Felz

I believe this is part of the AWS's explicit policy, possibly also in the terms of service:

https://aws.amazon.com/compliance/data-privacy-faq/

It would be crazy suicidal for their cloud business to break it, and possibly open them up to lawsuit.

Oh cool, my project is on HackerNews! I was wondering about the sudden uptick in user signups, and then I checked HN...

I'm Scott, feel free to ask me anything about the service.

Wait, they only made this attack 2000 times harder? That was the whole fix?

If they curated like that, esp with employees, they'd open themselves to an infinite number of lawsuits.

I've been using a reactive websockets over HTML in Kotlin implementation that I wrote myself, and it's been the least unpleasant web development experience I've ever had. There really is a huge advantage when you don't have to marshal/unmarshal/validate a request for a simple callback on a button, and page loads can be much faster.

That said, there is more operational complexity (though given my site is relatively small, not much yet). State that lives with the server dies with the server, so deployments can take longer if client sessions live for a while.

Average user here. Not really interested in jumping through hoops to get addons that seem like they were arbitrarily taken from me. The mobile rewrite shouldn't've been released until it was actually ready, because average users _do not care_ about Mozilla's development convenience.

Keep in mind that how strict the speed limits are seems to vary significantly by local laws and driving culture.

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

So what might be unacceptable "speedo pedo"ism where you live might be completely normal going 5 miles over the posted speed limit elsewhere. My personal experience is that people tend to drive at a speed they feel safe at, and don't pay that much attention to the signs.

Wouldn't DKIM verification require having the public key of Gmail from 2014 on hand? I looked up some old 2014 era emails I got from gmail and it looks like they were using the "20120113" selector at the time, which is no longer available through DNS query.

Funny enough, a lot of the complaints I heard were about how Android has made their slightly older phones unusable, taking up a lot more storage space while delivering no perceived benefits.

My friends accepted that Google Maps adding eg useful info about stores was a useful improvement. I guess we'l see if it gets ruined in the future.

I live in Benicia and I think the alternate plan is slightly disingenuous in removing Vallejo as a consideration for the Benicia stop. Much of Vallejo is only 20 minutes away from the rail line.

Hmm, wikipedia (and the ASN used as source) seems to describe this incident a bit differently. I didn't see anything about core temperature rising uncontrollably, it's slightly inaccurate to say that control of the rods themselves was interrupted (only a positioning indicator; they could've emergency shut down the reactor with the rods if they needed to; not sure why you say the control rod failsafe didn't work), and the "emergency flood" of boron which "luckily worked" was standard operating procedure. Do you have a better source?

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

Thank you for this comment. As someone who played a bit with GPT, it was very poignant for me. I still think it's incredible that GPT can put up such convincing facades, that it can generate genuinely novel and interesting text... but it's bittersweet, too, that it can't go any further with them. The ideas are lost in the context window.

I play AI dungeon on occasion, which uses GPT2 to generate freeform adventures. And I find over time that it's not really GPT2 that's writing stories, it's me. GPT2 is putting out plausible strings of words, but I'm the one giving them meaning, culling the parts that go off track, and guiding it in a direction I want to go.

And it is a bit melancholy. You see possibilities, nuances, subtexts, and meanings. The neural net sees words.

It sounds like Amazon is adding quite a surcharge onto their SMS, then.

https://aws.amazon.com/sns/sms-pricing/

You might not realize just how expensive SMSes can be. Based on AWS SMS pricing, they cost 6/10ths of a cent to US destinations each, so each deal sent to 90,000 people in the article would cost twitter $540. That's not even getting into Europe that can cost over 10 cents per SMS, but I doubt twitter supported that.

Sure. I'll add that in to the company policy documentation I'm going to draw up. (Or put it in the TOS itself?)

I do try to answer questions on HN. It's a little hard to keep track of threads over time, though. (Am I missing something?)

Anyway, good luck getting away from gmail!

I wrote a blog post on this a while back:

https://news.purelymail.com/posts/blog/2019-06-21-deliverabi...

Unfortunately there aren't any easy answers. You'll have to be persistent.

Could. I'll see about locking one down.

Cool! I'll take a look.

No offense taken! I think you're broadly right that I could be doing a better job on reassuring messaging, and I'll put that on my work queue.

I also hope to solve any bus factor problems before leaving beta. Right now if I were struck down, I think I'd still have about 9 months worth of free AWS credits on the account running the service, plus any profits. (It's not quite a self sustaining business yet, but it's not that far from it either.)

Hi! I am the "Scott" from Purelymail in question. In the hopefully unlikely event I'm hit by a bus, I do have friends who could step in to keep it running for a while. One of my infrastructural goals is also to have it run itself without manual intervention, where feasible. It's not 100% there yet, though. (Hence the beta.)

As to the terms of service, as far as I know the clause you quoted is fairly standard cover-your-ass. I've probably seen it in a few other service terms. Presumably, Fastmail words it as they do because they've covered all the reasons they might want to do so in their terms already, and have decades of legal experience.

When I worked there a few years back, the spin was that it'd be letting the lawyers themselves get more done, by automating or easing rote but necessary tasks like reviewing company documents &etc. I think the story around what paralegals would do was more nebulous, though.

Last I heard part of the idea behind the lawyering was that it'd give valuable in-house experience to develop the tools. What changed there? Did they get a big contract with Orrick to replace the in-house lawyers?

What's interesting is that it specifically looks like pre 1.4.0 Roundcube with the Larry skin, and not the shinier Elastic skin.

It probably wouldn't take a billion years. The last common ancestor of humans and orcas seems to have been about 95 million years ago.

For reference, a billion years ago, there were no animals or plants yet.

Given that, maybe there are a lot of species that really weren't too far off from human level intellligence, but we just got there first?

To be clear, the previous owners of .org didn't actually do the technical parts of running their TLD, they subcontracted it out: https://www.reddit.com/r/programming/comments/bi6gg5/help_ke...

Given that they're selected for being the very best returns, it's probably a decent (but not completely exceptional) combination of every factor- skill, luck, laundering, connections, etc.

https://www.lesswrong.com/posts/dC7mP5nSwvpL65Qu5/why-the-ta...

I think Project Loom would fit. It's a remarkably sane approach:

https://cr.openjdk.java.net/~rpressler/loom/Loom-Proposal.ht...

I'm not sure how South Australia works, but I think in general you pay for the cost of electricity generated (here going to zero) AND for the infrastructure required to get it to your house.

In part that's why rooftop solar is hazardous for a grid in high amounts- you don't pay the infrastructure price half the time, but the infrastructure still has to be there for you.