Hacker Newsnew | past | comments | ask | show | jobs | submit | Imustaskforhelp's commentslogin

Damn, Not really the most user-privacy caring decision taking by the french courts here. It might put user privacy in geopardy.

Technically someone can run their own DNS servers who can be hosted on DMCA ignored lands (Moldova,Maybe China) but this feels like a net negative for the privacy. Beyond that, the most effective things become VPN for the end user but I wonder if blocking law might extend to VPN as well. The judgement from this certainly makes me feel a bit off-putting for VPN's.

It's very rare to see these companies fight for the good of the consumer (well technically they did it only because they wanted to save man-power/money but even then, it was pro user-privacy) and this has now been shut down :-/


Kulesh, You had made a previous post 1 hour ago about Tamp which linked to its website and now you have made this post which links to its github in even less than an hour of the previous post.

You are a member of this forum for a long time, I hope that you realize that HN rule of having a certain time limit till you can re-post a link is for a reason.

I appreciate it if you can please wait for the hackernews limits or atleast wait for sometime before doing this from now on. Good day.


Hey looks cool, https://type.a2n.dev/test?repo=ohmyzsh%2Fohmyzsh&mode=timed&... but I reloaded the website and I got this error:

Object not found! The requested URL was not found on this server. The link on the referring page seems to be wrong or outdated. Please inform the author of that page about the error.

If you think this is a server error, please contact the webmaster. Error 404 type.a2n.dev Apache


You can access the page directly from this url, I'm gonna fix that ! Thanks :)

> The speaker seems to be generated by AI

This is Patrick Boyle, he's not AI generated? Why did you feel like the speaker is AI generated?


And Boyle, IMO, is a great presenter, in part because he is so deadpan

I've watched Patrick's videos for enough years that I know he is not, but I still wonder from time to time. His voice is incredibly flat and uniform, he always uses fake backgrounds and there is extremely high use of jump cuts in his edits.

The jump cuts have been there since almost the beginning.

The community noticed he rarely blinks and he ran with the gag and edited all of them out.


I suspect that he at least uses AI for scripts. He tends to repeat the same thing worded slightly differently a few times.

Hey Michael, that's alright but can you perhaps edit the article to have all the facts clear out there in the manner that Markus has said.

There are times to be satirical, don't get me wrong, but those are usually when the dust is settled and maybe a reminiscence on the past.

Have a nice day and I hope that something positive comes out of all of it. I always believe that there are only few projects which get to the eyes of the general public enough to get funded, LibreOffice is one of the very few. People trust Libreoffice with donations and money to fight against Microsoft and show a path of freedom.

For the document foundation to betray the people who programmed the code in the first place, is also, a betrayal of the people who have funded libreoffice for years, who would love to demand more answers and I hope that in the article, that you can talk _effectively_ to them. It's really sad to see all of this happen and I wish if something happens as I don't wish for people to lose hope in open source foundations with cases like these.


I am vegetarian and unless I am missing something which can be very possible, but the only charts India is missing are when they are related to fishes and meat consumption. To then have, India in the final chart require >100% of land feels a bit unsusbtantiated to me because that could've been a possibility if most of India was non-vegetarian but India is one of the most vegetarian countries in the world.

https://en.wikipedia.org/wiki/Food_security_in_India

From as far as I can tell, India is decently self sufficient in food consumption but there are still people who lack food & nutrition because of a lack of access to affordable and healthy foods is widespread throughout India (Wikipedia)

Now India's fertilizer industry is also although impacted by the strait of hormuz being closed, for the most part, India has tried to make a neutral stance within the conflict and Iran has opened up a ship or two to India, so I suppose it might be the govt.s priority to get fertilizer to pass through to our nation.

"Essentially, even if countries are theoretically self-sufficient, in our current system they still rely on the global fertilizer trade. In other words: Self-sufficiency is still only as good as the supply chains that support it."

I sort of agree with this line but that is because fertilizer can be essentially treated as by-product of oil and oil production within the region that fertilizer has the supply-chain issue.

If I remember my lessons of economics right, Indian agriculture is still very much lower level and some/many own cows (dairy) so they can use the dung products of cows as a natural form of fertilizer as an alternative form

The issue becomes really evident, within large scale farms though.

Within India, Despite rising global prices, the federal government is continuing to sell urea and DAP at subsidised rates[0]. They are able to do so because they can tax us in other areas of our life and Indian govt. as a whole is now working towards getting more urea and they are saying that they have enough for this season even with conflict and are taking multiple measures.

As such, The food prices within India at the very least are going to be unaffected but this definitely is putting up a strain on the indian budget.

[0]: https://www.reuters.com/world/india/india-tapping-alternativ...


> Hey HN — I'm Fayner, creator of js-cookie (22B+ annual jsdelivr hits).

Hey Fayner! Nice to meet ya, I have signed up to Hutch, I think I might be one of the first 100 users so are you suggesting that I might have the product free forever or till a certain time?

> Then Pocket shut down. Then Omnivore. And the web kept filling with content nobody asked for.

I used to use Shiori until I one day had accidentally removed that docker container, nowadays, I either star if its a github project or I even had a tampermonkey script which created a button in my screen which I can click to store links in json format in a private secret gist of mine. Personally I really like shiori and I have talked to their developers on matrix too and they seem nice, I would love to hear your thoughts on Shiori.

> This is a genuine v1 — the advertised features work, everything else is roadmap. What I want to build next: a preference learning layer ("more like this / less like this"), Gmail integration for newsletters, and highlights. I don't know which matters most to the people who need this.

Personally, I would love if there was a thing similar to linkhut (you can see my linkhut at https://ln.ht/~imafh), this is a place where I share all the really cool things that I found on the internet and the best part absolutely about linkhut is that I can write "notes"/"text" within Linkhut about a particular link)

So anyone, who is reading my profile can exactly know why I vouched for a project. Combining this with tags and videos and search and profiles with it being open source/with api, I Really Love linkhut but (its link aggregators features are a little not so good)

I am not sure of how similar linkhut and hutch can be, but I would really love to see a feature for example where I can get suggested an article by other people who have written some genuine things about it in notes and then I can comment about it within maybe hutch itself even, or if not.

I feel like having these human elements gives a sense of community. I would love to hear your thoughts on as well!


Hello!

Notes and highlights are on the roadmap — you'll be able to annotate as you read.

The social discovery angle is interesting. Hutch is deliberately private for now but curated recommendations from real readers with genuine notes is a different thing from algorithmic suggestions. Worth thinking about.

On the free tier: the first 100 users get free access, no time limit.

I haven't tried Shiori, I'll take a look.

Welcome to Hutch!


> Worth thinking about.

Thanks, I have been thinking about it from a more human perspective as in that I have some cool things to share to people and those don't fall exactly into Hackernews at times (suppose a struggling musician who has an interesting story to tell), Usually I then use linkhut with their channels with a note so that people can know why I have bookmarked them in linkhut if they go to my profile and see my linkhut there. Its definitely really interesting to think about.

> Welcome to Hutch!

Thanks! Happy to be a user here!


Thank you! Feel free to send an email at any time hutch@hutch-app.com if you have any other suggestions/feedback :)

I am now waiting for, Claude Code rewritten in assembly project :)

(Maybe it can be a good project for anyone who works low-level)


we’re going to have a bunch of CC rewritten in <obscure language>

Waiting for Rust.

Claw-code (reached 100k stars on github and I have made an article about them[0]) are currently re-writing claude-code in rust (they have already ported it to python)

Its worth following: https://github.com/instructkr/claw-code

[0]: my article: https://smileplease.mataroa.blog/blog/claw-code-reaches-100k...


I have been using @freakynit's runpod as well all be it, I like making working pomodoro apps as my own custom test, and although its not good for it (none of the prototypes work), I feel like it can be good within a specific context like Sql as you mention.

I imagine this being used as sub-agents with some sota models directing them but I wasn't really able to replicate it personally (I had asked Claude to create a detailed plan for a pomodoro app and then passed it to Bonsai)

I also tried its writing skills and actually they are kind-of decent, I also found that this model actually uses very comparatively little em-dashes.Its fine tunes are gonna be some really amazing things to come out. I hope someone makes a fine tune for website/tampermonkey extensions ;)

I remember using chatgpt-3 to use svelte/sveltekit to make a green button to blue button and having the text inside those buttons change and it's my personal wow moment from gpt-3 (This wasn't really able to accurately replicate it even in plain js), but I think that maybe the current model isn't good at writing html but the possibilities with custom-training these models and the idea of 1 bit model feels really great to me.

Especially with the idea of Ngram-embedding[0] (Meituanlongcat/LongCatFlashLite) and its idea. I imagine a 1 bit model + Ngram-embedding idea and I feel it can have many endless possibilities.

[0]: https://news.ycombinator.com/item?id=46803687 (I had submitted this but it seems to have had no attention during that time)

Maybe a 1 bit model like this and diffusion models for coding purposes might also go hand in hand, there are many experiments which can be done with this! (Also yes, many thanks to @freakynit running the runpod, I think I really learnt many things about this model in particular because of his runpod)

TLDR: I feel like this model is good within writing or atleast better in it than usual and it can be good asking it General purpose questions default but I feel like its not good at making html which can be fair, good to see that they are good in sql, but, not sure how they might approach in normal coding tasks. But either way, its an extremely fun model to play with!

(Edit: After some more tries, I have been able to make even one prototype of it after Gemini had holded its hands/giving it the code/errors, its not the best at this but still it works, just barely, https://gist.github.com/SerJaimeLannister/e90e8a134e4163f205...)


> I feel like it can be good within a specific context like Sql as you mention.

Yes I think very constrained task: known data universe, well known language etc should be the best possible place for small language models to play


Yes, I think though that, maybe 1) this shows 1-bit llm models working so more companies can do that so that we can get more competition within this space (+ ngram-embedding idea)

Another point, but I feel like, we can see some really good fine tuned models out of this model, the community feels excited about 1-bit LLM architecture. We are gonna see some good innovation within this space in the upcoming future.


Kind sir, May I say to you thanks for doing so! I really appreciate it :D

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: