HNHacker News
TopNewBestAskShowJobs

fullstackchris

742 karma · joined February 28, 2018

Owner of Full Stack Craft fullstackcraft.com

fullstackchris.at.hn

meet.hn/city/at-Feldkirch

Interests: Freelancing, Investment, Mobile Development, Open Source, Programming, Remote Work, Running, Startups, Technology, Travel, UI/UX Design, Web Development, Writing, Philosophy, Outdoor Activities, Hacking, Fintech, Cycling, DevOps

---

wonderfuldev_cn693xi6wq7203nxugqijvyc

submissionscomments
fullstackchris··on Svelte is surprisingly easy to learn
if anyone is wondering why i am perma supporting react and combating seemingly random react haters all throughout this thread, i have to say that i'm exhausted from sneaky anti-react posts popping up approximately every 2 weeks at this point

theyre always exciting and fresh at first glance, but if you dig into any real engineering or large scale use cases, 99% of the arguments fall apart. reply to this comment if you have doubts, i'll share my decade+ experience with counter arguments

if you stop and think for just a moment, a single moment, why react is the most popular frontend framework...

fullstackchris··on Svelte is surprisingly easy to learn
> proven unstable

care to share a link, blog post, or study? i've built at this point dozens of apps on react, no "instabilities" to call home about

fullstackchris··on Svelte is surprisingly easy to learn
> everything gets replaced by AI

oh god, what would that mean for the web? an unreadable mangle of JS, HTML, and CSS, even when it's in source code format? count me out!

fullstackchris··on Svelte is surprisingly easy to learn
but there are literally libraries on top of libraries to immediately provide in react what svelte might provide "out of the box". the best thing too is that these hooks are all open source - so you can either install a package directly, or pick and choose and install whatever hook you need.

idk, i'm dying on the hill that react is unopinionated for a reason, i think it can be morphed and molded to so many use cases

fullstackchris··on Svelte is surprisingly easy to learn
I kinda find this a weird comment... you say productivity is way better than React but then go on to say you still sometimes need to resort to the Discord forum to figure out what is going on (!)

To be fair, React used to have a documentation problem as well, but at this point I think anything and everything under the sun is documented excruciatingly well.

Of course people will comment I'm shilling React, but, reading the top comment, its kind of like trying to shill Svelte but immediately coming up short.

fullstackchris··on AI isn’t good enough
AI won't work to reeingineer society, just like the last 100 "latest things" didnt. There is a massive undercurrent that always goes in the opposite direction of technology advancement: tradition!

You need new people growing up and living in a new paradigm to get these changes, you cant just ask (or expect) people who have lived their life one way to change completely to some other way of living

fullstackchris··on GPT-3.5 Turbo fine-tuning and API updates
but how is the speed here? does it feel fast "enough"?

looking into to running llama on prem / private cloud but i have no idea where to start in terms of sizing, do you have any details or posts on to what the minimum / recommended hardware requirements are?

EDIT: just looked myself, not as encouraging as I'd like: "For good results, you should have at least 10GB VRAM at a minimum for the 7B model, though you can sometimes see success with 8GB VRAM. The 13B model can run on GPUs like the RTX 3090 and RTX 4090"

definitely borderline dealbreaking for solo hackers / small teams

fullstackchris··on Upcoming .com and .xyz domain price increase
Wow! Porkbun! I nearly forgot about that place!

Shamelessly been using Namecheap for a while now... their UI is a bit old school but they have some of the best prices around (at least did)

fullstackchris··on Structuring Your Infrastructure as Code
if you're talking yaml... you could argue that config file is also code, just a different language
fullstackchris··on Things you forgot (or never knew) because of React
At some point I'm really going to try to build my next SaaS app with Svelte. I'll write an equally dismissive post as soon as I run into the first roadblock, which I guarantee won't take long. I need big, bad, and fully customizable frontend/backend frameworks for building some giant static websites!

Let's take a look at the frameworks available for this...

Gatsby > 50K stars Next.js > 110K stars (!) Remix > 24K stars (impressive, only 2 years old!)

and looks like Svelte has... ultimately... Sveltekit. Svelte has Sveltkit:

Sveltekit > 15K stars

React is great; not because of React itself but the massive ecosystem and number of frameworks behind it. There isn't a component or hook under the sun you can't find.

I grow weary of such posts, still waiting for an objective truth as to what really differentiates a frontend framework and makes it so much better than all others; still haven't seen it.

fullstackchris··on Michael Burry shorts Major index in sizable bets
that's what everyone said in 99 when nasdaq basically doubled in a year... then continued to more than double the year after that! a lot of guys blew up being too early trying to guess when the selloff would finally happen...

they're called bulls for a reason - they're not the brightest, but you gotta be sure to either join 'em or get the heck out of the way while they're in control!

fullstackchris··on Backward Compatibility, Go 1.21, and Go 2
darn i was hoping for at least a hint or mention when Go 2 might be released... no dice
fullstackchris··on The Carrot Problem
"willingness to take risks" and betting the farm on a tech stock _continuosly_ over 20+ years are two very different things
fullstackchris··on Numbers every LLM Developer should know
I'm curious about the point on the embedding lookup cost... in my experience for an embedding lookup to be accurate, you have to include your entire document dataset to be queried against... obviously this can be just as expensive as querying a full cloud model if your dataset is very large. Interested if anyone had thoughts about this.
fullstackchris··on Show HN: Easyful – A Free Gumroad Alternative
Definitely noticed that too. Seems like a cheeky tip of the hat towards Gumroad.
fullstackchris··on We now believe that the game is over. LK-99 is not a superconductor
Definitely not. This definitely sounds like recency bias. I'll bet 9/10 of your non hacker / engineering family/friends don't even know what LK-99 is. I know thats the case for me at least.
fullstackchris··on Show HN: Axilla – Open-source TypeScript framework for LLM apps
I'm no expert in actual ML implementations but I was under the impression that Python (i.e Tensorflow) is actually C/C++ based under the hood. I just meant I can't imagine the V8 engine can be as performant for all that matrix math in those models.

But now that I'm looking at the actual code samples, I'm not even sure JavaScript is doing any of the actual heavy lifting? (I see you use OpenAI's embedding) so this tool is more of the glue connecting all the parts? Again, I'm out of my wheelhouse here.

fullstackchris··on Show HN: Axilla – Open-source TypeScript framework for LLM apps
I'm gonna be that guy who will probably show up sooner or later anyway, but... I can't imagine performance can compete with other languages? What were your findings or experience with that?

Still, I'm a huge fan of TypeScript and will give it a try anyway :)

fullstackchris··on LK-99 is an online sensation but replication efforts fall short
haha, now even nature has to have a commentary. a bit hypocritical title eh?
fullstackchris··on LK-99: The live online race for a room-temperature superconductor
gotta say, this is slowly looking like a giant nothing burger
fullstackchris··on Ice core scientists in East Greenland reach bedrock
isnt that the case with any advanced technology? (including wind and solar)
fullstackchris··on Ask HN: Is the market bad, or am I having the worst luck job hunting?
Can't help but feel this is often the case. A bit of "oh they are not desperate enough yet to join us"
fullstackchris··on Ask HN: Is the market bad, or am I having the worst luck job hunting?
Applying and interviewing for software jobs has become possibly THE most toxic thing I've ever experienced:

- ghosted by companies if you don't EXACTLY match and 'gel' with their stupid full stack takehome projects to the T

- multiple cases of simply no response

- hours of interview rounds taking my time away, when it's clear they won't take you

The list goes on, there is truly nothing I hate more.

fullstackchris··on Ask HN: Is the market bad, or am I having the worst luck job hunting?
I don't know, why should we have to be "motivated" to get a job, when the company can just get rid of us at any time? What's the point?

Seems like a double standard.

fullstackchris··on How to build toxic software teams
Like the clowns I used to work for, cancelling daily standup with 5 minutes notice or just not showing up at all

its really the little things of being professional

fullstackchris··on Is technical analysis just stock market astrology?
So... a lot of smug comments on both sides claiming what technical analysis is and isn't. Let's get something straight, there are only two ways of analysing markets: fundamentally or technically.

I've been actively trading nearly every day for the past 6 months and can tell you there are people who extract profits from the market every single day (I'm not one of them yet, but I'm at least modestly profitable the past few months). Being profitable every day can ONLY be done using technical analysis. You're not going to profit every single day doing intraday trades on fundamentals. People think 'technical analysis' is as simple as what this blog post mentions - using something as rudimentary as SMAs. Unfortunately, this is case for most people. Come back when:

- you know what auction market theory is (AMT)

- you know what one time framing is (OTF)

- you know what delta divergence is

- you know what poor market structure / single prints / excess all are

- you know what low / high volume reversals are

All these are NOT fancy indicators or contrived signals but are quite literally real structures and residuals of actual BUYERS and SELLERS in a market (i.e. exactly what a market is, NOT astrology!) that can help show where the markets MAY go (of course the future is never certain)

Further more, you have to synthesize the context of all of these things in real time as the session unfolds and block out any of the emotions or psychology of putting money on the table entails! To get to this top tier (being profitable nearly every day) is as hard and takes as much time as any other performance profession (athletes, sales, etc.)

Of course you're going to have an extremely bad time if you try and implement something like a SMA crossover!

Think of it this way: any automated technical strategy you can think up within even a few days of effort have been done over 1000 times by every financial firm on the planet - and you can guess they don't work - if they did there would be a lot more billionaires on the planet! Most of the automated strategies that are actually profitable are designed by engineers with the help of decades+ traders who have been consistently profitable. Even then, you'll be lucky if in a raw win/loss ratio is over 50% (but this gets into expected value of the trade, etc. too long to post about here, you can be profitable with just a 30% win rate, for example if you're reward to risk is 2:1)

Anyway, thats my 2 cents from someone who trades with real money on the line. It's probably pretty easy to say that TA is made up and use it as a scapegoat when you YOLOed a bunch of money on GME and lost it a few years back.

fullstackchris··on Don't Take VC Funding – It Will Destroy Your Company
I think it falls into a few categories: and you can _feel_ the difference. There is type one: the company just going through the motions because they're just the NPC that feels "oh I now its time to 'do' VC because that's what everyone does"

Then another type: some of the deep tech startups full of super smart people, even publishing papers, and are usually backed by one or more major institutions (universities, companies, etc.) you can literally feel the passion pouring out of their employees

Then even still there is a third but very rare type: that startup that bootstrapped itself to profitability without any VC at all! (IMO most impressive and difficult)

fullstackchris··on Definition for Open Weights Licensing
This is a great idea, but I'm thinking weights aren't enough. If the recent rumors are true of the latest models like chat GPT-4, for example - they have a realitively new style of overlapping _a cohort_ of models, not just utilizing a giant matrix of weights (as has hereforto been the defacto architecture) - or is this still the case even for chat gpt 4 after training?

I'm no expert in ML, so perhaps I am ignorant as to what information is truly needed to reproduce results. I just cant help but think you'd need some hints at the model architecture itself as well.

fullstackchris··on The Concise TypeScript Book
yeah I took a look and couldnt help but feeling like it was a laundry list with minimal examples and not always telling you why or when you would use a certain feature of TypeScript. Example: TypeScript _does_ enable you to overload functions, but overloading functions is nearly never the best way to do something, often cited as a spaghetti code pattern
fullstackchris··on Ask HN: Problems for the next decade?
learned about the Wendelstein X-7 this weekend (and stellarators in general) and they've accomplished some pretty impressive acheivements... i think there are many breakthroughs and things to be done in this space
← PreviousPage 11 of 20Next →