> none of this is going to improve people's lives.
I have some old borderline senile relatives writting apps (asking LLMs to write for it them) for their own personal use. Stuff they surely haven't done on their own (or had the energy to do). Their extent of programming background - shitty VBScript macros for excel.
It also helps people to pick up programming and helps with the initial push of getting started. Getting over the initial hump, getting something on the screen so to speak.
Most things people want from their computers are simple shit that LLMs usually manage quite well.
Good question whether or not this (outsourcing their thinking) actually just accelerates their senility or not.
As someone who likes to solve hard or interesting technical problems, I've long before LLMs often been disappointed that most of the time what people want from programmers is simple stupid shit (ie. stuff i dont find interesting to work on).
>OS manufacturer can’t be bothered to interact with their own UI libraries to build native UIs something has gone horribly wrong.
I honestly think that has way less to do with Microsoft, more of a representation of "software engineering" practices these days.
For example, Gnome shell has bunch of javascript in it, GTK has layout and styling defined in some flavour of CSS, etc.
I'm of opinion if you start writing OS userland in either javascript or python (or both),
you should be fired on the spot, but I don't make the shots.
Most technical decisions aren't really driven by what makes a better end-user experience or a better product, it's mostly defined by convenience and familiarity of substandard software developers - with mostly and primarily web-slop background.
Cosmic (from the PopOS folks) is getting rid of the crappy javascript from GNOME Shell. And the CSS in GTK+ themes is just for the sake of syntactic convenience.
Cosmic is quite nice. There's some polishing left to do, but it's already pretty solid. The app store is a bit of a turd, but I bet that's just because it's by nature connected to the internet. More could surely be done with caching and pre-loading, but not sure if I want my computer to pre-load app store content all the time just in case I open it.
Compared to Windows it's of course absolutely unreal.
But the difference is that none of the CSS or Javascript usage in gnome is tied to a webview. They are all binding in some way to GTK and much simpler rendering routines.
> I'm of opinion if you start writing OS userland in either javascript or python (or both), you should be fired on the spot, but I don't make the shots.
KDE Plasma, which is in my opinion the most advanced desktop environment is written in Qt QML which is JavaScript. There are advantages to that over C++, namely your session won't simply crash.
(While you can use some JavaScript from QML, the application still have a C++ core. QML applications can still crash. There is no DOM with QML, no browser overhead)
QML is absolutely not JavaScript. It's a markup language to describe user interfaces, spiced with JS for certain interactions. All heavy lifting behind the scenes is done in C++ - the QML runtime as well as the application logic and data models.
The software industry has always had more juniors than seniors so this issue of juniors calling the shots is not a new one but it does feel like it's been getting worse and worse... Now it's basically AI slop vibe coders calling the shots about coding best-practices.
When I read comments like this, I honestly think that people are only complaining about this because the "bad people" are doing this (in this case Microsoft/Gnome Team).
Neglecting the fact that almost everyone else is doing similar things.
> For example, Gnome shell has bunch of javascript in it, GTK has layout and styling defined in some flavour of CSS, etc.
What GTK is doing isn't really any different than how many UI framework work and have done so for quite a while now.
Almost every desktop UI toolkit/library/framework in the past 15-20 years has the following:
- Markup interface for defining the layout. If they don't have that they have a declarative way of defining the UI.
- Some sort of bindings for popular scripting language that hook into native code.
- Some of styling language that isn't that different from CSS.
This has been the norm for quite some time now. It works reasonably well.
Futhermore there isn't much difference between what desktop developers are doing and what web developers are doing.
> I'm of opinion if you start writing OS userland in either javascript or python (or both), you should be fired on the spot, but I don't make the shots.
Why? I find Gnome works really well on Linux. I have a pretty nice desktop environment after adding two extensions (Dash To Dock and App Indicators). Gnome runs well on relatively ancient hardware I own (2011 Dell E6410) with a garbage GPU (it isn't OpenGL 3.3 compliant). It actually performs a lot better than some other DEs that are 100% native.
JavaScript is indeed a slow language. However in Gnome that isn't the bottleneck. People have been making UIs with JScript (basically JavaScript) using WSH back in the 90s on Windows 98.
> Most technical decisions aren't really driven by what makes a better end-user experience or a better product, it's mostly defined by convenience and familiarity of substandard software developers - with mostly and primarily web-slop background.
What makes a better end user experience has nothing to do with any of this. There has to be an incentive to create a good end user experience and there simply isn't in the vast majority of cases.
In many cases it doesn't matter really what the tech behind something is. Most popular programmings and associated frameworks all work reasonably well on machines that are over a decade old. I am running Discord on a 15 year machine dual core laptop processor and it works "ok".
So this sort of complaining about "modern devs" I've been hearing about for almost 20 years now. The issues I've faced with doing quality work has been almost always to do with how projects are (mis)-managed.
>But the checker can smile at me. Or whine with me about the weather.
It's some poor miserable soul sitting at that checkout line 9-to-5 brainlessly scanning products, that's their whole existence.
And you don't want this miserable drudgery to be put to end - to be automated away,
because you mistake some sad soul being cordial and eeking out a smile (part of their job really) - as some sort of "human connection"
that you so sorely lack.
Sounds like you only care about yourself more than anything.
There is zero empathy and there is NOTHING humanist about your world-view.
Non-automated checkout lines are deeply depressing, these people slave away their lifes for basically nothing.
You think those people won't need to enslave themselves somewhere else if the checkout line is automated? Asking for your job to be automated in a capitalist economy is putting the cart before the horse.
I don't know. You are correct that the Japanese/Korea has this mentality to spend time at work just to spend time but there is a noticeable delta in effective results between those countries and modern China.
> Some people don't have the luxury of preaching for whatever ideals they have without a need to release anything in 10 years
Wait, how did they gain this "luxury"? Are they trust fund babies or something?
Or did they earn their big stash of money by producing "garbage" and now retroactively are preaching ideals that they themselves didn't follow or what?
This line of "criticism" doesn't make any sense whatsoever.
After all both in question live off money they've made and/or are making from their (arguably) uncompromised quality work.
That is to say their uncompromised quality work has directly resulted in them being able to not release anything for close to 10 years, and practice their ideals in software they ship even if the "shipping" takes 10 years to do.
It would be more fair to say, that most people don't have the craftmanship and skill (and not the luxury) to be able to produce high quality work and software that enables them the so called "luxury".
>Or did they earn their big stash of money by producing "garbage" and now retroactively are preaching ideals that they themselves didn't follow or what?
In the JBlow case - yes, he made his money using C++. So far, he hasn't shown that using Jai is particularly productive for software engineering.
> So far, he hasn't shown that using Jai is particularly productive for software engineering.
And how would he do that exactly to whatever ungodly standards you are setting for the man?
Many people have criticized C++ in past (which is very easy to do), yet he's practicing what he's preaching in the most direct way humanly possible, he's both (1) designed and implemented a new programming language (that has directly addressed most of the issues), whilst (2) also making a complete non-trivial game in the newly designed language at the same time.
His games have always taken long time to make, and now he's making game + engine + programming language. At the same frigging time!
The only "luxury" JBlow has is that he's an exceptional individual and you're not. He has rare combination of ability, perseverance and work ethic, and by all accounts most people are neither of those things at once.
Most criticisms 99% of time are either misrepresentitive, misinformed jealousy or something to do with politics.
I have no issue with personally acknowledging that some rare individuals are simply way better than me.
And to prevent sounding like a gushing-fanboy, I suspect that his newest game won't sell very well, because his first two games have atleast something to appeal to general public (either visuals of Witness or time travel mechanics (somewhat novel at the time) of Braid) while this game doesn't appear to have the same draw.
This game has too much of a generic-sokoban puzzler vibe to it to appeal to the general public who aren't already ardent puzzler fans (and are there enough of those and can he reach enough of them? etc). And the trailer doesn't help to change this perception.
>And how would he do that exactly to whatever ungodly standards you are setting for the man?
By providing a result in a way that will be superior to the current status quo. Maybe there will be results, but right now there are none.
I have no idea why you are so invested. I don't care about the man's personality or whatever qualities he has. I look at what he does, and so far, he spent 10 years making a game that you yourself admit won't be even that good.
Of course, you could say that changing the course of the industry not possible in one man's lifetime, so you'll need to gather round more people to get the action going, but this tone actually prevents you from starting a Jai revolution.
>I don't care about the man's personality or whatever qualities he has.
The only thing I'm addressing is the so called "luxuries" you alluded to, and the alleged "luxuries" he has is directly a result of his personality and his qualities.
The only reason you don't have those so called "luxuries" is because you're not even in the same ballpark as good. It really is as simple as that.
> By providing a result in a way that will be superior to the current status quo.
But he's done exactly that.
> I look at what he does, and so far, he spent 10 years making a game that you yourself admit won't be even that good.
I'm not saying that the game won't be good necessarily, I'm saying the game probably might not sell very well (atleast not to justify the amount of money spent from purely business perspective, etc)
He hasn't. He made a programming language that allows making a sokoban game in 10 years. That's probably not what people need. The industry can make similar games in a course of several months. It doesn't look like a groundbreaking achievement to me. A monumental amount of effort, sure, but the _result_ isn't there.
Plus, _in the past_, he made Braid, in C++, in a relatively practical way. He made money using the industry standards, now he loses money deviating from the industry standards. The question I'm interested in is: why would anyone listen to what the man _says_ if his own preaching makes him lose money?
But okay, you don't want to hear any of that. You keep fixating on the "luxury" part. The reason we talk about JBlow is because he made Braid back in 2008, and it was an awesome game, and it sold well. More importantly, the timing when it released - it was what kicked off the boom of the indie game development back then. He also made The Witness, and although it was also a good game, it was most likely not as groundbreaking as Braid, considering that he chose Braid instead of The Witness for a remaster. And then he complained that it, quote, "sold like dogs**", end of quote. Unfortunately, what was the jewel of the indie game development in 2008, doesn't really excite the audience that much in 2024. The world has moved on.
The music indsutry is well aware of a phenomenon of a "one-hit wonder". If the JBlow's qualities were the only reason he could make Braid and get rich enough to not release anything for a decade, then surely anybody with these qualities could make Braid 2 and do the same thing, correct? Well, nobody can do that. Not even JBlow himself. Not anymore. It's not 2008.
> The music indsutry is well aware of a phenomenon of a "one-hit wonder".
He made two hit games, Witness was released 7.5 years later.
> Within a week of release, Blow stated that sales of The Witness had nearly outpaced what Braid had done during its first year of release.
> The Witness is widely regarded as one of the best games of the 2010s. The game appeared on 'Best of the decade' features from IGN,[103] Polygon,[104] NME,[105] CNET,[106] and National Post.[107] Edge considered the game the 22nd-best game of all time in 2017
Calling him "one-hit wonder" simply has no basis in reality.
He's at minimum a two-hit wonder.
> it was most likely not as groundbreaking as Braid, considering that he chose Braid instead of The Witness for a remaster.
Now you're making shit up on the spot to make an argument.
Think for a second will you, how exactly would he remaster Witness? Braid Anniversary Edition was announced on 2020,
at which point Witness would merely have been ~4 year old game.
Braid was also made for a 720p console, the Xbox360 Xbox Live Arcade service, so remake atleast makes some sense.
> The question I'm interested in is: why would anyone listen to what the man _says_ if his own preaching makes him lose money?
What exactly is he _preaching_? Not what you have cooked up in your mind, but actually _preaching_?
Why would anyone pay attention to the man who has made TWO hit games in a row, and a third one in his own programming language (that has inspired countless other programming languages like Zig and Odin), yes, why indeed people would listen to an exceptional guy who has repeatedly demonstrated competency and delivered results, whilst always putting it all on the line?
Can you make atleast one hit, not two, just one? Or anything of note?
No you can't, you can do nothing, that's why you don't have the "luxuries" and people don't listen to you, but pay attention to him. You might not like it, but it is what it is.
And you like to comfort yourself with the thought that you don't have some sort of unearned "luxuries", because otherwise you would do great things.
But the reality is that he's exceptional and you're not.
That the game development industry requires a new programming language. So far, the evidence for that is slim. (I mean, metaprogramming with #run is cool, I'm also fed up with cmake. But surely we don't need to throw away all of our C++ tooling for that? Nah, we probably need something more incremental.)
> Calling him "one-hit wonder" simply has no basis in reality. He's at minimum a two-hit wonder.
Okay, I've been corrected. The Witness also sold really well. So he's a two-hit wonder, he clearly had developed a process to make great-selling indie games. I admit that, I admire that. (I only said good things about the guy anyway, why you would call me a "hater" is beyond me.) But now, he deviated from this process. His primary goal now is clearly not to create a good game, but to promote Jai.
> why indeed people would listen to an exceptional guy who has repeatedly demonstrated competency and delivered results, whilst always putting it all on the line?
Because there are limits to everyone's competence. It's like a generalized Peter's principle - being successful in one area doesn't mean you'll succeed in all others that you put your hand in. Even John Carmack didn't really succeed in rockets.
After all, the game dev industry is showbiz. Its ultimate goal is entertainment. JBlow is an entertainer, first and foremost. There are a lot of musicians and actors more influential than JBlow, does that mean I won't be a fool if I listen to their opinions on anything more important than what to eat for breakfast? No, not really. And in the same way, not a lot of people will choose Jai for programming, not in the next 20 years for sure.
> Can you make atleast one hit, not two, just one? Or anything of note?
No, absolutely not. I'm actually the most useless creature of all, good for nothing (other than keeping you engaged, apparently). You got me. And I'm not even trying. I'm not trying to preach for anything, develop new industry approaches or whatever. I'm just humbly making a point: but even if I weren't the most useless, I wouldn't be able to reach the JBlow's heights. Even if I had the same set of skills that JBlow had in 2008. For example, a notable part of the success of Braid was thanks to a contract with Xbox Live Arcade, and where is XLA now? The world has changed. The market has changed. The audience's needs have changed. Becoming an indie dev of such caliber now requires a different set of skills, one that a single person might not even physically have.
At some point, you'll have to admit that (1) it's not only the qualities and the hard work that brought JBlow to where he is, but also sheer luck, and therefore (2) yes, it's a luxury. If you don't believe in (1), well, okay then. But if you agree with (1), from that (2) trivially follows. If it doesn't for you, then it's purely semantics, I guess.
> That the game development industry requires a new programming language. So far, the evidence for that is slim.
I love how you one hand acknowledge your severe lack of ability and achievement. And yet at the same breath you confidently put forward to know better - than JBlow no less - what the game-dev-industry or world at large needs. Or that you'd even have the ability to gauge evidence for it(or lack of it).
What evidence would even qualify as proof that game-development-industry (or world at large) requires a new programming language?
What is the exact threshold of "suck" that you have to cross before you go "yup, we need something different"? Does such threshold even exist?
And how do you measure it?
> There are a lot of musicians and actors more influential than JBlow, does that mean I won't be a fool if I listen to their opinions on anything more important than what to eat for breakfast?
Is John Blow making bold opinionated statements about fine-dining or something?
No? Then what are you even talking about?
Why are you constantly making shit up to discredit the guy?
This is NOT rational behavior, its some sort of ego defense:
"like how dare he say bad things about C++, who does he think he is (just some one hit wonder game-designer, just got lucky!)? He has no idea what he's talking about!"
Except, he making statements about a language that he has used extensively for more than 25years at this point. And used it to ship large, intricate, largely succesfull hit-games all with their own engines where he has done bulk of the programming work.
That is to say, you can HARDLY find anyone more competent and suited to comment on deficiencies and shortcomings of C++, and how to improve them and fix them.
Now, just because he makes astute observations about various defects in C++ doesn't make him special, after all C++ is extremely badly designed mess, and it is very easy to do so, and thousands of people have done so.
What makes him special - is that he - has mostly delivered on this (stretching himself thin in the process), whilst also making a large game at the same time. This is very rare and exceptional.
> metaprogramming with #run is cool, I'm also fed up with cmake. But surely we don't need to throw away all of our C++ tooling for that? Nah, we probably need something more incremental.
Who is this council of "we" you're refering to? A council of average Joeys that haven't shipped anything of note and is more concerned with whats "cool"?
You have roughly zero idea what the actual painpoints of making and shipping large games are.
His latest game does full rebuilds in 2 seconds, so he can iterate and make changes quickly.
There are no "incremental improvements" that can be done to C++ to suddenly make builds not take MULTIPLE MINUTES.
> JBlow is an entertainer, first and foremost.
This is what you have got wrong, JBlow is an exceptional programmer first and foremost, who also happens to be a pretty good at thoughtful gamedesign, and pretty good at doing public speaking, among other things.
> a notable part of the success of Braid was thanks to a contract with Xbox Live Arcade
Notable part of success is that he made Braid interesting enough to win "innovation in game design" at IGF.
Winning IGF ment he got contract with XLA (interested in making money and promoting platform)
This is a deterministic process, there's no dice rolls or lottery draws involved here.
If you're exceptional and you make exceptional things you succeed sooner or later, statistically speaking.
The whole thought process that if you spawned another much younger JBlow in 2026, he would be attempting to make another verbatim Braid, instead of something completely different - way more attuned to current market conditions is not very bright. He (the young JBlow clone) might not even choose to do games in these market conditions, he might chose to do exceptional, highly influential work in a completely different domain.
What however is highly likely is that he'd be highly, highly successful at whatever it is. Because highly exceptional hardworking people just succeed (unless they are born in Mumbai or Karachi)
I mean, if you're born as an average Joey, instead of being born exceptional, it is _luck_.
After all, who would choose to be average when they can be exceptional and bright?
But it is important to acknowledge at which point luck materializes.
And the lucky event wasn't XLA at 2008, the lucky event is beign born exceptional.
Most people would call being born rich a luxury. And not - being born exceptional and applying the said talent and hard work to ever more ambitious projects.
> Even John Carmack didn't really succeed in rockets.
He was very successful at engineering aspects of rocketry considering his very small and completely self funded budget. Just not comfortable burning 1mil of his personal funds / retirment money per year (that was still considerable money to burn out of personal stash in 2007/2008)
Just to be clear, your comments are implying everyone who doesn't write everything from scratch is shipping garbage.
Ignoring how misinformed that opinion is, I would say The Witness is a very compromised game. Maybe if less focus went into the technical aspect, it could've been better.
How is this my fault as customer? This a predatory practice in tech.
I work in automotive, the hoops you have to jump through in order to push a SW update are enormous. One of the first rules is: if the owner of the vehicle does not consent to an OTA update, you're out of luck.
The industry is obviously unable to self-regulate, so it is time for an external regulator, e.g. the EU, to jump in and mandate that SW updates cannot be applied without explicit consent and an explicit explanation of what is being changed. Of course, security updates must be maintained separately from feature updates like this.
As a consumer, I always want the latter, rarely do I want the former. My device, my choice.
GIMP has god horrid UX, there's no way it could have eaten Adobes anything.
There's lineage of FOSS apps that stick by the "we're not X, we're different from X." mantra.
The discomfort, frustration and unintuitiveness you're feeling from using our app? It's just you!
No, that's not bad design and bad UX! its simply because we are different! We aren't X (Photoshop), we just do things differently here!".
We've implemented a number of items from the issue posts once consensus was reached, and we hope more people will participate and help improve GIMP further.
People don't buy into this kind of signaling these days. It just does not work anymore.
Which e-waste are you currently running on?
You're just not. I know it.
Instead you run out and buy the newest shiniest thing so you can put a docker container into another docker container. And fill landfills with ewaste as a result.
Your engineering practices most directly contribute to ewaste, because extremely powerful PCs from 15 years ago doesn't hold up anymore to ever more shitty layers of javascript and vibecoded python stuffed recursively into ever more docker containers.
geohot is just being honest. That is respectable. All the signaling bullshit - is not.
Could you please not post in the flamewar style to HN? You've done it repeatedly in this thread, unfortunately, and we're trying for the opposite here.
I'm tired of this insane right-wing opinion that having an opinion against waste is just virtue signalling, so we shouldn't care about anything because it's all just "signalling" and it's somehow more honest to not give a shit about anything. It's tired and it's disgusting and it's how we get a world where we don't care about improving our society or our environment.
I do care. You might not, but that's a you problem and not anything to do with me signalling anything. I'm being honest about everything I say. You not accepting that says more about you than it does about me.
Could you please not post in the flamewar style to HN? You've done it repeatedly in this thread, unfortunately, and we're trying for the opposite here.
That's unfair. I'm going unfairly called out even though I stand by my beliefs and I'm only trying to defend myself. How are the replies to my comments in any way acceptable and according to HN guidelines???
"Put your money where your mouth is", come on. That's not acceptable and it's provoking. How can you defend bullies like this?
I specifically responded with an identical reply to the other main commenter you were arguing with, who was also breaking the rules.
Still, someone else breaking the rules doesn't make it ok for you to do so, and pointing the finger at others instead of taking responsibility is not a helpful response.
"Normal" people will just buy the game if it's good.
So it's irrelevant if it uses AI or not. Ie. it's not a sales pitch and not part of decision making process when making the purchase.
There are increasingly more games that use some form of AI generated content, voice lines or otherwise, and nobody could care less, except the people outlined above.
By your own admission its not irrelevant, there are a small group of people who do care about that kind of thing on either side. For an indie dev, that matters. AAA studios can pretty much guarantee at least a few thousand sales, indie devs, especially the less established ones, have far less. For first timers, there'll be none at all.
The thing is though, appealing to the pro-AI crowd is much more difficult. They want a game thats a shining example of what AI can be in gaming. The anti-AI crowd doesn't need that, they've got examples of that for decades. A few AI generated voice lines won't do much to appeal to the pro-AI crowd.
Nobody is trying to appeal to "pro-AI crowd" (whatever the fuck that even means) when they use AI tools.
If an indie (or even less of an indie) is using AI generation, they are doing so to save costs or work around their very limited budget. Or using it to work around some limitations where voicecasting every line would be infeasible, etc.
And losing the small portion of the miniscule-vocal-always-complaining crowd (who odds are - wasnt part of their audience to begin with), to be able to use AI-gen is not a loss at all.
Data on Steam is telling, these tools are becoming increasingly prevalent.
Oh yes they are, there's a lot of games (or at least, promises of future games) that promise to be 100% vibe-coded or that make heavy use of AI in a way thats very prominent to the player. There was an example just last week:
> And losing the small portion of the vocal-always-complaining crowd (who odds are, wasnt part of their audience to begin with), is not a loss at all.
That seems like a very different crowd to me. I've been around the industry long enough to see the signs of that, and I don't see that much from the anti-ai crowd, or at least not in any more significant numbers. See: the project zomboid AI art issue
But like I say, for an indie, yes losing a small audience can still be a big loss.
It seems like you're way too bought into warring internet-weirdo tribe dynamics.
If you use or don't use a tool (your choice), it doesn't make you pro or an anti.
It's basic pragmatism, if a tools is useful to you, you use it, if it isn't, you don't.
The consumer base mostly doesn't care, nor should they. They care about end result.
Or else nobody would buy iphones, nikes and what not.
The moment you bring up "pros and antis" and tribe dynamics, I smell a brainrot from a mile away. You do you I guess.
"Normal" people already hate AI being chowed down their throat. The won't mind AI, when it doesn't feel like AI. As soon as it does, that is a bad not a good feature.
Only a small number of indie games will go mainstream enough for that to matter, I think. If your likely outcome is selling 10,000 copies getting in with the reviewer and blogger crowd is probably helpful.
I have some old borderline senile relatives writting apps (asking LLMs to write for it them) for their own personal use. Stuff they surely haven't done on their own (or had the energy to do). Their extent of programming background - shitty VBScript macros for excel.
It also helps people to pick up programming and helps with the initial push of getting started. Getting over the initial hump, getting something on the screen so to speak.
Most things people want from their computers are simple shit that LLMs usually manage quite well.
Good question whether or not this (outsourcing their thinking) actually just accelerates their senility or not.
As someone who likes to solve hard or interesting technical problems, I've long before LLMs often been disappointed that most of the time what people want from programmers is simple stupid shit (ie. stuff i dont find interesting to work on).