one step further:

Timing is not the only skill that is fun and developers should acknowledge this.

The reason we talk about parrying this way is it’s everywhere now and people have decided it’s what’s needed to make turn based games compelling, but it isn’t.

The reason they think that is that people have decided that timing a thing is the only exploit-free, skill-driven action you can have at the core of gameplay.

But it isn’t.

I find it a bit lazy to default to this approach on everything and it’s certainly on everything. Timing is a big part of gaming, particularly in real time action stuff, but there are other tools in the toolbox.

Here’s an observation: the reason people keep trying to make metroidvanias and being worse than SotN and its handheld sequels is that modern designers can’t get over adapting Soulslike combat where the Igavanias were more concerned with giving you a ton of options. Timing is there, but there’s a ridiculous amount of self-expression through tool selection. You can go for a tanky build, you can break DPS and movement a hundred different ways. Getting good at those games makes them look like a broken mess, but it’s self-expressive and fun.

Timing a marker to a yellow bar in Clair Obscur or trying to guess when the ridiculous fifteen second windup animation of an enemy is going to trigger a five frame active window is not self-expressive and fun. It’s a QTE.

Clair Obscur’s combat has other issues with balance beyond that, and you can certainly spec to trivialize the parry even without changing the difficulty level, but the annoying part is the focus everybody puts on the timing minigame versus the actual (pretty solid) turn based game design running underneath.

Fully agreed. Everybody is betting it’ll get there eventually and trying to jockey for position being ahead of the pack, but at the moment there isn’t any guarantee that it’ll get to where the corpos are assuming it already is.

Which is not the same as not having better autocomplete/spellcheck/“hey, how do I format this specific thing” tools.

Yeah, the AI corpos are putting a lot of effort into parsing big contexts right now. I suspect because they think (probably correctly) that coding is one of the few areas where they could get paid if their AIs didn’t have the memory of a goldfish.

And absolutely agreed that making sure the FOSS alternatives keep pace is going to be important. I’m less concerned about hating the entire concept than I am about making sure they don’t figure out a way to keep every marginally useful application behind a corporate ecosystem walled garden exclusively.

We’ve been relatively lucky in that the combination of PR brownie points and general crappiness of the commercial products has kept an incentive to provide a degree of access, but I have zero question that the moment one of these things actually makes money they’ll enshittify the freely available alternatives they control and clamp down as much as possible.

Sorta kinda. It depends on where you put that line. I think because online drama is fun when we got to the “vibe coding” name we moved to the assumption that all AI assistance is vibe coding, but in practice there’s the percentage of what you do that you know how to do, the percentage you vibe code because you can’t figure it out yourself off the top of your head and the percentage you just can’t do without researching because the LLM can’t do it effectively or the stuff it can do is too crappy to use as part of something else.

I think if the assumption is you should just “git gud” and not take advantage of that grey zone where you can sooort of figure it out by asking an AI instead of going down a Google rabbit hole then the performative AI hate is setting itself up for defeat, because there’s a whole bunch of skill ranges where that is actually helpful for some stuff.

If you want to deny that there’s a difference between that and just making code soup by asking a language model to build you entire pieces of software… well, then you’re going to be obviously wrong and a bunch of AI bros are going to point at the obvious way you’re wrong and use that to pretend you’re wrong about the whole thing.

This is basic online disinformation playbook stuff and I may suck at coding, but I know a thing or two about that. People with progressive ideas should get good at beating those one of these days, because that’s a bad outcome.

It’s pretty random in terms of what is or isn’t doable.

For me it’s a big performance booster because I genuinely suck at coding and don’t do too much complex stuff. As a “clean up my syntax” and a “what am I missing here” tool it helps, or at least helps in figuring out what I’m doing wrong so I can look in the right place for the correct answer on something that seemed inscrutable at a glance. I certainly can do some things with a local LLM I couldn’t do without one (or at least without getting berated by some online dick who doesn’t think he has time to give you an answer but sure has time to set you on a path towards self-discovery).

How much of a benefit it is for a professional I couldn’t tell. I mean, definitely not a replacement. Maybe helping read something old or poorly commented fast? Redundant tasks on very commonplace mainstream languages and tasks?

I don’t think it’s useless, but if you ask it to do something by itself you can’t trust that it’ll work without singificant additional effort.

Sorry, but no. My current path is

C:\Users\<user>\Documents

I literally just navigated to my documents folder in a explorer window and copy pasted the path here. I swear I’m not making this up.

I’m pretty sure that these days syncing your documents folder with OneDrive is two separate opt-ins: one to log in to OneDrive at all and one to select whether to sync your libraries.

I am not going to set up a VM just to check this, but I have multiple Windows machines in operation on Win10 and Win11 at home right now and none of them are syncing my libraries. That ranges from five year old Win10 installs to Win11 installs as recent as a couple of months ago.

For the record, I don’t think you’re crazy either. There’s definitely something baked into the install we’re doing differently or some version difference or whatever. It’s surprisingly hard to suss this out at this point, since there’s a fairly complex set of could-backed choices, first time setup choices and maybe even regional changes, I’m not sure. The one interesting, kinda shocking takeaway is how differently our machines can be set up based on probably some checkmark we each set up differently once ages ago or whatever the hell this is.

See, then you fell for a dark pattern, because I refused to use it on first boot and there is no One Drive folder in this brand new computer I’m currently using at all. It isn’t the standard My Documents folder, it doesn’t have a folder at all and the application icon isn’t on my system tray.

That’s why I was asking about the Win10 install being different, but I’ve installed Win 11 twice this year, once in a computer for personal use that currently doesn’t have any One Drive folder at all and one for work where One Drive is logged in to a work account (along with Office 365) and it only syncs that work folder, not the personal folders, photos and whatnot. You can absolutely have a Windows (11, anyway) install with no One Drive synced to anything at all, or even running.

It sucks that Windows designs its install process as a dark pattern-ladden attempt to get you to sign up for crap, but you can reject all of it. Maybe I do it enough I have the habit and I underestimate how hard it is to choose what you actually want. I guess that’s the equivalent to having a working Windows 98 key memorized in the early 2000s.

No, you’re thinking about the Microsoft Account login (which still has workarounds but whatever). OneDrive needs to have its own separate login, in case you, like me, have a separate account for work or need to have multiple One Drive accounts or if you have paid One Drive, 365 and whatnot.

So you can absolutely log in to Windows with a MS account and log in to One Drive with your work account… or not log in at all and just not have it running, which is what I do.

I have installed Win11 on a new computer build this year. I promise I’m looking at my system tray and there is zero One Drive icons on it. No One Drive folder in my Windows file explorer, either.

I’m confused. Are you logged in to OneDrive?

I mean, don’t get me wrong, I loathe OneDrive. It’s the flakiest, most unreliable piece of software in MS’s current end-user-focused stable…

…but it still needs you to log in. If you log out from OneDrive it does nothing. It’s a separate login from the Windows login, too, if you’ve used one of those to install Windows.

I genuinely haven’t used Win10 in enough time I don’t recall if it gave you more notifications to re-enable it, but after refusing to log in and taking OneDrive out of my startup tasks I don’t think it’s come up again on any of my Windows devices.

Man, people really overstate the barrier to entry to the terminal. Windows troubleshooting is full of command line stuff as well.

It’s not the terminal, it’s the underlying issues. Having more GUI options to set certain things is nice, but the reality of it is that if an option isn’t customizable to the point of needing quick GUI access it should just never break, not be configurable or at least not need any manual configuration at any point. The reason nobody goes “oh, but Windows command line is so annoying” is that if you are digging in there something has gone very wrong or you’re trying to do something Windows doesn’t want you to do.

The big difference is that the OS not wanting you to do things you can do is a bug for people in this type of online community while for normies it’s a feature.

Well, I think there’s a nuance between the notion of people hosting files and sharing them for piracy purposes, which is technically no different than hosting and distributing any other file, and a platform for digital distribution of media.

You could argue that peer-to-peer services with built-in search, like Kazaa or eMule came a lot closer and that’s defensible, but the birth of modern platforms is less on the tech to host and share the files and more on the ability to do DRM and sell access tokens digitally. Modern digital distribution is less an iteration on piracy software and more a response to it to provide something that could compete on convenience while being monetized, having DRM and, yeah, cutting a lot of people out of the money loop.

And on all those counts… yeah, it was Steam. Valve did it first and did it effectively while music labels and movie studios were still hoping lawyers would fix things for them.

You could download games from multiple publishers years before Valve did it, too. Doesn’t mean Valve didn’t come up with the first, largest digital distribution platform for games that was then the template for every first party (and most of the digital media distribution in other media industries).

Best I can do for you is let you have that and agree that piracy invented it and Valve monetized it, which is actually worse?