This is cool. One feature that would be awesome would be to query the "graph" of points.

I have point type A and I want type B - what's the most economical path to walk between the two point types.

Thanks for the feedback :) Yes, it’s on the roadmap.

However, in practice, indirect transfers usually don’t give you good exchange rates - I might be wrong though

+1 for this! Seems useful

I was curious the last time I sold tickets so actually called them asked about this. The emails are aliases to verify the tickets are valid, then they are forwarded to the buyer.

Usually the format is <email prefix of buyer>@<stubhub controlled domain> and the tickets are forwarded pretty quickly. This way stubhub can actually validate you sent over a PDF + the contents of those PDFs or images if the buyer disputes delivery.

but why would they use shady looking domain names for this?

I’d guess so that ticketmaster etc. can’t identify which tickets are being sold via StubHub. If the domains were identifiable, ticketmaster could block the transfers.

But if StubHub (MarkMonitor Inc.) is visible in WHOIS, then ticketmaster could do a `whois emeraldsummitadvisors.com` and block still block it?

Cat and mouse game with Ticketmaster team..... next up anonymous registrations.

Yeah. That's a slightly more expensive endeavour. The question is is they can scale enough the number of domains and connect to send emails. New domains + not known can have not so good sending reputation. If they use low number of domains, Ticketmaster could start blocking new domains with more than average number of ticket purchases.

To most people it isn't shady looking? It gives the feeling that you're emailing a person, which might be what they're going for.

Totally just guessing: Maybe to reduce inbound spam that's forwarded to the buyers? If it was one central email address, spammers could just send a ton of images/PDFs to all known prefixes of that central domain, but needing to guess the prefix + right domain adds a layer of indirection? They may not do much validation on the forwarded tickets, just store it for disputes

Wouldn't it be much cheaper and effective to add a random secret to the bit before the at rather than after?

Shady looking domain names are usually cheap and available?

That's my best guess at least.

That makes no sense, why wouldnt it just be @stubhub.com or @stubhubtickets.com

Wow, this makes running LLMs locally significantly more approachable. Very cool!

Agreed!

He's doing a great job of only looking a little bit uncomfortable with the questions.

Seriously this Q&A is incredibly odd.

TLDR; it's because they're not pushing new features but rather bug fixing through the proper channels (Webkit framework or JavascriptCore).

The interesting bit being that they can patch native methods using JS. You have to write a specific patch in JS though and still fix the original native code before pushing out the next version.

I've been pleasantly surprised at how fast Github has been moving since that critical open letter came out. For a giant company like Github they've been releasing developer tools very quickly.

Makes you question what exactly they were doing before the letter came out.

Taking naps in big piles of cash? Heh.. Enterprise-only stuff?

Yes. They were/are heavily focused on Enterprise features.

Also interesting to see in this earnings report is that Google's "other bets" aka moonshots are 3.5 billion in the red...[0]

[0]https://abc.xyz/investor/news/earnings/2015/Q4_google_earnin...

I'd expect years, if not a decade for those to make money.

I'm fond of how Google's voting structure is designed to prevent common stockholders from changing the direction of the company (for the same reason Elon Musk won't take SpaceX public until there are ongoing flights to Mars): investors are too concerned with short-term enrichment.

If all of your ventures are making money hand over fist, you're not trying anything hard.

I totally agree, this is just the first glimpse we get into the finances of the moonshots.

I'd be interested to see if this is just because it's new and novel or because of some sort of human psychological difference. The latter would be really interesting.

I believe it's the latter, but there is no way I'm discussing that viewpoint in here.

Woah

The only way to bring about that kind of change is from the regulatory level, the companies benefit from obfuscating the changes they're making you agree to. Making it easier to see that they just added a single provision allowing them to sell your data is not something companies will want to implement themselves.