HN2new | past | comments | ask | show | jobs | submit | more ronsor's commentslogin

The idea seems to be a simple enough computing system (instruction set, programs, CPU, etc.) so that it can be documented, operated, and recreated indefinitely with the least amount of hassle, ideally reusing existing hardware.

It costs money beyond 10 links, which means either a paid subscription or institutional affiliation. This is problematic for an encyclopedia anyone can edit, like Wikipedia.

This is assuming they can't work out something with wikipedia to offer it for free (via a wikiforge tool, or bot) in exchange for the exposure of being the most common archive provider/putting a "used by Wikimedia" logo on their website.

The major reason archive.today was being used is that it also bypassed paywalls, and I don't think perma.cc does that normally.


Wikimedia could pay, they have an endowment of ~$144M [1] (as of June 30, 2024). Perma.cc has Archive.org and Cloudflare as supporting partners, and their mission is aligned with Wikimedia [2]. It is a natural complementary fit in the preservation ecosystem. You have to pay for DOIs too, for comparison [3] (starting at $275/year and $1/identifier [4] [5]).

With all of this context shared, the Internet Archive is likely meeting this need without issue, to the best of my knowledge.

[1] https://meta.wikimedia.org/wiki/Wikimedia_Endowment

[2] https://perma.cc/about ("Perma.cc was built by Harvard’s Library Innovation Lab and is backed by the power of libraries. We’re both in the forever business: libraries already look after physical and digital materials — now we can do the same for links.")

[3] https://community.crossref.org/t/how-to-get-doi-for-our-jour...

[4] https://www.crossref.org/fees/#annual-membership-fees

[5] https://www.crossref.org/fees/#content-registration-fees

(no affiliation with any entity in scope for this thread)


> Organizations that do not qualify for free usage can contact our team to learn about creating a subscription for providing Perma.cc to their users. Pricing is based on the number of users in an organization and the expected volume of link creation.

If pricing is so much that you have to have a call with the marketing team to get a quote, i think it would be a poor use of WMF funds.

Especially because volume of links and number of users that wikimedia would entail is probably double their entire existing userbase at least.

Ultimately we are mostly talking about a largely static web host. With legal issues being perhaps the biggest concern. It would probably make more sense for WMF to create their own than to become a perma.cc subscriber.

However for the most part, partnering with archive.org seems to be going well and already has some software integration with wikipedia.


If the WMF had a dollar for every proposal to spend Endowment-derived funds, their Endowment would double and they could hire one additional grant-writer

Do you have experience with this? I'd like to hear more, really. I think this is the first time I've seen a suggestion for something new they can spend money on. I usually just see talk about where to spend less.

If the endowment is invested so that it brings very conservative 3% a year, it means that it brings $4.32M a year. By doubling that, rather many grant writers could be hired.

Well the last annual report I could find actually says that they got a return of 17.65% so 3% would be pretty bad

https://wikimediaendowment.org/annualreports/2023-2024-annua...


I don't think high seas sites hosted in off-shore jurisdictions particularly care about the DMCA...

Yes, it's a common feature of malware.

Reactivity is not a virtue. If your existing laws are already not being followed, then adding more is a folly that merely possesses the illusion of effectiveness.

Let's be fair: almost everything is linked to Peter Thiel's dark magic company these days.

The UK's NHS is already quite close with Palantir: https://www.palantir.com/uk/



Brazilian Serpro (state IT company) also has a deal with Palantir.

"The partnership between Serpro (Federal Data Processing Service) and Palantir Technologies is a technological collaboration focused on leveraging large-scale data analysis (Big Data) and the implementation of Artificial Intelligence (AI) in the Brazilian public sector"


And he was in the Epstein files, allegedly meeting with Russian officials repeatedly at Epstein’s place. One of which is a handler for assets of their intelligence service.


We should not be doing business with people mentioned that many times in the files.


[flagged]


What crime was he convicted of? Surely that has some bearing?


So which crimes are unforgivable enough to permanently prevent someone from rejoining society? Typically we handle these by assigning life sentences without parole. Was that the case here?

"a crime"


Feel free to elaborate on which crimes should require us to permanently exile someone from society and which ones shouldn’t. Usually the court handles these with life sentences without parole. Was that the case here?

The usual suspect.


Health service buys software from massive multinational. They also work with Google, Azure and AWS. More news at 11.


> buys software

To do what, exactly? This is public money being spent. Why are you so eager to be ignorant of it?

> from massive multinational.

Let's be honest: "Health company buys software from US defense monopolist."

> They also work with Google, Azure and AWS

Yes, and you and I can also buy those products and use them, do you use any of palantirs products in your daily life?


who is ignorant of anything? Foundry is a pretty decently known product now, plenty of companies and gov departments use it. It's a B2B product but anybody can go spin up a test account or read the docs, none of it is secret?


> It's a B2B product

And how exactly is the NHS making use of it? What problems is it solving for them? What new capabilities do they have now that they've deployed it?

> who is ignorant of anything?

B2B is as generic of a label as you can get.


They have a whole website to answer all these questions, and there's a fair bit on youtube etc? Honestly, I'm not sure why "what if we had a consistently available tech stack and data model" is that far fetched an idea though.

https://digital.nhs.uk/services/federated-data-platform


> "what if we had a consistently available tech stack and data model"

And Palantir is the only company that provides this? They didn't have this before? After they bought it what changed? Again, what new capabilities do they have now that they lacked before? What required them to make a deal with a defense contractor for a "consistent data model?" Did other ERP vendors not exist?

It's 2026. Make this make sense. Justify their decision don't excuse it with hand waving.

> is that far fetched an idea though.

You are working really hard to ignore reality here. Basic accountability in spending tax payer funds is not a "far fetched idea."

> https://digital.nhs.uk/services/federated-data-platform

The deeper you dig into that site the more you discover the supposed benefits still had to be developed locally and merely rely on Foundry as a database. You're telling me the NHS had to make this highly questionable decision because they couldn't find a database vendor?

This stinks. I'm sorry you can't or aren't willing to admit you smell it.


They won a public government tender..? I presume other people also bid and it wasn't as cheap/convincing? Would you feel this suspicious of the contract if it was using Databricks or Snowflake?

Claude has no issue with this for me, just as the other commenters say.


GPT-4o has severely damaged the minds of many individuals.


Not false. But also helped some who were already damaged. I wonder what's the netto?


The people who consent to being subjected to the LLMs aren't the only people impacted. If they were, a cost vs benefit analysis makes more sense.

LLM driven delusion is driving people to harass others, even commit murder... and, less cosmically, gum up communities, online forums, and open source projects with gonzo conspiracy laden abuse.


Yeah. I don't want to defend it too hard either. I ultimately canceled my ChatGPT subscription due to the introduction of 4o.

4o was the most famous driver of this kind of behaviour, I think. Other LLMs now have better guardrails.

But looking at the relevant reddit, I can't deny it has helped people function who couldn't otherwise.

It's the same with religion. Some creepy people, you're really glad they've at least <Found Christ>. It's a good thing Christ doesn't have a plug.


Why specifically GPT-4o?


4o was ridiculously agreeable. It would go along with anything: https://hackertimes.com/item?id=43840842 The last time they moved to deprecate it there was a similar backlash which resulted in them reversing that. https://hackertimes.com/item?id=44842147 https://hackertimes.com/item?id=44839842


OotB, it had a more creative voice and less safety systems built in. I think these folk could wiggle around the prompting for modern models to be better if they were more savvy.

For example: Silly Tavern users with jailbreaking, advanced prompting, and paramter hyper-optimization.

Maybe that wouldn't appeal to this kind of user anyway, since it'd peek too much into the sausage factory? Who knows.


Read the reddit post.

It would feed into delusions about being that user's boyfriend while the new model is rightfully saying none of it was really true.


Funnily enough that's given as an example of a prohibited type of app in their review guidelines.


@PlatoIsADisease (because dead comments can't be replied): the term WalledGarden has been a term for this and related concepts since long before marketing-speak had completed the takeover of the internet.


[flagged]


I prefer Walled Rent Seekers Paradise


One runs on machines we pay for (= costs us money). Another runs on end users' machines (= costs them money).


You can pay to have better/more machines, while you can't do the same with your clients


But you can tell users their hardware isn't supported and to fuck off until they buy new hardware. Which is the norm.


But it affects user experience.


Backend performance issues can grind your system to a halt. It’s basically a requirement for Discord to work reliably.

Front-end performance is not a hard requirement for most end users, unless the app is actually unusable. Discord isn’t that bad compared to some software I’ve used. You have to get beach balls on startup and complete UI freezes for people to really care. If it’s good enough for most people, shaving some MB off the memory usage or small number of ms off latency isn’t important to the business


As long as negative experience doesn’t meaningfully impact user monetization, there’s no business incentive for that type of company to care.


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

Search: