sprobertson

Looks like a great resource. Maybe include a "last updated" so users can know if the site has gone stale?

Edit: Nevermind you're a step ahead - I was looking on the individual items but a global one makes sense.

I was contacted by someone in another country and we agreed to do it on an escrow site. Worked out fine, made some money (tbh probably not much given how many years I had held on to it). I have not tried purposely selling on other sites.

Depends on your monitor? It's always been #ff6600

Japan is also very small which makes it easy to "understand"

Every website has the ability to know which country you live in, likely which city, just using your IP address. Use a VPN if you're concerned.

So a git log?

imo linear algebra is the main thing to refresh yourself on, as most of NN is based on matrix multiplication. The calculus is mostly in the gradient descent which in practice is abstracted away, so having a few decades old understanding might be good enough. For practical implementation Andrej Karpathy has a good series of videos going over a lot of the concepts starting here https://www.youtube.com/watch?v=VMj-3S1tku0

Simply using a calmer custom alarm (some soft music that slowly builds) before the gunshot alarm

It might just be the word "like" said with a certain affect, I've seen it glow in the wrong context a few times.

The M in MANGA is an update from Facebook to Meta

This is great, I could see myself using it daily. The only hindrance I've found so far is navigating large responses. Would be cool to have some way to collapse chunks of JSON (a la https://github.com/antonmedv/fx), or even just more vim key navigation, like G/gg, %, {/}, and search.

I like the micro-adjustment on that first one. It's still limited (I assume by frames?) but easier for me to use than dragging to get an exact frame.

I feel like I learned everything except the question in the title. I still don't know why it's long, just a lot of facts about how long it is.

More like comparing Instagram to Signal

Where would one find more?

The maximize button has an option to go into a split screen mode where you get two "half full" screens (i.e. side by side but still hide the menu bar, window chrome).

It's the exact same SVG with two stroke colors changed. I believe this is known as "stealing"

https://pocketbase.io/images/logo.svg

I think that's Grant's point - there's a (potentially) false dichotomy in those descriptions.

Did you read past the title? The main point of the article is a syntax for knowledge bases - plain text is just an implementation detail.

Well done. Matching the HN style with Tailwind is a nice touch.

One thing I've been wanting for with search in general, especially LLM-powered, is having some kind of date relevance - especially with the fast moving world of technology.

For example I want to know how ProductX and ProductY compare. Last year ProductY didn't have FeatureZ, but they implemented and announced it last month. There might be several comments lamenting the lack of FeatureZ from 2 months ago, but they shouldn't be considered with the same weight now that it does exist.

I don't have any ideas for how this should be done but it's something I'd like to see tackled in RAG systems in general, and wanted to put it out there.

Do a spell check on your 3 sentences haha, you have "wesbite" in the instructions

What? This is a paper describing a system/technique that they happen to call ZenDB.

This is the most technically correct answer - language models are at their core trying to continue a string of text. If you had trained the model on knowing exactly which letters represented the other letters in this particular puzzle, it might work.

On the other hand it seems like LLMs' inability to generate perfect boilerplate, requiring me to go through and fix all the bugs and nits, is all that's keeping my "sharpness" at this point.

I think that is both the basic point of and minimizes the interesting point of the website - the actual element doesn't matter, it's the things you can do with them.

> engineers who poured hours of their lives into building this

Not to mar these specific engineers, but that's an empty phrase that can be said about anything ever built. It doesn't somehow make the idea or implementation good.

Couldn't build up to doesn't mean they didn't exist... it took me a while to make a "Sandwich". But after spending enough (too much) time with it, I was able to construct some things that show it's not disappointingly finite:

* "Bollygraff"

* "Teen Mom 12"

* "Hackimus Prime"

* "Billionaire King + Sushi Burger Venus" (sic)

One potentially frustrating part is that some things turn out to be black holes... for example, combine "Trump" with anything and it tends to return "Trump". There are also plenty of loops, and some that refuse to combine at all.

I've found retrying after a bit will often return the new result. I suspect the frontend is timing out before the backend comes up with a result, which is eventually cached by the time of the second attempt

Not that interesting if you look at town density https://www.visualcapitalist.com/wp-content/uploads/2018/10/...

I've been working on something like this but it's of course harder than it sounds, mostly due to how few example use cases there are. A dumb false positive for yours might be "you tend to turn off the lights when the outside temperature is 50º"

Anyone know of a database of generic automations to train on?