Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
tomaytotomato
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
4 ms
·
31.
▲
by
tomaytotomato
4mo ago
Not a criticism, but why would I use this instead of locking down my Claude using the allow/deny permissions list? "permissions": { "allow": [ "Bash(npm run lint)", "Bash(npm run
32.
▲
by
tomaytotomato
4mo ago
How useful is spatial data over time, does it decay or age much? Is the geographical data more useful, or are buildings and other structures more important? Genuinely don't know much in this space.
33.
▲
by
tomaytotomato
4mo ago
Agreed, I have been trying to get back into blogging on my site. I found LLMs good for spitting out a structure for my posts but if I ask it to write for me I end up just fixing or correcting it and then lose motivation and never finish the
34.
▲
Ask HN: The next evolutionary step in LLM usage?
3 points
by
tomaytotomato
4mo ago
|
2 comments
35.
▲
by
tomaytotomato
4mo ago
"You have 20 seconds to comply"
36.
▲
by
tomaytotomato
4mo ago
Anecdotal but I find people in the tech world hostile to ex or serving military, at least in the UK. It might just be the liberal bubble of London, but the sentiment I saw with some colleagues was. - "Why would I want to serve, I am no
37.
▲
Total War: Warhammer 40000 – Alpha Preview
(youtube.com)
3 points
by
tomaytotomato
4mo ago
|
0 comments
38.
▲
by
tomaytotomato
4mo ago
When Agentic browsers become the norm, surely we will go back to the days of super plain HTML pages?
39.
▲
by
tomaytotomato
4mo ago
Nice story, I don't personally agree with your choices but you are now back on the straight and narrow, big kudos to you and your wife.
40.
▲
Bringing a dead Spring Boot project back to life with Claude
(tomaytotomato.com)
3 points
by
tomaytotomato
5mo ago
|
1 comments
41.
▲
by
tomaytotomato
5mo ago
Always thought it would be cool to make an RTS or simple city builder game in the browser. Would you recommend this for hacking around or not?
42.
▲
by
tomaytotomato
5mo ago
That's cool, I just tested it out and it is fast but unfortunately its accuracy is not great.
43.
▲
by
tomaytotomato
5mo ago
LLMs cannot think, they just guess with some magic with the help of things like softmax ( https://en.wikipedia.org/wiki/Softmax_function ), that makes it seem intelligent. We aren't enslaved or at any risk just yet
44.
▲
by
tomaytotomato
5mo ago
Fair play to the chap, it was refreshing to read a scan of a letter typed from a typewriter. <joke> I just hope he doesn't start mailing packages to people in the tech industry in the next few years.</joke>
45.
▲
by
tomaytotomato
5mo ago
A UH60 Blackhawk with DEVGRU is heading to your location now </joke>
46.
▲
by
tomaytotomato
5mo ago
Yes as BenjiWiebe said. I was meaning more removing your reliance on Google in your life for search, photos, entertainment. You can start small with Duckduckgo etc.
47.
▲
by
tomaytotomato
5mo ago
The paranoia is completely warranted, but there is a solution. Just root your Android phone and put a custom ROM like LineageOS etc If you want a stretch goal try and de-Google yourself, I have tried but failed twice now.
48.
▲
by
tomaytotomato
5mo ago
Strong disagree. How many times have you written a boring CRUD endpoint that returns a list of results? Half the stuff I lose time on is the boring boilerplate code (despite using frameworks). Now I can focus on the fun stuff like determini
49.
▲
by
tomaytotomato
5mo ago
Indeed, more of a frambled egg. Lets see what happens in two years time.
50.
▲
by
tomaytotomato
5mo ago
This had to happen, for many reasons: - Its a throw thing at the wall and see what sticks situation - LLMs will improve* - Using LLMs in an agentic way will improve (git worktrees, sliced PRs, spec driven steps) So what happened here is a m
51.
▲
by
tomaytotomato
5mo ago
AI has always been dangerous, but not existentially dangerous. Mythos is dangerous but it's not going to Skynet us. Just the same as the military drone using some sort of OpenCV library and target prioritisation loop isn't going t
52.
▲
by
tomaytotomato
5mo ago
Congrats, its a small step in the right direction. The UK it seems has dropped the ball on the whole training and building models part, although we are punching up in other areas now. We really need to get our own equivalent to Mistral, and
53.
▲
by
tomaytotomato
5mo ago
I am a Java dev and Spring user for about 10 years now. If you want to learn how Spring framework and Spring boot works, the best thing to do is build your own library and then learn to add it to a new spring boot service. https:/
54.
▲
by
tomaytotomato
5mo ago
Looks like the repo is here - https://github.com/Classic7-Mod/WindowsOOBERecreation The repo is only 8 months old, which could be seen as good or bad.
55.
▲
RFV-0001: Request for Vibes
(github.com)
1 points
by
tomaytotomato
5mo ago
|
0 comments
56.
▲
RFV-0001: Request for Vibes
(tomaytotomato.com)
2 points
by
tomaytotomato
5mo ago
|
0 comments
57.
▲
by
tomaytotomato
5mo ago
Do Java engineer at Google not use IntelliJ?
58.
▲
by
tomaytotomato
5mo ago
As a senior dev I would say there is a risk of muscle memory atrophy, but the learning opportunities outweigh the risk. Today I learned about a more elegant helper method in Apache Commons' StringUtils library for Java. The function wa
59.
▲
by
tomaytotomato
6mo ago
I would recommend you consider a Claude code hook, you could then specify a particular part of the agent lifecycle to invoke the fortune as an alternative to manual invocation of the skill. You could do it on `SessionStart` to give a horosc
60.
▲
by
tomaytotomato
6mo ago
<tangent> Github actions is running like treacle now. Even when our company pays lots of money for cloud and private Github runners. I know its the go-to punchbag but I think enabling Copilot reviews globally for a large proportion of
More ›