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

For example, this blog may have had syndication (RSS/atom/jsonfeed) by default if a blogging framework was leveraged.

Do you know of a CEC adapter which can do HDMI2.1 4K120Hz? What has held me back from this setup is only being able to find 4K60Hz CEC adapters.

The CEC adapter shouldn't care? CEC is its own pin, everything else just goes straight through. (or maybe through twists).

You can also probably maybe do terrible things like CEC on one port and signal on another, it's mostly a bus.


Read it while you can, third-level .name TLDs are being discontinued next year by Verisign. [0]

[0]: https://en.wikipedia.org/wiki/.name#Discontinuation_of_third...



I hope whoever is running na.me (the domain registration was renewed a month ago) is preparing to register domains like joe.random.na.me if you can demonstrate a registration of joe.random.name, for an affordable amount.


I'd imagine that would be infinitely more risky to rely on, plus probably nobody (figuratively) would actually migrate


Thanks for letting us know about that.

Here is an archive link of the original blog post:

https://archive.is/zi6qn

Going to post it here in case someone several years from now comes to this thread and wants to read the original blog post.


> https://archive.is/zi6qn

> Going to post it here in case someone several years from now comes to this thread and wants to read the original blog post.

I wouldn't count on short-links living long. Generally, the point of link-shortening is that they're temporary. "zi6qn" might later get reused for some other page.


How are we so sure archive is going to stay up? It’s already under legal attacks.


You probably want archive.org: https://web.archive.org/web/2/https://ken.arneson.name/2015/....

IIRC, archive.is actually explicitly rejects the idea that they're an actual archive for the preservation of material.


So is archive.org. They take down articles when journalists ask them too and block legal sites from being archived. It is more likely for archive.org to take down or make inaccessible a piece of content than archive.is.


> They take down articles when journalists ask them too

They take them down when their respective author / copyright holder wants them down. That wouldn't apply in this case where the author wants them up, but a DNS infrastructure might cause the links to fail.

> and block legal sites from being archived.

Not sure what this refers to.


>That wouldn't apply in this case

I was contesting the claim that archive.org wants to preserve material. It's hard to trust that an archive will preserve the page you archived if the owner can come along at any time and wipe it. Some people want an immutable copy of a page that they can reference in case it gets changed or taken down in the future.

>Not sure what this refers to.

For example you can't even see 4chan's homepage. For such a notable site in internet history, it is a shame that it is not publicly archived by them.


> I was contesting the claim that archive.org wants to preserve material. It's hard to trust that an archive will preserve the page you archived if the owner can come along at any time and wipe it. Some people want an immutable copy of a page that they can reference in case it gets changed or taken down in the future.

You're letting the perfect be the enemy of the good.

Sure, archive.org may not meet your perfect definition of an archive, but it does see itself as an archive that's preserving things for the future. My understanding is archive.is explicitly rejects that role (e.g. IIRC, they make silent changes to pages, their ownership is opaque, and they reserve the right to just disappear without a trace tomorrow).

This is how archive.is describes itself: https://www.reddit.com/r/DataHoarder/comments/1i277vt/psa_ar...:

> Yes, the word “archive” in the title is misleading. The main purpose is to hold up ephemeral web pages for latecomers. Shots taken more than a few days ago have almost no visits.


>e.g. IIRC, they make silent changes to pages

There is one documented case of this and it was in response to harassment. My usage of archival doesn't involve me harassing the person who is graciously offering that service to me so it is outside of my threat model. This is actually a case of letting perfect be the enemy of good. I frequently archive posts where the owner may want them to be removed so what archive.org allows is against my threat model against the archive. They are not the same.

>they reserve the right to just disappear without a trace tomorrow

I already gave you an example of how they made the archive of 4chan disappear out of the blue with no warning. Archive.org is no better in this respect of archives going away with no warning.

Archive.org also can randomly lock your account and delete all of the stuff you archived with no way to appeal. An organization that actually cared about preserving things wouldn't be so haphazard with the collateral damage of deleting everything from users. The site cannot be trusted for long term preservation of data.

Ultimately everything is temporary so if you really care about backing something up you should do it yourself. I believe both these sites do give you a download option to save the pages locally.


>> Not sure what this refers to.

> For example you can't even see 4chan's homepage.

That's likely the same thing, the 4chan admins asking IA to omit them.

> I was contesting the claim that archive.org wants to preserve material. It's hard to trust that an archive will preserve the page you archived if the owner can come along at any time and wipe it.

Being able to preserve material depends on their surviving.


What makes you believe they don’t own the second-level TLD here?


I do own the second-level TLD


Modern PCIe NVME drives typically see a few microsecond latency, but even DDR2 latency was around 10 nanoseconds. Memory remains top dog by a long shot.


Fall of Civilizations: https://fallofcivilizationspodcast.com/

Each episode is of exceptional quality and retells history in an engaging manor. Since it's history, the entire backlog is still relevant.


> Tax dollars don't fund the government. The government funds the government. That's what "Fiat currency" means.

There was $4.9 trillion in revenue and $6.8 trillion in outlays in 2024 [1]. 95% of that revenue was from taxes. In spite of the high deficit, it remains a true statement that the federal government is funded by taxes as they account for the majority of funding.

[1]: https://www.cbo.gov/publication/61185


The money would be spent regardless of what is collected through taxes. It's not as though taxpayer money goes into a gigantic bag in the treasury, and then those same dollars get pulled out to fund public spending until the bag is empty.


Those lending to the US government are doing so by purchasing US Treasury bonds, which have due dates and earn interest.

Based on the credit rating of the US government, they will certainly not end up with less in nominal terms.


Considering that the current deepest borehole in the world is 12km [1], 20km is a lofty goal.

[1] https://en.wikipedia.org/wiki/Kola_Superdeep_Borehole


> The deepest reached 12,262 meters (40,230 ft) in 1989

Hopefully those 30 years of technology will be of service


Those statistics are for New York State, not NYC.

See page 27 (PDF page 33) of [0].

Commercial natural gas usage in NYC is roughly 40% of non-power generation natural gas usage. The actual proportion varies greatly depending on time of year, presumably due to natural gas heating in residential properties in colder months.

[0] https://www1.nyc.gov/assets/dep/downloads/pdf/air/nyc-combin...


I believe part of the reason for American companies to not list teams publicly is to guard against easily poaching talent using that data.


LOL. Like that information isn't readily available on LinkedIn, FB, Twitter, Github, etc. There are ways ($$$) to retain talent.


Consider applying for YC's Winter 2027 batch! Applications are open till November 2.

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

Search: