

The squidgy gel look Vista/7 had was pretty nice, too.


The squidgy gel look Vista/7 had was pretty nice, too.


They’d arguably stopped some time ago. I have a Thinkpad T490s, and a fair chunk of that isn’t upgradeable without swapping a fair bit of the body.
The keyboard, for example, is a permanent part of the chassis. Replacing out requires you to swap the entire shell out.
The Ethernet port is some proprietary gubbins, because Lenovo wanted to be funny, and use the same protocols and pinouts as regular Ethernet, but used a special physical connector.
Half the RAM is also permanently soldered into the motherboard as well, so you can’t properly upgrade that either.


I have no idea who thought it would be a good idea to give an error code to a user in Hexadecimal form, with no other information.
An error occurred: 0x 80070003
is hardly helpful at all.


It is, as is the native powershell terminal, thankfully.
I don’t think it was. It was them putting the triage tag on themselves that seems to have put it into a loop, at least going by the progression of events.
The user put on the tag, the bot removes the tag, because only maintainers can put that on, and then puts on the tag, after which it removes it because someone who wasn’t a maintainer put the tag on.
I could see exit being an issue if you’re doing something in the CLI that might cause you to type that without passing it as a command (“What do you want the program to do after the task is complete?” “Exit”), but I’m not familiar enough with the program to say one way or the other.
In fairness, it is off-topic, since a lot of it is more commentary about AI, rather than talking about the repo, or the issue. The only comment that could arguably be relevant is the person saying that the user could also use CTRL-D to exit the program.
The repo might be for an AI tool, but I’m fairly sure that the bot isn’t itself LLM-powered. It’s just your basic generic bot.
From what I can tell, they want the program to close when you write exit instead of /exit. Guessing it currently does the latter, and does a “did you mean /exit” sort of thing.
In fairness, this isn’t an LLM issue, but a poorly made bot issue. An old fashioned bot would be equally vulnerable to doing it, assuming it isn’t one.
Burger. Especially if it’s not too tall, and thus a perfect height for eating with.
Sometimes, you just need a flat food structure.


Though the check isn’t very sophisticated, if memory serves. It more or less checks whether / is passed to rm -r.
If you did something like rm -r $VAR/*, but didn’t check to make sure that $VAR was set and not empty, it could still fire, since rm wouldn’t see that root got passed, only a bunch of directories in root.
The personal project is a matter of personal pride, whereas for work, any old thing will do, as long as it meets the requirements.
Del is files, Rmdir is directories.
Running del on folders just leaves an empty tree.
Thing go up instead of down.
It’s Google’s version of an IDE with AI integrated, where you type a bit of code, and get Bard to fill stuff in.
It would at least be a little more understandable, what with the whole aborting, terminating, or killing children before the parents to prevent zombies.
Brain and Limb.
Being able to just enter a partial command, and hit [up] to jump to prior commands that started in the same way in zsh is a godsend.
:x is also an alternative to save and quit.
Equally valid for the facial expression you’d make upon finding that out.
Cloudflare’s is a corporate solution from the company that man-in-the-middles half the internet and makes me click shit every fucking time. I see it whenever I make the mistake of following a stackoverflow link.
It is also not very useful if you don’t use a PC. Every time I look up a Cloudflare-gated site on my iPad, I usually have to jump through a few captchas before it will let me in, if it doesn’t decide to be a grump and decide to put you in a sisyphean cycle of captchas, constantly refreshing without end.
Or if you use some software. I have citation software that gets stuck in the loop because Elsevier puts their journals behind a Cloudflare wall, and when it pops up the prompt to prove you’re not a bot, just refreshes straight into another prompt.
Pink is, after all, not a colour of light. In which case, it is entirely reasonable for a pink unicorn to also be invisible.