harrisi

Perhaps. The thing I was trying to highlight more is comments (particularly disapproving ones) about the intersection of JavaScript (especially React, apparently), TUIs, and LLMs.

I just thought it was interesting. I'm a fan of TUIs, both positive and negative about JavaScript (the modern language is quite nice, given you avoid the historical warts, in my opinion - the ecosystem is unfortunate), and I think the best description for how I feel about LLMs is that I'm bearish. Also Simon is a fantastically prolific and intelligent person though, even though LLMs are not my cup of tea.

https://news.ycombinator.com/item?id=48960912 is a recent comment by someone I also respect related to how LLMs recreate human-like content (Widgets, really), making actual human-created content feel fake.

Sometimes it sure does feel like ELIZA is the main news source these days.

I don't want to edit this to change the narrative, but I wanted to clarify that I'm not not a fan of all of the technologies or people related to this. As far as compan{y,ies} go, though, well..

Maybe I'm taking crazy pills, but I swore there have been very similar comments I've seen as the current top post by weakfish:

Maybe I’m taking crazy pills, but I’m still stuck on “why the hell does a TUI need to run in terminal React by way of JavaScript”

The fact that Anthropic felt the need to buy a runtime so they could make their TUI better speaks more to the quality of engineering than anything else IMO.

If rewrites are so easy, why not rewrite CC in a native language? Would’ve been a hell of a lot cheaper.

It turns out, there's some similar sentiments in the last year ish:

https://news.ycombinator.com/item?id=47043325

https://news.ycombinator.com/item?id=47105720

https://news.ycombinator.com/item?id=48001113

https://news.ycombinator.com/item?id=46134570

https://news.ycombinator.com/item?id=46124596

I'm not a fan of any of the technologies, companies, or people related to this. I just couldn't shake the feeling that I've seen similar comments.

September and all.

Interesting. Well, for information about the project: https://lib.rs/crates/iocaine

The site linked is excluded from the wayback machine.

I believe they're referencing iocaine[0], which I believe I've seen on codeberg before. I may be wrong though.

[0] https://iocaine.madhouse-project.org/

A hint would maybe be good. You could highlight the first letter and a direction, for example.

The share functionality could be improved. Sharing the score and time to the specific puzzle would be nice.

Also I came back to this thread to comment that this is the most engaging, fun, and satisfying word puzzle game I've maybe ever played. My only complaint is that there aren't more puzzles!

Thanks for the fun :)

It's so interesting to me how people consider writing bash scripts to be some sort of absurdist gimmick. I've made some scripts like this for blogging, as well as many other things.

It's often not the best choice. There are situations where, I'd argue, it is. One of the main reasons, in my experience, is for very lightweight servers. I don't want python or node or anything. I want to run a web server with 128mb of ram and a 1gb disk. Somewhat niche, but it's useful to just use the base system.

The other useful thing about it is knowing your shell in general. Even if you're not on the infrastructure team, being comfortable knowing how to navigate a system is both fun and sometimes critical. You never know when you'll need it, and it's an interesting look into your tools.

It's akin to saying "why know how to use a saw when you can use a chainsaw" or "why know how to make a fire when you can use lighter fluid and a lighter?" Sometimes things come up. And it's fun to know how we got to where we are today.

https://github.com/fusionlanguage/fut/blob/master/doc/refere...

Probably my favorite thing about the author's posts is that it's all handwritten, vanilla code - as far as I can tell. It seems like all the HTML, CSS, and JS are just plain, standard, universal code. It's one of the few "advanced" sites that just works on an old iPhone 7, for example. So many modern sites that use modern frameworks just don't work anymore.

Also just a huge fan of using the existing infrastructure of browsers - even older ones - in effective ways. Browsers have been quite capable for awhile now.

Millions of people died when a possible solution of "hey, everyone, let's stay inside for a couple weeks" could've possibly effectively eradicated the virus does not seem like a great example of humanity's ability to cope with imminent global existential threats. The potential solution(s) to the massive brick wall we're speeding towards are far more inconvenient than "everyone just hang out for a minute." It's radical change to everything in society.

The comparison with Covid is also striking because the only reason a global "don't travel too much" solution couldn't work is due to the nature of capitalism. It's not like we couldn't feed everyone. It's just that some people with too much wouldn't gain as much for a little bit. Which is the same root cause of why solving climate change is impossible without radical change.

Unfortunately, I don't know the answer. I'm quite certain it's not to maintain the status quo, though.

Both the GitHub repository and the website explicitly claim the entire project is entirely generated by Claude.

We exist in a world where the exchange of goods and services is inherently oppressive. Some people draw the line at working for Northrup Grumman, some at RJR, some at Meta, some at Starbucks, and some at the local farm. I'm not one to judge where the line should be - I'm not even sure if there is a moral or ethical line that exists in the system.

It is odd. There's significantly more new users commenting here than every other submission on the front page, both in absolute numbers and proportionally.

¯\_(ツ)_/¯

It's interesting that there are, at the time I'm commenting, 11 new users commenting on this submission, some commenting multiple times. I wonder what the effect of "share my score" type pages have on account creation.

Just want to say thank you for sharing your project. Very fun, and I wouldn't know about Matthias Wandel's version if not for yours!

Also, both of these tickled my brain in a great way. I think a potentially fun continuation would be to "eyeball" physics. For example, throw a ball and pause the physics before it hits something (ground, object, who knows?) and guess the location. Or show two objects about to collide with certain shapes and masses and guess what one of them will hit first and where.

I know who Petzold is but I can't remember if I read this at the time. I do love it though. I was thinking a lot about generated code in Visual Studio a couple years after this article.

It's kinda funny to think about various forms of code generation. From compilers to IDEs to parser generators to, now, LLMs. Even several higher level languages that compile to lower level languages are generative, essentially.

Still not a fan of LLMs, but it's always a good to remember that the concept isn't entirely new or unique.

Ironically, my first thought was using Automator or AutoHotKey (there's a different one for macOS I think? But you get the point) to just identify those dialogs and click yes/allow/whatever.

Even though a bunch of the responses are "well you don't want a keylogger" when the first solutions I can think of are also (potential) keyloggers. :)

The post says the purpose of exploring this space (which is a fun one) wasn't speed, but representation. The speed gain was an added value.

I'm not saying SQLite's varint implementation is ideal for every application. It's just an implementation that is one of the most used implementations, if not the most (I'd bet it is by a large margin though). It just seemed like a missed opportunity to compare it with the implementation they landed on.

EDIT: Just wanted to add, thanks for sharing that link. Interesting!

I'm surprised there's no mention in the post or here about SQLite's varint encoding. Not that it would necessarily satisfy the constraints, but it's one of the most used varint implementations.

I see what you mean. That article makes a few things seem more dramatic than they are as far as I can tell. The report says there was a 2% reduction in canopy coverage between 2010 and 2020 for Wake county. Or 3.6%, relatively. I can't find exact numbers for Raleigh.

I'm not trying to take the stance that there's no problem. I think we should spend far more resources, time, and education on the importance of nature. I was just pointing out that urban forestry has been a thing for awhile. The two unfortunate sides to it, at least if you ignore all the rest of modern infrastructure, is that one, there's only so much space you have to add more trees in urban areas, and two, it's treated more as a necessity which has a lower-bound to reach than an ongoing project to improve.

So, succinctly, we should plant more trees. :)

Side note: it's unfortunate that the news article doesn't link to the report, as far as I can tell. I believe it's referencing this: https://s3.us-west-1.amazonaws.com/wakegov.com.if-us-west-1/...

Your prediction came true decades to centuries ago, depending on where you are in the world. Urban forestry has been an important part of city planning for these (and more) reasons for quite awhile.

You're off by an order of magnitude, sadly.

Unfortunately this is a good example of kicking the can. Not to the next administration but to after the next elections. Some aspects are felt already but not all.

It's a good time to be kind to your neighbors. No matter their background, they're almost certainly not the ones to be upset at.

It can be quite politically valuable to kick the can to the next administration.

Depending on storage constraints, you could always dualboot. That would give you the exact same hardware to compare, and it's not a full commitment.

Anecdotally, I find that getting Linux on somewhat older or underpowered hardware is always a massive positive. Better performance as well as battery life. I'm not as familiar with modern hardware's relationship to either OS ("OS vs. some flavor of OS based on a similar or same kernel" - I know) with modern hardware. Worth a shot though!

Every supercomputer seems to do quite well with Linux kernels. Probably good enough for Crysis :)

I was curious about this from another video in relation to this case. I have no legal training, but I think there's no reason the witness would refuse to answer when counsel objects. There's no judge in the room as far as I know. The court handles the objections after the fact, I think, which could potentially have implications on how the trial proceeds.

Again, I have never practiced law, so I may be entirely incorrect. Also, I am not defending the witness or their actions.

I wrote a fix for node that got upstreamed a few years ago on a Lenovo Thinkpad 3 Chromebook. I'm actually commenting from it now. It's not a workhorse by any means, but for $99, it's not bad. A 1.1GHz Celeron processor with 4GB of memory is able to compile projects like node, python, Erlang, etc. without much hassle. It just takes a lunch break :)

Any modern Mac is more than capable. I had the baseline M1 Macbook Air that I did work on as well, just to see how that fared. Much better than this machine - 10x the price, but more than 10x the performance. This one is great as a "I don't mind if I break it or lose it" device.

It's also built on an unstable language. I'm a big fan of Zig, for what it's worth, and also it's important to take into consideration.

Testing is not a proof a software system is correct. Also, if tests are generated as well, there's no trust in how anything works or if the tests are covering important aspects.

The aspect of "potentially secure/stable code" is very interesting to me. There's an enormous amount of code that aren't secure or stable already (I'd argue virtually all of the code in existence).

This has already been a problem. There's no real ramifications for it. Even for something like Cloudflare stopping a significant amount of Internet traffic for any amount of time is not (as far as I know) investigated in an independent way. There's nobody that is potentially facing charges. However, with other civil engineering endeavors, there absolutely is. Regular checks, government agencies to audit systems, penalties for causing harm, etc. are expected in those areas.

LLM-generated code is the continuation of the bastardization of software "engineering." Now the situation is not only that nobody is accountable, but a black box cluster of computers is not even reasonably accountable. If someone makes a tragic mistake today, it can be understood who caused it. If "Cloudflare2" comes about which is all (or significantly) generated, whoever is in charge can just throw their hands up and say "hey, I don't know why it did this, and the people that made the system that made this mistake don't know why it did this." It has been and will continue to be very concerning.