Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
olliej
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
50 ms
·
661.
▲
by
olliej
3y ago
Wow “no nacelles have overheated or fallen off the max”. Really? Not even when a different single point of failure hidden “feature” flew a bunch into the ground killing all on board?
662.
▲
by
olliej
3y ago
you're right, it's not an sms replacement: it's actually secure, and it has no requirement that you have a sim or phone number. Something RCS does not have. You can use SMS (and apparently RCS in future?) to talk to others, a
663.
▲
by
olliej
3y ago
Isn't 30% standard across pretty much every App Store? Steam does not have lock in and yet I thought charged 30%, GoG is 30%, etc. Obviously other app stores could in principle charge lower amounts because they don't actually have
664.
▲
by
olliej
3y ago
You mean "Apple should be required to build messaging services on top of apple's infrastructure". Google also prevents me from making a company that uses its search engine infrastructure and unlike iMessage it is actually the
665.
▲
by
olliej
3y ago
I don't understand why people are cancelling their streaming services, the increasing prices, additional advertising, and spreading the content among different services creates a much more traditional viewing experience. This is more r
666.
▲
by
olliej
3y ago
They make the key exchanges much larger for the same level of security, but the implementations are generally much harder to ensure are secure and safe as RSA is fairly susceptible to side channel attacks. The old argument that ECC is slowe
667.
▲
by
olliej
3y ago
Ugh, so many RSA certs :-/
668.
▲
by
olliej
3y ago
Everytime I see JSX I think of E4X, which is a now defunct standard that contributed to JS having for(of) instead of for each() :-/
669.
▲
by
olliej
3y ago
Fun story, when first developing the web socket protocol we had to deal with kind of nonsense (especially for large commercial routers that did "deep packet inspection"). The solution is that the websocket protocol essentially doe
670.
▲
by
olliej
3y ago
> would likely be for courts to decide. this is the core issue, essentially if you want to publish OSS code you now need to know you can afford a lawyer, because even if a court decides you aren't liable, getting to that point requi
671.
▲
by
olliej
3y ago
Like having a Patreon or similar sponsorship? Sounds like a recipe for licenses that say "you are not permitted to use this software if a bug could cause you any harm"
672.
▲
by
olliej
3y ago
If you have any kind of Patreon, GitHub sponsorship, buy me a coffee, or anything to "support development" it would seem you have met the requirement of being a commercial developer who should be liable.
673.
▲
by
olliej
3y ago
Here's the problem as I see it: * Person A makes OSS project P * Organisation/Person B uses P * A vulnerability in P causes financial harm to B Is person A now liable under this law? What happens if person A has a Patreon or GitHu
674.
▲
by
olliej
3y ago
Since when has stock based compensation ever cause misaligned incentives or negative externalities? :D
675.
▲
by
olliej
3y ago
Wait, so now we're saying stock based compensation is bad? does that mean we'll stop giving executives stock based compensation?
676.
▲
by
olliej
3y ago
You can already remove a number from iMessage (because this exact problem came up back when iMessage first came out)
677.
▲
by
olliej
3y ago
That’s something you can do from iOS or macOS, or you can change your account password (which can be done on the web)
678.
▲
by
olliej
3y ago
Because google is an advertising company, and makes money by spying on users. Google being found to scan emails received in gmail and advertise on that basis is why amazon receipt/invoice emails no longer include details inline. GMail
679.
▲
by
olliej
3y ago
As a dumb question, how does Apple make a change that would allow android messaging to work without a cellphone? How does apple make a change so that android's messaging system isn't unencrypted by default? RCS is not encrypted, g
680.
▲
by
olliej
3y ago
Ok, should Sony be _required_ to support playing playstation games on an Xbox if people are willing to pay?
681.
▲
by
olliej
3y ago
So if a company is large enough it should be forced to provide free use of its infrastructure? Can I also demand google provide me with free use of their infrastructure?
682.
▲
by
olliej
3y ago
I'll try phrasing this differently: should any company be required to provide free use of their infrastructure to non-customers?
683.
▲
by
olliej
3y ago
This is like saying "is X is better than IE, why do the sites that work in IE but not X".
684.
▲
by
olliej
3y ago
"AI in the cloud" with user data requires users to give up any pretense of privacy so it's amazing to me that people are ok with this? If you build "AI" into an OS, and that "AI" is on the cloud that means
685.
▲
by
olliej
3y ago
Also who is paying for all this? iMessage is not a free service
686.
▲
by
olliej
3y ago
No, the point of the ruling is that given an arbitrary amount of data, if a machine can generate the "invention" (in the patent sense) then you cannot patent it. Claiming that you invented the thing because you setup the AI would
687.
▲
by
olliej
3y ago
When you file a patent based on a "novel" application or combination of existing ideas, or ideas from other sources, those ideas are not something you can patent. e.g in "somebody came up with a wheeled cart, somebody came up
688.
▲
by
olliej
3y ago
The patent office says "if an AI comes up with it's not patentable" so you turn around and sign a piece of paper that says you came up with it. How is that not perjury? I'm not making a statement on "should a machin
689.
▲
by
olliej
3y ago
That would be perjury?
690.
▲
by
olliej
3y ago
I love that I can't see CNN content because it needs me to disable my ad blocker to ensure my privacy. :-/
More ›