0x008

In my humble opinion the sound of the headphones is just preference. Of course some people will tell you the arya is great or the hd800s or whatever, but in the end it just is a sound preference. Do you prefer airy highs or warm mids or whatever the terminology is.

When it comes to DACs and amps its just simple: get the one that has the best signal to noise ratio. Usually they should not alter the sound profile at all.

Also most DACs you will find on audiosciencereview for example have such a good signal to noise ratio (>100 or whatever) that is is imperceptible to the human ear anyway. You can go higher, but you won't hear a difference. And some of these cost less then 100$ or even less.

I would go for something that looks ok, is not too large and fulfills your user experience preferences.

In my example I bought separate DAC and AMP to have the best signal to noise ratio, but I have more cables and 2 boxes on the desk and the AMP is huge. I would rather go back for an integrated solution that maybe can also drive my monitors. But here we are...

it's not all about perfect measurements.

then, of course, if you buy into all the snake oil it's a different story ;-)

what the fuck that USB-C Dongle has great Signal to noise ratio. wow. probably because it runs on battery?

the biggest problem is stress. Once you get into the stress treadmill every hour feels like 5.

It is totally fine to be placed on the backburner -- but then you should get backburner hourly rates and not pay the same as the big clients.

Because they are spending all their money on data scientists.

What is the benefit of having a terminal float on top of your existing window as opposed to just switching to another window and switching back?

I don't really get why you would bring a floating terminal into your "main coding screen" to do something in the background. What problem does it solve compared to just switching to another window, running the command, and and switching back?

Also in tmux you can just attach a existing window as a split to your current window and then pop it out again using aliases/bindings.

You can use the same hotkeys to switcch between windows and panes that you use on the ssh box :=)

danswer has been referenced many times on hn already. I personally am not really a fan of seeing repeating ads for the same thing here.

GPT3.5 is the most popular MoE model probably.

Just a great time.

you could try to use ranger or a similar terminal filemanager. it's basically like nvimtree or nerdtree

I find it hard to read with those AI-generated summaries unfortunately. But I appreciate the effort of collecting some high-quality links for us.

Makes sense. Probably on stackoverflow or in forums people tend do give more effort in their answers when the questions are polite.

What is this title? It's a pull request that contains a great investigation which outlines the effects which various mixes of quantization have in one metric. The article draws no conclusion and the title is "k-quants".

I like https://haxplore.com/

Well, now they made sure it is impossible to control that underaged teenagers can continue using that stuff.

But the question is rather of the cost are peanuts compared to expected cash flow for the next 12 months.

Google and Facebook are not interested in pooling all their resources in order to build the next big thing. They are just interested in doing “enough” so that people keep using their platforms. The race is about how often a day every person on the planet spends on either google or Facebook/instagram. It’s about who is “the homepage of the internet”. They just need to be good enough so that traffic doesn’t move off to chat gpt.

Because the cost of entry to the market is so absurdly high right now, it is seen as a good investment opportunity. If you throw enough money at it you can make your place no nice and early and then win later by sheer experience in the field. That is the idea.

I don’t think you can really compare that. The streaming service market is not elastic because you cannot easily interchange one series for another.

If other vendors LLMs become good enough it will actually be easily to interchange and then the race for the best UX and integration will be upon us (which the other commenter alluded to).

I think Meta (facebook) is already working with android phone manufacturers to include chips with specific instruction sets to support the llama architecture.

probably intel is working on the same. I assume the next generation of mobile chips (laptop or phone) will all include special LLM processors.

continue.dev can complete code as well, but it is not a mere code completion plugin like copilot, it is more like an interactive chatbot where you can edit and talk about parts of the code which you have selected.

for mere code completion there is this: https://github.com/FarisHijazi/PrivateGitHubCopilot

but I haven't tested it.

I think my critique to the article would then be that there is not distinction between actual objective utility of the solution (I.e. expected outcome) versus the own standards and fear of being not good enough (which is totally unrelated to the actual utility or greatness of the solution).

Well not really what I mean. I mostly have faith I create a pleasant outcome, but the way to get there is painful.

I feel the simple equation fails to account for that.

There is nothing new in the article, it is basic motivational “science”.

However, I feel this does not apply for me and greatly misses one of the main aspects of procrastination: the fear of disappointment. Can I fulfill my parents expectations? Can I fulfill my own expectations? Will I fail? Will it hurt to face my own demons and my own fears of not succeeding?

Am I just different? I thought this was what procrastination is about for everyone.

until you need a dependency…

What is so horrible about it? Our devs are happy they can focus on fixing unit tests in pycharm, while others care about where the code is going.

Overpass is great, thanks for sharing that.

This looks really cool and I am looking forward to use the same file manager on all OSes I am using, but I fail to see the relevance as a Mac OS user. Finder is really good and has the great builtin preview feature, it can connect to clouds just fine etc.

Only thing from the Roadmap in spacedrive I see, which looks really interesting, is encrypting on the fly using stored keys. A file Manager with built in truecrypt like functionality would be great, let's see how they implement it.

Workers also sounds nice, I would like to see multithreaded file operations. Maybe they can followup with wildcard operations on the filesystem.

But in general it feels like they are trying to cram a lot of (already existing) different tools into one. I am not rally seeing the value proposition here.