For anyone interested in push-to-talk/intercoms we built a free tool at the beginning of quarantine last year called Squawk - https://www.squawk.to

Looks intriguing, but from the website I can't really tell what it does/doesn't do. Maybe a quick youtube video would help make this concrete? For me, the question is: if I put this on an old smartphone and put it in my kid's room, can I/she use Siri to send each other message? Or do we have to separately open the app, wait for it to load, and then say the message?

Interesting. How do you make money on it?

When something is free it usually is paid for by selling my data, as the free-tier of a paid product, or the owners haven't figured out how to make money on it and are hoping to figure that out later.

We don't but it helps with marketing and has synergies with other paid products

Most of the traffic is P2P. We operate a couple of geo distributed relays for NAT'd traffic. But everything is 100% e2e encrypted thanks to webrtc.

It's mostly a tool for us and something that helps sell our other products, but we decided to split it out as a separate app so that it could be used independently. All we collect is a name and email, and even those can be pseudonymous.

We have a free version called Squawk over at https://www.squawk.to

I think the stripped down hardware is quite an important part of the offering.

Wow...an S-1 with positive net income.

Yes but they don’t really have a way to expand outside of the US, they are basically an answer to a uniquely American problem unless they’ll go they’ll target the self medicating crowd that uses the relatively dodgy online questionnaire/chat prescription pharmacies.

Why is that a problem? Companies don’t need to be international.

This is not really a problem. If they were in a situation where they didn't have a way to expand outside of say, Burundi, that would be a problem, but the US is 25% of the global economy.

> Yes but they don’t really have a way to expand outside of the US, they are basically an answer to a uniquely American problem

The US accounts for 45% of global pharmaceutical sales. Even if they're limited to "only" half of the global pharmaceutical market, that doesn't sound like the worst position to be in.

I think that’s for North American and it’s in revenue not in actual sales, but yes that is currently the case but it’s also clear that the US health system cannot continue as it does for much longer. So overall while they can clearly profitably operate in the current market I’m not so sure about their longevity.

The market size and business opportunity is not measured by units sold but by overall revenue. Am I missing something?

Yes you are missing the point that this company can only operate because of just how utterly broken the US healthcare system is, and sooner rather than later there will be sufficient political pressure to change that.

Because their risk is dramatic health-care reform? In the USA?

Medicaid for All almost made it into the Democratic Party manifesto this time, it was cut but realistically it will make it into the 2024 or 2028 one especially once the longer term effects of COVID-19 will be more apparent because many of the recovered will suffer long lasting effects potentially for the rest of their (hopefully long) lives, and overall once all the bullshit settles down people will see it as a pretty good wake up call.

You're almost proving my point. Look at the numbers; between now and any reform in 2024 they'll be making $100MM/yr in Net Income -- by 2028 they'd have a war-chest of $800MM -- cateris peribus.

"sooner rather than later" won't be in 4 and could take 8! years (that's seems later to me)

If you're concerned about their longevity it looks like they'll have plenty of cash to figure it out over plenty of time.

It depends on the time scale you are looking at 4-8 years isn’t a lot, and also there are plenty of other changes that could happen in the prescription drug market and there have been many pushes for that which don’t require a full on single payer system.

Most of these changes still maintain the necessity for price transparency, which GoodRx still offers. There will always be some need for an online portal through which to shop for medication, regardless of whether or not there are price controls or if the payments are reimbursed/subsidized by the state.

8 years is a bit less than half of Google's existence.

A political “manifesto” is meaningless.

GoodRx doesn't cover physician administered drugs, which makes up a sizable proportion of that 45%.

So they're a mature, profitable company going public and that's a "Yes, but..."?

They seem pretty vulnerable to being obsolete if there is any sort of substantial change in the prescription drug market. And there seems to be quite a push to change the prescription drug market. Just look at how convoluted their revenue generation is, there are like 5 middle men.

That's why they use the money that they mad to hire lobbyists to ensure that policies won't change in any meaningful way.

I hate how this world works :/

This is true of basically any major healthcare play. On the (at this point extremely) off chance that we move to single payer, for instance, insurance companies face an existential crisis. On the off chance that we fix prescription drug price gouging and absurdly inconsistent pricing and all of these problems that GoodRx solves for patients simply by _cutting them in on the scam a bit_, then GoodRx faces an existential crisis.

The fact that there are 5 middle men is... more of a moat in this case? They've solved a complex, intractable problem profitably. Does this mean they're a 20-year company? Eh, not without some significant additional revenue streams or pivots into different areas, and to that point I agree with you about their vulnerability. But for the next 10 years? These are good margins building a business that only incrementally reduces the fucking over of the consumer.

Also there's a large selection bias: people using less common or more esoteric/demanding languages tend to be better engineers who have invested the time in learning something off the beaten path. That absolutely does not mean it's the right decision since most employees will statistically be average and you want to have a stack that appeals to a large enough talent pool.

We've got a core product running on Django, and one thing the author doesn't mention is testing migrations. The migration system is wonderful, except for the difficulty in testing migrations. There's no sane/official way to do this. And it's such an important thing that I don't get why the Django crew haven't tackled it.

I've used django-test-migrations before and was really happy with it. https://pypi.org/project/django-test-migrations/

We've played around with that and it's on the roadmap. But it just adds another set of fixtures to herd.

What sort of proposal would you put forth? I generally install one of my production backups onto a staging server (secured appropriately for production data), and run the migrations against it to test.

That's not the intention on the code, I thought he is talking around writing a migration and a test that ensures the data migration is correctly applied

My strategy here is to have a staging server where migrations are applied before they can hit production.

As someone else pointed out, migration without CI can be dangerous and difficult to coordinate with deployment.

This is what we do but it's too manual for my taste.

Hi Vlad - your anecdote about ship tracking is interesting (my other startup is an AIS based dry freight trader). You must know the Vortexa guys given your BP background.

How does QuestDB differ from other timeseries/OLAP offerings? I'm not entirely clear.

thank you, life is an interesting experience :) I used to work with Fabio, Vortexa CEO and had to turn down an offer of being first employee there to focus on QuestDB. They are an absolute awesome bunch of guys and deserve every bit of their success!

What makes QuestDB different from other tools is the performance we aim to offer. We are completely open on how we achieve this performance and we serve community first and foremost.

The prompt contained a few react examples with annotated English descriptions that were similar enough. You’re not seeing the entire prompt and as impressive as this is, it’s likely that you’d end up writing more code trying to get it to do what you want than if you just wrote it yourself. Future GPTs may of course be better, but there’s a bit of “magic” to these demos that the author isn’t super upfront about.

We're incredibly excited about GPT-3. I think there is a fair bit hype exhaustion, especially from the likes of OpenAI ("our AI is too dangerous to release"). So this is completely understandable.

However I think what's missing here is our benchmarks (a la Turing test) are about negation as opposed to affirmation. We tend to evaluate AI on whether or not we can discern the fact that it's AI. We seek to negate it as human, as opposed to affirming it as human (or close to). And this is not the right mindset when it comes to AGI because the gap between "obviously not human" and "human-like" is enormous. These are all definitely steps in the right direction, and the applications for even robotic process automation will be huge. But we're not even close to having nets that can reason about even the most basic things.

> However I think what's missing here is our benchmarks (a la Turing test) are about negation as opposed to affirmation.

I would question the value of the Turing test, and maybe think that's not a great example for AI.

There's always been this assumption that passing the Turing test would mean we had AI, but I think that was always predicated on the machine generating the outputs. With the GPT- models, it's not clear that this isn't a form of compression over an immense data set, and we're sending pre-existing _human_ responses back to the user. It implies to me that we can pass the Turing test with a large enough data set and no (or very little) intelligence.

All of this makes me believe "These are all definitely steps in the right direction" is questionable.

I’m not really sure what containers/k8s has to do with JVM? Containers are just name spacing, not virtualization. If you’re running in the cloud, either way you slice it you’ll have two virtual machines: hypervisor and JVM.

I’d argue that being able to herd your JVM procs like cattle makes them good candidates for k8s because you can always just set resource limits so they get purged when the heap becomes too large.

I think this is pretty myopic. As someone who straddles both sides of this, the world is not that simple.

If the "business side" as you say isn't viable, then nothing matters. This has nothing to do with late stage capitalism...this is capitalism 101. If you're not hitting critical mass, you're dead.

This take also continuously misses that most people, and thus must customers, are willing to trade privacy/performance for convenience. If that's not your cuppa, then hopefully there's an alternative.