yurymik

Or easily re-connect AirPods from the phone to AppleTV and watch at night without waking the whole house.

> why isn't that standard operating procedure

That says nothing about system bandwidth: if normally they would approve 10 drugs in N months, but now they threw all available resources on one project to get N/10, it would make sense.

> thousands of people, you see in the end credits

if you look closely, those usually on-set people, medics, drivers, catering services - those who have unions. VFX companies in most of the cases have a dozen of entries out of hundreds who actually worked on the title, or just a single line company mention. I could understand this practice in cellulose era, but saving a few megabytes at the cost of disrespect to your workers is simply despicable.

Now replace all inner slashes with U+2215 (copy-paste from here https://www.compart.com/en/unicode/U+2215)

I meant overhead in a broad sense - both packet size and CPU load combined - what end user actually care about.

My test is what I have to do fairly often: use Windows Explorer to copy 70-100gb file from a network NAS to a local drive. Every so often I click on the wrong network share pinned in the Explorer and see slow transfer speed.

I observe there's about 37% overhead when using TS connection on a local gigabit network.

Copying large file from Synology DS1821+ NAS (Amd Ryzen V1500B) to Windows PC (i7-6700K) is about 111-113 MB/s when accessing NAS directly and 70-73 MB/s when traffic goes through TS (different large files, so no caching here).

Did you try https://www.youtube.com/watch?v=nIoXOplUvAw ?

It looks you can query it further:

  > Show options for intent variable

  Persuasion
  Information
  Anecdote
  Reflection
  Inspiration
  Nostalgia
  Humor
  Heartfelt
  Political
  Spiritual

You can go other way around and use other TUIs for GDB:

* https://github.com/pwndbg/pwndbg * https://github.com/longld/peda * https://github.com/hugsy/gef

Hi Adam! I really enjoy listening to your podcast, but if I may I have some feedback:

* Every time your guest specializes on something low-level - like Rebecca Heineman from most recent episode, David Shayer from last year, and others - you keep saying "yeah, this is a very complicated subject and I don't understand anything about it". Everyone has they own area of expertise, no one knows everything - and that's OK. But when you as a host constantly put yourself down, it leaves a very strange impression on the listener: "well then, invite someone who does".

* When you do a voice over for the interview and queue interviewee responses - the whole sequence feels doctored. It just doesn't sound as a normal conversation, more like out of context clips forced into pre-made narrative. Again, I'm not saying it _is_ out of context, it just feels that way.

Please don't take these two the wrong way, you're doing great job.

Also, I would love to support the show and you seem to have a Patreon account, but prices there are a bit too steep for a casual listener: $6/14.50/36 per month (that's in Canadian dollars). For example, semi-weekly Darknet Diaries has $3/7.50/14.50 tiers.

In the last episode you've announced bonus content with Rebecca, but there's not even an announcement on the Patreon page and it's not clear what tier would allow access to it.

Sorry for the rant, but I hope it provides outside perspective and you might find it useful.

It seems to be a third book in a series. Do you need to read Singularity Sky and Iron Sunrise before this one?

I wonder why that's the case. If you own a DNS name, you can set NS record to point to magic-blah.tailscale.net, so they'll have a chance to geo-locate you and return IP address of the closest VM.

Well, it didn't go that well [last time around](http://www.x86-secret.com/pics/divers/v56k/finalv5-6000.jpg).

There Is No Antimemetics Division.

There's Ring Video Doorbell Elite, so the market isn't completely bare. Yes, it's Amazon and pricey, but it does exist.

Mac laptop? ".local" seems to from mDNS that requires multicast that currently doesn't work in Tailscale: [1], [2], [gh].

1: https://forum.tailscale.com/t/mdns-over-tailscale-network/22...

2: https://forum.tailscale.com/t/bonjour-via-tailscale/3237/2

gh: https://github.com/tailscale/tailscale/issues/1013

There are cheap low-tech solutions [1] and more expensive high-tech [2]

1: https://www.homedepot.com/p/LABTECH-Radon-Detection-Test-Kit... 2: https://www.homedepot.com/p/Airthings-Corentium-Home-Battery...

You don't need to avoid _all_ system calls. Just call sbrk/VirtualAlloc/vm_allocate once and then dole out smaller chunks yourself.

That way if you need to add new platform, it's only one place; you're shielded against weird platform bugs (Apple goes one way, Microsoft the other), so the app behaves the same; you can tweak system to your particular use.

I am a bit confused here. Does it mean that all tires should be single use and no one is allowed to have summer/winter/performance sets?

Video is paywalled and only pptx file is provided there, but you can find pdf with speaker notes at https://www.gdcvault.com/play/1027553/Adventures-with-Deferr...

Even if you do have a physical sim card, finding sim card ejector isn't trivial, especially in a situation when you do need to call 911.

On the other hand, what guarantees do they have that you haven't tampered with the readers? Are they able or can afford to disassemble every single one of them to inspect?

Just copy-pasted a piece from the other HN tread (https://news.ycombinator.com/item?id=31630193) and it's not even close to AWS:

https://unrealspeech.com/k6g8a

Now we have `winget install VideoLAN.VLC` :)

For anyone wondering, there's a missing 'l' in the handle. Correct one seems to be `@kamilkazani`

I still can't find any mentions of external memory mapping without copying it into the heap. _Very_ disappointing as it makes yet another version of WASM completely unusable for many high-performance applications :(

Yes, but once malware is on the other side, it could be caught by the AV doing static analysis/signature match. This point of this technique is obfuscation and evasion.

> With ultra settings, it runs at 4k 30fps (average of 40fps) on an i7 cpu and rtx 3090.

by rob cupisz (Tech and Rendering Lead at Unity Demo Team) https://twitter.com/robcupisz/status/1505875759243612172

There was a patch that adds QUIC support, but recently steering committee have announced that they are not going to merge it and will implement their own version instead.

In the mid 2000-s we were researching what video card to use for a product with very long life time and extremely tedious certification. Once completed, you can't replace one component with another without invalidating the certificate. While both Ati and Nvidia would offer better performance and features, Matrox was the only supplier promising to keep stock for the next 10 or so years.