Get the top HN stories in your inbox every day.
trizoza
paufernandez
It is increasingly important to be able to see that many things are true. There is no single "truth". Many things are true at the same time, and in all aspects of life. Each brain is like a band pass filter, and the effort we should make is to try to imagine the points of view of others, which are just different slices of the same world. Then embrace the slices we like, and just ignore the ones we don't, but don't argue or fight for our slice as it if was the only one.
armchairhacker
To clarify, there are formal truths: widely-accepted hard science, e.g. “2 + 2 = 4”. Technically, there’s a point where we can’t fundamentally prove anything (“if a tree falls and nobody hears it, does it make a sound?”), and rarely we get things wrong (e.g. classical physics)….but in practice, these are true, end of discussion.
Then there are informal truths: e.g. “the Earth is round”, “the sky is blue”, “Gala apples are red”. You can nitpick them (the Earth isn’t a perfect sphere, the sky is only blue during the day in areas without high pollution, Gala apples may be pinkish or have yellow blotches, or exceptional discoloration), endlessly or until they become formal (possibly by becoming self-referential). But in practice, these are also true (like formal truths; although it’s important to know the difference because…)
The problem is, there’s no line between an informal truth and uncertainty/opinion that isn’t true. Like you know ##FF0000 is red and ##00FF00 is not, but there’s no exact color that separates “red” and “not red” (it depends on person, mood, surroundings…) Consequently, unlike formal truths, informal truths have false implications (“fuzzy logic”). An informal truth can be phrased in a “misleading way”, priming the reader for a false implication (a formal truth can be phrased in a convoluted or unintuitive way, but interpreted formally, never leads to a false implication).
The vast majority of discussion is not formal. Even the smartest people constantly fall for false implications. And this isn’t completely solvable, because we fundamentally can’t formally define everything (too much detail): we tried with GOFAI, it failed and its successor, neural networks, informally defines things like us (by forming a lossy model of the world, then generalizing it).
ahepp
> rarely we get things wrong (e.g. classical physics)
Curious, do you think quantum physics is the end of the line? I wouldn't claim to have a good understanding of anything beyond classical physics, but I've just assumed it's turtles all the way down and at some point we'll find serious issues with the quantum model
ksec
>widely-accepted hard science, e.g. “2 + 2 = 4”.
Except we live in a world where people do argue 2+2 could also be 22 ( Because they use Javascript /s ) Which is basically people believe what they want to believe in. Rationale rarely works.
d4ng
2+2=4 is only a formal truth up to the axioms of arithmetic and how we denote numbers. The statement is not statement with regards to objective reality.
pmarreck
I think the biggest problem is not that there may be many truths, but that people don't even bother seeking out the best evidence or reasoning against their truths.
Especially in tech, for which there are basically only tradeoffs.
I use Zig in most things lately, and I use AI. I have a high quality standard (that AI honestly sometimes makes difficult to meet), but my github has never been more active:
I have a really good code-review skill (which I'm actually in the middle of updating, but it's here): https://github.com/pmarreck/llm_skills/blob/yolo/deep-code-r...
I also have a pretty neat (although arguably janky) way for LLM agents in different tmux terminals to talk to each other: https://github.com/pmarreck/llmsend
jasonjayr
The problem arises when there are contradictory truths, and defenders of one or both sides refuse to dig deeper to both self-reflect on what they believe to be true, and perhaps come to a deeper more correct understanding.
hallole
There is only a single source of truth and that is objective reality. Maybe you agree with that, but your wording is messy. It's true that different perspectives can yield their own particular bits of truth, if that's what you're saying.
coldtea
"Objective reality" is only the source for the least interesting truth. The truths that really get people fighting over concern best courses of action, moral matters, aesthetic issues, and things like that, where there isn't some singular objective truth (and even if there was, nobody has access to it).
rrgok
Uh, what is objective reality?
galleywest200
Different opinion != being weird.
sinpif
Yeah, well, that's just, like...
gr8painz
[flagged]
paytonjjones
It's fun when subreddits or Discords are premised on a specific subtype of weirdness. There, you'll get mobbed if you're normal.
yapyap
> nternet is beautiful because it's ok to be weird - this is often the opposite on twitter, fb, reddit and many discords where if you have a different opinion you get mobbed by angry comments making one feel worse about their own weirdness
which is ‘funny’ because by offline standards the average redditor will probably be seen as weird
dieseleration
I think it makes perfect sense for Zig to have their stand against LLM contributions while consumers of the compiler/Zig project overall use whatever code aids they like. Building a language is not a matter of churning out as much greenfield code as possible, but in careful consideration of whether or not some feature and its implementation fits coherently into the entire overall language. It's upstream of so much, and we now have decades and decades of examples where just letting rip with new additions renders a language schizoid and unergonomic. An LLM's tendency to "yes, of course, and," to any suggestion is not what a healthy language project needs, but it can be tremendously useful for someone employing a balanced and ergonomic language to generate products. I'm glad to see Mitchell keeping a cool head as the unfortunate tendency in so many devs to take sides and get dogmatic plays out yet again.
joaohaas
This is not the main reason for the ban. You can read the linked post in the article that explains the AI ban thing in more depth.
nilsherzig
If you're unsure about spending the time to learn Zig, I really recommend watching the following interview with the creator of Zig https://www.youtube.com/watch?v=iqddnwKF8HQ convinced me more than any design doc or blogpost could
zazuke
It's soo good, just watched it before.
Jonovono
Very interesting, but wild to me how much this guy resembles Elon (in not only appearance but also mannerisms and way he speaks)
bayarearefugee
> this guy resembles Elon
What an awful thing to say about a person.
heldrida
Finally someone else notices it!
strix_varius
Yeah he's like the light side Elon. Happy with his life, his family, his choices. Content. Zen.
keepupnow
Absolutely
Lerc
It's great to be in a position to do this, however I'm beginning to think that their greater contribution is ghostty
I don't really know how to value things any more when I see someone develop a tool that is kind-of useful that then gets acquired for half a billion dollars. As someone with a decent number of decades of terminal hopping, the improvement that ghostty has brought a breath of fresh air. To me it has represented more utility that a few of those acquisitions.
johnwheeler
I use Ghost TTY coming from iTerm for no other reason than I saw everybody else using and praising it.
Is there some special feature I'm missing? I would only call it a marginal improvement. If that. I fail to see what the big deal is.
neobrain
> Is there some special feature I'm missing? I would only call it a marginal improvement. If that. I fail to see what the big deal is.
Among the "GPU rendered terminal" options, afaik Ghostty is the only one that has proper search/context menus, tabs, and scroll bars. I'm sure it's easy to get by without, but compared to the overall value-add of these terminals (which exists indeed but isn't tremendous either) I find it quite a significant downgrade, so I appreciate that Ghostty has both.
novafunc
For me,
* available on Linux and macOS
* settings easy to transfer, just a file
* comes with Jetbrains Mono Nerd font built-in, no need to install it separately
* supports ligatures
fridder
I personally like how I barely had to configure it, how nerd fonts just worked, and how nicely it renders text
mixmastamyk
It’s not quite finished, give it time to mature. But pretty good already.
johnwheeler
Yeah, it's a good polished piece of software no doubt. I'm not denying that, but the hype it gets is just... I don't know.
kyrra
input latency. the time from pressing a key to showing on-screen is much lower with ghostty (I can't find exact number, but it seems to handle input 2-4x quicker. So around 15ms instead of 60ms).
Also just the general render pipeline is way faster in ghostty. There are things you just can't do in iTerm because it's so slow. Ghostty is attempting to improve the experience to allow for more things to be built in the terminal.
rafram
> input latency.
I guess, but I have a hard time caring about 1/22 of a second of additional latency when iTerm works so well overall.
> There are things you just can't do in iTerm because it's so slow.
Such as?
anotherevan
The ctrl-tab tab switching behaviour being round-robin instead of stack[1] is what is keeping me on Konsole. None of the newer consoles (for Linux) seem to have this.
throwaway2037
I can see that both styles could be useful to different types of users. Why not make it an option?
wickrom
I'd love to hear what made you settle on ghostty. There is not dearth of terminal emulators out there, each claiming performance or batteries included.
dust-jacket
I'm not the commenter, but for me ghostty was good for being a Very Good terminal experience with almost no config required.
Just checked and the config file for my daily use terminal setup is 3 lines long. 3! That means I know I can chuck it on any system, any clean re-install, and it'll be Fine. That counts for a lot when you've grown tired of endless config tweaking.
zemnl
Same for me.
My config is a couple lines longer, but other than font-family, font-size, color theme and a couple of other settings I didn't need to change anything else.
I definitely spent way less time configuring it to suit my needs that I did with any other terminal I used before.
ngrilly
Same. Almost everything works out of the box, with great defaults.
noisy_boy
Seconded. I keep hearing about ghostty but I have yet to see a strong enough justification about how it is _that_ better. I use konsole and has significantly more user friendly screen to manage settings. I heard about ghostty's performance so I did some timing tests and ghostty was faster than konsole but not that much - not in any perceptibly significant measurable sense.
warmwaffles
I went from Alacritty to Ghostty for ligatures and some other small goodies. I could probably get those same goodies with Kitty, but I didn't want to try nor have the desire to try. I may go back to Alacritty if I grow tired of Ghostty.
beepbooptheory
I never got the speed thing. Ghostty at least seems slower on my machine compared to foot(client).
binarin
libghostty is a bigger contribution, it's being embedded left and right.
hack1312
i switched from iTerm 2 on macOS because it would get bogged down sometimes or occasionally lag. it’s been noticeably faster and i appreciate the file-based config as well as the defaults, leading to my config being under 5 lines.
on linux i use the default terminal in gnome which is ptyxis now iirc and haven’t felt any need to switch.
copperx
Don't you miss the crazy configuration options that iterm2 has?
rofrol
Splits can have different font sizes
pelasaco
> however I'm beginning to think that their greater contribution is ghostty
Is that meant as an argument in favor of "another $400k for Zig" or against it?
I like Ghostty, but investing $800k to develop a programming language that is primarily known for producing a terminal emulator doesn't sound like a particularly strong argument in its favor.
Genuine question. I’m not trolling.
Lerc
It's just a recognition of things done.
If there is a greater message, most people could probably put a similar amount of work on a project for the community.
Not many can write a lege cheque for a cause. Many who can, do so and consider their job done.
It's a credit for someone to be able to do one but choose to do both.
Imustaskforhelp
> I use AI heavily. I've written about my AI adoption journey and shipping real features with AI assistance. I'm also quite vocal about remaining rational about its capabilities and frustrated with its negative impacts on open source.
> The point is that I have opinions. Those opinions don't fully align with ZSF's approach. And yet, I have nothing but respect for ZSF: the people, the policies, and the project. Part of what makes the internet and open source great is that projects can be weird and different. They can set unusual boundaries, build their own culture, and pursue quality in ways that won't make sense to everyone.
Mitchell does feel like the adult in the room when other people are having chain-saws and acting irrationally for a lack of better term (for example jared/bun controversy which the post just somewhat touches on)
(Mitchell's tweet about AI psychosis is genuinely influential and is now a pointer to what this phenomenon might be)
I really think him and simon's opinions are somehow decently nuanced opinions on AI that the internet has to offer.
Now glazing of mitchell aside, I am happy that zig foundation gets such amount of money and I am really excited that Zig an independent language is able to get the level of love that it does.
There is a famous talk by the creator of Elm on the economics of independent programming languages and how its hard for them to get sponsored if they aren't already working at a company (Rust was created at Mozilla, Golang was created by Google)
This is a real issue that is true for most of open-source and I am just happy that we are atleast moving slowly towards some good as well. Its an uphill battle with multiple lows but I am happy for the positive changes as it gets as open source does have a special place in my heart as it taught me about privacy and many of your hearts as well.
Arrowmaster
As things are right now, I see this as a respectable way of operating.
Michael has made his views and usage of AI known. The Ghostty project has a detailed AI policy for users to see and the team is willing to devote resources to enforcing a middle ground policy. The Zig project has a detailed policy taking a strict stance and as a result I expect they do not have expend as much resources when a contribution is suspected of being AI assisted.
A strict policy on either side is easier to enforce based on finite resources (mostly people). I'm sure many projects would like to have a middle ground policy but cannot currently devote the resources it would require long term. We might never see a shift in moderation abilities and this remains for the longer term, or there could be advanced in moderation that allows projects to adopt a more nuisanced policy that's right for them.
mortar
> Michael has made his views
Assume you meant Mitchell?
Arrowmaster
Yes. I think autocorrect got me on mobile keyboard.
GodelNumbering
I have been experimenting with modifying Ghostty lately. It's a well attended codebase and a pleasure to work with, props to Mitchell.
Since Ghostty is written in Zig, I ended up adding native Zig AST support in Dirac (https://github.com/dirac-run/dirac/blob/master/src/services/...)
One thing the has been a little unintuitive is the pattern of all code and tests in single files, which makes the filesizes grow much larger. Also if you're coming from inheritance supported languages, Zig forces a different way of thinking
Hasz
Mitchell Hashimoto, talk about putting your money where your mouth is. What a cool dude. Much respect!
osigurdson
Zig is really nice. I enjoy using it a lot. Glad to hear that it is getting a little more funding.
teekert
Adults responding in adult ways. Respect.
RustSupremacist
Donations to the Zig Software Foundation largely go to contributors to the language. Meanwhile, donations to the Rust Foundation largely go into pockets unknown.
The key difference is that the Rust Foundation is a 501(c)(6) and not a 501(c)(3). The Rust Foundation would do better for the community if they were a 501(c)(3) and more transparent about finances. Follow this example for the greater good.
acedTrex
I started using zig more heavily for some edge device ML inference projects lately after watching Andrews jetbrains interview and it really really resonating with me on a personal level.
Am also really overall enjoying the language, it def has some rough spots regarding documentation and the stdlib but overall has been very nice to work with in neovim.
I can't throw 400k but I'll go ahead and pledge some dollars towards it as well.
qudat
Major props to Mitchell (and his family) for these donations.
Get the top HN stories in your inbox every day.
What a word of wisdom right there, the bit about internet is beautiful because it's ok to be weird - this is often the opposite on twitter, fb, reddit and many discords where if you have a different opinion you get mobbed by angry comments making one feel worse about their own weirdness.