mewse

I remember usenet where every forum was exactly the same

alt.confident.assertion.question.doubt.disagree

;)

In this thread, we pretend that the difficult and time-consuming part of a code review is all the reading you have to do.

Or, since we're apparently playing the game of maybes in this thread, maybe the LLM was only trained on the teams grandmothers' spaghetti recipes, so that new hires can learn to make the best bolognese sauce.

That headline feels like a really clever metaphor for something but I can't figure out what.

Green bubble is messages you sent via SMS (and so may have been charged by your carrier depending on your cellular plan)

Blue bubble is messages you sent via iMessage.

All incoming messages are grey, regardless of whether they were sent to you via SMS or iMessage.

I suspect a lot of people 'invented' the effect at approximately the same time. Honestly, the Dreamcast was the first piece of hardware really capable of doing the effect to a high level of quality in real-time.

I developed the cel shading effect for the Dreamcast game 'Looney Tunes: Space Race' (developed by Infogrames Melbourne House) literally during the first week we had access to a Dreamcast development kit. Infogrames Sheffield (devs of Wacky Racers) were shown an early version of our implementation, and added the similar effect to their game. It looked great, but went into their game pretty late in production, so the game hadn't really been optimised for it the way that ours was.

And the folks behind Jet Grind Radio came up with the effect on their own as well, and beat both of us to market. They were using exactly the same algorithm, but were using it in a very different way; they were fully embracing and leaning into the uneven, wide and jagged outlines, where Sheffield and we were fighting against them and trying to match a more uniform and traditional art style.

And then only about a year later, somebody seemed to have figured out how to make the edge-detection cel shading approach work in real-time on Xbox, for the game "Dragons Lair 3D". I had done a test implementation of that approach on the Dreamcast, but it wasn't nearly performant enough for us to run it on multiple characters at once while playing a game too! Not sure whether it was due to the Xbox being more powerful or them just having a smarter algorithm than mine, but you can't argue with their results! If you're making a game that you want to look like an actual hand-drawn cartoon, that is still absolutely the best quality way to do it, IMHO.

Someday I'll find an excuse to try my hand at implementing one of those again. Performance shouldn't be a problem at all any more, I imagine!

As a practical matter, studies about links between major diseases will always always be talking about correlation.

To reach 'causation' would require intentionally giving your experimental subjects Covid-19 (and in a way that didn't result in them knowing they'd had it!), and that's unlikely to pass muster with the ethics review board.

I can't stop thinking about this throwaway parenthetical at the start of the blog post:

[...] for many writers, writing a book is about the last thing they should do (unless they feel a book bursting out of them, much like a facehugger).

Now, as we all know, the aliens that burst out of people in the Aliens franchise are called 'chestbursters'. "Facehuggers", by comparison, are hatched from alien eggs.

So in this metaphor, since we're told that novels are facehuggers, the writers must be the eggs. And by process of elimination, we can deduce that the innocent starship crewmembers being attacked by facehuggers (novels) are innocent readers.

The metaphor actually contradicts the author's main thesis, since every egg (writer) does in fact contain a facehugger (novel). But contrariwise, all the human characters (the rest of us) would be much better off if those novels just stayed inside the writers and didn't insist on being written or read.

Metaphors are like scissors; they're twice as much fun when you run with them!

As a game developer, I'm absolutely not going to risk my players' engagement with the game by putting character personalities, dialog, and (therefore) plot into the hands of an AI that's going to play out differently for each player. No way, no how. That's where all my game's value is - that's what pays my rent - and I will not be handing it off to a glorified RNG.

Not the person you're replying to, but I agree that for me (as another heavy shift-G user), vim-racer having `:relativenumber` turned on is where my own troubles are coming from.

I'd really love to be able to specify whether I want absolute or relative numbering (or both or neither); the :set commands don't seem to be implemented?

Pedantry: "Craftsman" was never a company and didn't "go out of business".

"Craftsman" was originally a store brand used by Sears, under which it sold tools that Sears contracted various third parties to produce.

In 2017, Stanley Black & Decker purchased the "Craftsman" trademark from Sears (Sears Holdings at the time), giving Sears a long-term royalty-free license to continue producing and selling tools under that trademark (I think that license lasts for about another decade at this point? And then there'd be a small royalty payment for using the trademark beyond that point.)

As a result, several different stores now each carry tools all called "Craftsman", they're generally not the same tools; different stores sell different tools produced in different places, and all sell their own versions of the tools under the same "Craftsman" brand name.

This undoubtedly causes absurd levels of consumer confusion. As you say, it's absolutely causing tremendous reputational value for the trademark.

That's so clever! I would have guessed that film reels would be too fragile for that sort of continuous usage, though I'll confess that I have no personal experience with them.

I was really curious about how they would handle 'rewinding' a film reel between plays; assuming 16mm film, I figured, it would have been about 30 meters of film, so I wondered whether you could splice it together into a full loop and concertina it inside the cabinet so that it doesn't actually require rewinding between plays, and can instead just keep playing it continuously in a forward direction, and... after a bunch of Internet trawling, I found a forum post by somebody who had purchased and restored two units, that's exactly what they'd done! (I had my calculations about the length of the film wrong, though; apparently the full film is only about 10 meters long, which would definitely make it easier to wrangle the loop!)

The world of these older electromechanical arcade games is seriously cool; so much history and ingenuity here that I hadn't been aware of!

Ah, cool, I wasn’t aware of arcade machines using recorded video before the brief laser disc era! If you have any references or links where I could read up about them, I’d be fascinated to learn more!

From the description, I'm guessing that the game your father played was "Firefox", a 1983 laser disc game by Atari.

The game synced dynamic computer graphics on top of pre-recorded laser disc background video. The "somehow switch to footage of the enemy aircraft crashing and burning" was, unsurprisingly, accomplished by simply jumping to a different track on the laserdisc, and typically happened when you completed one level and were about to begin the next.

I've never played it myself, but the gameplay actually looks surprisingly similar to Atari's vector graphic "Star Wars" game.

Or maybe not surprisingly once you notice that Firefox was created by the same team who had made that Star Wars arcade game the year before, and would go on to make the also-similar Empire Strikes Back arcade game the following year. I guess if you've got a winning formula, don't mess with it too much!

I remember once standing at a booth at the MacWorld Expo looking up at the massive display of a newly released game "Jump Raven" in which you were tasked with flying a hovercraft through city streets, shooting baddies, and all that usual video game stuff. I commented to the random guy standing next to me, "It looks a bit like Apache Strike, doesn't it?", referring to an old helicopter sim from like eight years earlier which had used a very similar viewpoint and game mechanics, albeit with only fairly primitive black and white graphics. And that random guy standing next to me turned and said, "I'm allowed to plagiarize from myself". Oops! (Sorry, Bill, I hadn't realised you were the programmer on Jump Raven too!) :D

It's kind of impressive how rapidly and how many rote processes built up around what began as "[We value] individuals and interactions over processes and tools"

I have started my own company and made a job that I'm passionate about and I think my friends would laugh at anybody who claimed that I wasn't burnt out basically all the time. ;)

Worth noting that while Hypercard was the glue which held everything together, most of Myst's functionality was implemented in native plugins. (or at least that's what Robyn and Rand told me when I chatted with them at a MacWorld Expo around its release). Their earlier games (The Manhole, Cosmic Osmo, etc) had been much more pure-Hypercard affairs.

Interestingly, for Riven (Myst's sequel) they were still using Hypercard to author the game, but the game no longer ran through Hypercard; instead, it ran on a custom C++ engine which ran on the data files output by the Hypercard authoring tools.

I agree that I've never had a problem with terminal speeds on Linux. If you're using urxvt, that's probably going to do just fine.

Where I have had problems has been on the Mac, where the system default "Terminal.app" or popular alternatives like "iTerm2.app" can each be catastrophically slow if you have a lot of control codes (as in, for example, rapidly paging through a large document in vim with an intensive color scheme active), and it could just take noticeable fractions of a second to redraw a fullscreen terminal window.

Moving to a faster terminal emulator like alacritty or kitty did make a good quality of life improvement for me in that specific use case.

It's.. kinda complicated?

"Craftsman" was never a company. It was a trademark under which Sears sold tools which they hired various tool manufacturers to produce for them.

Black and Decker bought the trademark from Sears in 2017, but don't use the trademark themselves; instead, they licensed the name right back to Sears for 15 years for free (that license will expire in 2032), and Sears (now Sears/Kmart) is still using it the way they used to; getting tool manufacturers to produce tools for them and selling them to consumers under the Craftsman badge.

So it depends on how you want to look at it? Sears never actually made the tools themselves, and don't own the trademark any more, but they're still the ones using the trademark and they're still selling tools made in the same way as they did before. So.. nothing's really changed? Except that the ownership of the name has been transferred to a different company.

Are you stating that this specific machine actually does this, or are you just theorycrafting that a hypothetical machine could do this?

Good point, yes!

Worth noting that there's an equivalent of epoll on most platforms.

On Windows there's IOCP, on Mac and BSD-derivates there's kqueue, and Linux has epoll, but to a first approximation they all do basically the same thing; instead of giving you a full-sized array of "active or not" results that you have to iterate across to detect activity on each of your sockets (as you get from the standard berkeley sockets 'select' and 'poll' APIs), they only inform you about the sockets that are actually active, so you can spend less CPU time iterating through that big results array.

I can say from personal experience that when you've got a single server process that's monitoring 25,000 sockets on a Pentium Pro 200mhz box, it makes a huge difference!

I'm a little surprised (and maybe skeptical) that it'd make a noticeable difference for the much smaller number of sockets that your average web server would be using, but.. maybe?

Sometimes. It depends on the UI requirement for the game.

I have seen some extremely-UI-heavy games which do exactly what you say; building a whole windowing environment entirely from scratch. (I'm working on a game that's doing that right now!).

But most major game engines provide a basic UI toolkit if your needs aren't too intense, and those are often more than enough for most games and are a whole heap less effort; they can typically be themed so that your UI doesn't look like everybody else's UI, and if all you need is a single window with some widgets and buttons and maybe a text field, that's probably going to be more than enough for you. Like, 95% of games would probably be totally fine with just that.

Maybe five years back I worked on and adjacent to a couple large games that were building their UI in an embedded copy of Flash (which wasn't as dead as people think!) Not sure whether large games are still doing that now, but it wouldn't surprise me if there are still some holdouts building their game UIs entirely in Flash!

If you go far enough back, everything is nothing.

Hobbyists who are sponsored by NordVPN. ;)

But there is no home button on Nintendo's first-party NES or SNES-style controllers, which I think are the ones more directly comparable to what we're talking about here.

(Not sure about the N64-style controllers; I've never actually seen one of those in the wild)

Since you're bringing up "capturing the data", it's worth pointing at Kagi's privacy policy, the very first point of which is this:

* Searches are anonymous and private to you. Kagi does not log and associate searches with an account.

https://kagi.com/privacy

Additionally the search page itself displays this notice:

Your searches are always private. We do not see them and they are not associated with your account

The "captured" user data you seem to be concerned about doesn't appear to actually exist in this case; the data is quite explicitly not being captured.

Describing something as "super good enough" is an amazing turn of phrase and I'm totally going to use it in future.

Zaphod: It’s a carbon copy of the Ravenous Bugblatter Beast of Traal - or I’m a Vogon’s Grandmother!

Arthur: The Ravenous Bugblatter Beast of Traal? Is it safe?

Ford: Oh Yes! It’s perfectly safe - it’s just us who are in trouble.

- Hitchhiker's Guide to the Galaxy, Fit the Sixth

I'd also like to highlight "non-hostile" as a reason why folks might prefer ChatGPT.

Stack Overflow has a lot of stridently opinionated jerks contributing to it, and if I can just ask ChatGPT a question and get an answer that works rather than having to deal with being belittled by those people, then I'm probably having a much better day as a result.