Hacker News new | past | comments | ask | show | jobs | submit | smingo's comments login

USA has increased its export controls on advanced process node chips, with an act that comes in to effect today, and will get published in a few days' time. The preview is available at https://www.federalregister.gov/public-inspection/2024-28270... .


Barber's Adagio for strings, although it gets referenced enough; Khachaturian's Spartacus and Phrygia, especially the brass version; Holst's planets, esp Uranus; Beethoven's 7th; Gorecki's 3rd.


For me it’s Bach: Erbarme dich, mein Gott.

In this version: https://youtu.be/BBeXF_lnj_M?si=kPTUsAB19_RyECO5


Some of mine include Bach's violin partita 3, Beethoven's 9, and Mahler 1.


indeed, with an incoming Teams meeting invite, it should be determinable from the sender's context which account should work on the meeting. Instead there is 2 minutes of waiting, and what seems like pot luck with the account.


Not so sure that's fair. There are papers cited from 2016 that are from other authors.


The dystopia in the film version of Ready Player One clearly borrows from this.


Putting the WiFi SSID and password on the side of the energy monitor sensor box allowed for this:

> The whole thing was quite dissapointing. However, I do have a few Raspberry Pico microcontrollers lying around at home. If I could connect to the WiFi network of the energy manager directly and get the data from the server, I could just extract kW consumption from the API, multiply it by a correct rate and then display it on some Grafana instance.

Love it!


Regulating here is necessary, but the challenge is steep! IoT devices may include a complex bill of materials (BoM) including software (SBoM). Vulnerabilities can appear in any of those components.

On the one hand, CVE and vulnerability databases are excellent, and with some automation of vulnerability and patch availability the's the possibility of automated re-build.

But the manifests can be huge. And some component could be vulnerable, but was never anticipated to be so, and perhaps doesn't even have the means to be patched. Update processes for sub-sub-components may not have been exercised, and could lead to bricked products.

So labelling and guarantees are welcomed. But the challenge is practically insurmountable, and until the entire industry steps up to meet it, labelling and guarantees are going to be 'best effort'.


Totally valid point. Escrow then open sourced, or perhaps even some insurance policy so that the future patching and vulnerability remediation is guaranteed.

There's a bunch of stuff consequential to EO 14028 which could allow for some automation of library vulnerability.


I was trying to think of ways to finance it, and "future patch insurance" is clever! There are other sorts of business insurance where the insurer is liable even if the business no longer exists, so it would be doable. Though a policy would require a level of technical competency that other insurance policies don't, since their providing a guarantee of service rather than a guarantee to pay out a certain amount in damages.


As a tool for quantifying the number of civilisations, agree, the Drake Equation is not so useful given the known unknowables.

But as a tool for stimulating discussion, creating engagement for SETI, or as an exemplar for Fermi Estimation, it's excellent.


Someone must have felt very clever when they concluded Erlang would be dimensionless.

  Calls per second X seconds per call = no unit.
But that's wrong. Clearly the number of calls ongoing maps on to lines being used for a circuit-switched network, or datagrams in flight, or whatever. Far from dimensionless.

  Call_starts per second X call.seconds per call_start = calls as the unit
Much more plausible, IMHO.


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

Search: