jstx1

36 ms·

451.
by jstx1 4y ago
If the backend, frontend, API and data layer are in Elixir, what's actually written in Rust?
452.
by jstx1 4y ago
Coinbase is not a scam - the service they provide is the buying and selling of cryptocurrency which they do just fine. I don't like crypto either but calling everything related to it a scam is just intellectually lazy.
453.
by jstx1 4y ago
It means that nobody could have prevented the crypto market collapse from the position of a Coinbase Chief Product Officer.
454.
by jstx1 4y ago
> most employees are spineless and will fall in line. I don't get this. It's a layoff - what do you think not spineless and not falling in line looks like?
455.
by jstx1 4y ago
No, you can obviously do your best to be employable and all that, but nobody is recession proof.
456.
by jstx1 4y ago
This is creative writing bordering on a conspiracy theory. You're stringing up a bunch of claims without evidence, presenting them as fact, and hoping that your tone is authoritative enough for people to not think about it critically.
457.
by jstx1 4y ago
Instagram is built with Django.
458.
by jstx1 4y ago
> What about web developers? It depends on the company, not on the domain. OP might as well be a web developer, most developers are.
459.
by jstx1 4y ago
Even if you have the free time, I'm surprised that they managed to batch the different stages and have all interviews in the same time period. If I'm interviewing, at any given time I'll be sending out resumes, and talking to
460.
by jstx1 4y ago
Aero press with freshly ground specialty beans. The aeropress makes it very easy to get a consistently good cup of coffee and I definitely prefer it over a V60. I guess that falls under "manual filter"? You're also missing fr
461.
by jstx1 4y ago
> I need to see these "I take full responsibility" people actually punished. At the very least, this sentence needs to be removed from their mouths. There's some infantile obsession with this phrase in every layoff thread
462.
by jstx1 4y ago
One hypothesis is that the big tech companies are just better at sensing what the market is doing and reacting accordingly as quickly as they can, while companies from other industries are sleeping at the wheel. If that's true, we can
463.
by jstx1 4y ago
Do we have details on which geographies and teams are affected?
464.
by jstx1 4y ago
> The market was extremely quiet last week, but it seemed to pick up quite a bit this week. Are you just counting your inbound messages? Because there's no way you can personally tell what's happening on the market on a week-to
465.
by jstx1 4y ago
It might be because there are fewer companies that hire at your current level of compensation - a few international big tech companies and some trading firms - and all of those are clearly slowing / freezing hiring or laying people off
466.
by jstx1 4y ago
But all the other files, and the dependencies, and where is it going to run, and does it have permissions to see all the data and account resources it needs, and the testing, and...
467.
by jstx1 4y ago
Why use Pydantic instead of dataclasses?
468.
by jstx1 4y ago
That’s what I meant by state of the labor market.
469.
by jstx1 4y ago
It's not necessarily for performance, more often these layoffs are a way to get rid entire teams and you can't do that through natural attrition.
470.
by jstx1 4y ago
It's the usual - be as employable as you can, search for a job, apply, interview, get hired. I don't think that you need to do anything special just because the market is worse. When you start applying you'll find out how dif
471.
by jstx1 4y ago
> Is the current round of layoffs an admission that the "software will eat the world" scenario will not happen or an indication that it can be done with even fewer tech employees? Either statement is reading too much into it. I
472.
by jstx1 4y ago
> When I think about this moment in time, the start of 2023, it’s showtime – for our industry and for Microsoft. He could have left that part out.
473.
by jstx1 4y ago
I don't know what you mean about Docker - in my experience it runs just fine on M1.
474.
by jstx1 4y ago
Kubernetes expects these to exist by default right?
475.
by jstx1 4y ago
Why though? What specific thing does PyTorch do that TensorFlow doesn't do in a large monolithic Python application?
476.
by jstx1 4y ago
I don't get the more native part - you can call a TensorFlow or a PyTorch model and get results from either in Python - what kind of integration are you imagining where the internal implementation of the model makes a difference?
477.
by jstx1 4y ago
It's a big leap from "this isn't the most loved tool" to "dying a slow death". Plenty of companies use TensorFlow, some of them at huge scale, and they're getting stuff done with it just fine.
478.
by jstx1 4y ago
> What's the point of training bigger language models when you can't even serve (inference) them economically? They're built to impress people and push the boundaries of can be done with ML. Economies, accessibility, scale
479.
by jstx1 4y ago
Not enough process can be a pain too. "Just do this for us" without specific requirements, and then the thing you've built is never good enough and needs to be continually refined anyway or it's never used. Those environ
480.
by jstx1 4y ago
No, I don't think it is. Longer discussion here - https://news.ycombinator.com/item?id=33787508

More ›