What is the point being made here? Some past technologies were overhyped, therefore AI is overhyped? Well, some past consumer technologies did change the world (smartphones, texting, video streaming, dating apps, online shopping, etc), so where's the argument that AI doesn't belong to this second group?
Also, every single close friend of mine makes some use of LLMs, while none of them used any the overhyped technologies listed. So you need a specially strong argument to group them together.
What a fantastic list. I'll be saving it to show the junior developers.
My only nitpick is that "reliability" should have been a point by itself. All the other "ilities" can be appropriately sacrificed in some context, but I've never seen unreliable software being praised for its code quality.
Which is part of why LLMs are so frustrating. They're extremely useful and extremely unreliable.
I agree with the general message, but I'm curious what ingredients go in your 800 calorie sandwich. That's more than a double Big Mac with 4 patties (780 kcal)!
The meeting notes in the repo was a nice surprise. Overall looked great, striking a good balance.
.input {$var :number maximumFractionDigits=0}
.local $var2 = {$var :number maximumFractionDigits=2}
.match $var2
0 {{The selector can apply a different function to {$var} for the purposes of selection}}
* {{A placeholder in a pattern can apply a different function to {$var :number maximumFractionDigits=3}}}
Oof, that's a programming language already. And new syntax to be inevitably iterated on. I feel like we have too many of those already, from Python f-strings to template engines.
I wish it'll at least stay small: no nesting, no plugins, no looping, no operators, no side effects or calls to external functions (see Log4J).
As a happy Bazzite user, I had no idea things were so bumpy. At least the migration to other os-tree distros is trivial (Fedora Kinoite -> Bazzite was one or two shell commands). My main reason for using the distro was the built-in nvidia drivers for my old graphics card.
I think it's also appropriate to use it when the rule is so strong that exceptions are famous because they are exceptions. "Birds are capable of flight" is strong enough that penguins and ostriches are famous for being counterexamples.
But that's not following the saying - it's still not proving, it's modifying the rule. It shifts the rule from "birds can fly" to "most birds can fly". Pointing out that penguins can't fly doesn't make the case that birds can fly stronger in any way.
You're right in a strict sense. But in my experience such strictness is only useful in hard sciences and (maybe) legalese. There are exceedingly few things we can claim to apply everywhere, and even fewer we can "prove" to each other.
Give it a try if you don't believe me. Even categories we take for granted, like trees and fish, are not perfectly crisp, and "obvious" facts like "humans need a heart to live" have surprising exceptions.
> Pointing out that penguins can't fly doesn't make the case that birds can fly stronger in any way.
I disagree. It's such a common rule that there's a long Wikipedia page for the exceptions[1], and the first photo is of penguins, labelled "penguins are a well-known example of flightless birds.".
If I knew nothing else about the topic, I would take it as evidence that it's common for birds to fly, otherwise that fact would have been unremarkable. Not hard proof of a universal quantifier, but a useful rule nonetheless.
> There are exceedingly few things we can claim to apply everywhere, and even fewer we can "prove" to each other.
Yes, this is why hard and fast rules don't make sense, and why they should have "generally", "normally", or "mostly" attached to them.
If you have two categories of birds, one with those that fly and one that doesn't, having that second list doesn't make the first stronger. At some point that second list dilutes that first one so much that it doesn't make sense anymore.
If my rule is that "white guys are named Dave" does my building a list of every example of a Dave and non-Dave make my rule stronger? When does the "strong" nature of the rule get watered down sufficiently? Honestly, a list of hundreds of birds tells me that it's a weak rule and that the "birds fly" rule is wrong.
Good, KYC exists for a reason. Why does AI need to open an account, anyway? Just give it a debit card with a limit, not a whole new account and contract with a bank.
Those are much easier problems to solve, and surely already solved by some fintechs, than bringing cryptocurrencies to the minimum legal compliance and meeting performance requirements.
Random drafting is only part of the problem. My biggest gripe was the lack of in-run saving. When a run can take an hour, that's just disrespectful in a single player offline game.
And another (smaller) issue is the step limit. End game has you running to pretty far away areas. The walking itself quickly gets old, but you sometimes waste the entire run because you didn't have enough steps for the required back-and-forth.
That being said, I greatly enjoyed how note-taking was rewarded. By the end I had over 600 screenshots organized in different folders.
Mid-run save is a legitimate gripe. Practically the reason it doesn't have that is that the internal game state is probably horrendous and restoring it would be a nightmare. The engineering in Blue Prince is terrible. Imagine the terrible spaghetti you've seen from self-taught programmers, now, imagine nobody more senior is in charge and remember it's a video game so there's time pressure. So yeah, that's a quality issue, and definitely a fair gripe, I could imagine a hypothetical "fixed" version where this works.
The step limit is an important resource. There's a reason one of the early goals of the game (in Bequest and to some extent Dare modes) is to have more steps at the start of each day and why an important penalty of Curse mode is that you only have 13 steps. As with other resources like keys, you can learn to make better use of what you have and also how to get more of it within reason. I don't think it's as good of a game without Steps. They're not (outside Curse mode) scarce enough to commonly end a run, but they matter.
Drafting is already a limited resource, the step limit feels like a hat-on-a-hat. The early game could limit on available draft pulls without the step limit.
But also, I am one of the people that the drafting mechanic directly conflicted with my interest in progressing the game. That lack of being able to focus on a particular thread of my choice affected my interest in the game. I didn't want to juggle every thread all at once, especially without knowing which threads are the most interesting to pull ahead of time.
Drafting doesn't cost steps. That is an important early game realisation.
I think this sort of "If I just keep banging my head against it, then it will break" attitude is a problem and Blue Prince was a much nicer experience for discouraging that but of course each person is different.
Yeah, I noticed the engineering issues. One of the few PlayStation games I've played where crashing was a regular occurrence, which made the lack of saves even more infuriating.
I wouldn't suggest removing steps entirely, but maybe something softer than abruptly ending the day. After exhausting my steps, let me walk around without drafting rooms and picking up items, for example.
And the late game puzzle quality was very hit-and-miss for me. I loved the sigils, for example, and appreciated the permanent upgrades/changes. Other puzzles required putting disparate items/ideas together, but by then the game had expanded too much and it was unclear what paths were exhausted, still useful, or simply fluff, and the randomness made every check time-consuming.
The Reddit community of Outer Wilds maintains an extremely comprehensive list of recommendations, in the style of "if you liked this aspect of the game, try these other games":
It asks for a very different skill level than Outer Wilds, for sure.
On the other hand, it gets extra points for the in-game manual booklet. The mechanics, the metaphor, and the gorgeous execution should be required material in game design classes.
Yes, but I wouldn't be surprised if the change is rejected. The crypto library is very opinionated, you're also not allowed to configure the order of TLS cipher suites, for example.
That is a completely valid threat model analysis, though? "Just hope no bad guy ever gets into the safe" is rather the entire point of a safe. If you have a safe, in which you use the contents of the safe daily, does it make sense to lock everything inside the safe in 100 smaller safes in some kind of nesting doll scheme? Whatever marginal increase in security you might get by doing so is invalidated by the fact that you lose all utility of being able to use the things in the safe, and we already know that overburdensome security is counterproductive because if something is so secure that it becomes impossible to use, those security measures just get bypassed completely in the name of using the thing. At some level of security you have to have the freedom to use the thing you're securing. Anything that could keep a bad guy from doing anything ever would also keep the good guy, ie. you, from doing anything ever.
Perhaps figuratively? I manage several servers where the majority of (LDAP) accounts have no special privileges at all. They get their data in the directories and can launch processes as their user, that's...pretty much it.
Though the upstream comment is gone and I am perhaps missing some important context here.
When the question is "how do I communicate securely with a third party," there's nothing you can do if the third party in question gets possessed by a demon and turns evil. (Which is what happens if an attacker has root.)
Random sysadmins who have access to your server have the permissions to steal whatever is communicated between third parties unrelated to this sysadmin.
Just because some random outsourced nightshift dude has the permissions to do "sudo systemctl restart" shouldn't mean he gets to read all the secret credentials the service uses.
As it is now, the dude has full unfettered access to all credentials of all services on that machine.
Also, every single close friend of mine makes some use of LLMs, while none of them used any the overhyped technologies listed. So you need a specially strong argument to group them together.
reply