zhobbs

"That being said, if you’re making a game, it probably won’t work in cross-platform"

I feel like gaming is the easiest thing to do cross-platform since you don't rely on any native UI and OpenGL is pretty similar on different platforms. Unity, Unreal Engine, and many others are designed to target many platforms.

I look at it more in the sense of Uber's product not being a rideshare app, but being a "move you from point A to B in the best way possible app" (best = meets your cost/time/comfort needs).

Here's a good writeup on the "jobs to be done": https://stratechery.com/2018/aggregators-and-jobs-to-be-done...

If you are a fan of niche or "underappreciated" programming and want more of it, I could see that argument being pretty compelling.

Fair point!

My first reaction was: "Wow, FB finally went over the line and is actually an evil spyware distributor."

Then I started thinking about what this app really is. At $20/month per user, it's clearly impossible to recoup that money on a per-user basis via better ad targeting. This app is a market research app with a very small opt-in panel, just like having a Nielsen box on your TV.

I've never felt like Nielsen's data collection is evil, so it makes me wonder if my reaction is rational.

Also, looks like Nielsen has a similar program: https://computermobilepanel.nielsen.com/

Technically, in Dota you only control 1 unit. That being said, I don't know if OpenAI's algorithm had 5 independent agents controlling 1 unit each, or 1 agent controlling 5 units.

I think it's just classic vertical integration, nothing to do with how animals are treated: https://www.investopedia.com/ask/answers/012715/when-does-it...

Agree 100%. This is the same bloatware Android has seen since day one, nothing new here.

China is investing a lot in Africa, good video on the topic, "How Africa is Becoming China's China": https://www.youtube.com/watch?v=zQV_DKQkT8o

If you do any machine learning you'll likely still need Nvidia for CUDA, it was the deal breaker for going AMD in my case.

I was thinking about this too. Search engines should regularly crawl a page and notice that the content has been updated right? So it seems like they should ignore pubdate as it could be gamed, and just measure the real "last updated" date based on content changes.

"it will take significant stagnation on Tableua's part"

As a former customer of Tableau, it feels like this has already happened right? I worked at a company where many teams used it, and there were zero defenders of Tableua when it came time to decide whether to renew...it's a mess, especially on Mac.

There is a really big metric on their dashboard, "Pageviews", and this is a best of all worlds solution (cheaper, more content, diverse content, diverse authors promoting their content, forbes.com SEO juice, etc).

Probably useful for justifying costs to managers if your company is currently migrating to Swift, which is pretty common in iOS shops at the time (and now?)

Agree that the lock-in is weak compared to other markets. Their strategy has been to incentivize drivers and riders to not split usage between other apps.

Uber drivers who do a certain number of hours or rides per day/week often get bonuses, so they'd have to forego these bonuses to split time with Lyft.

On the rider side, they're pulling a play from the airline playbook, with their version of frequent flier miles and status: https://techcrunch.com/2018/11/14/uber-rewards/

Personally, I feel like the correct thing to do for the user is to use SFSafariViewController. Custom webviews are pretty annoying IMO.

Fastly's CEO (a Cloudflare competitor) gave a pretty great tech talk on what it means to be isolated: https://www.youtube.com/watch?v=FkM1L8-qcjU

Guessing we'll see a competitive product from Fastly in the near future.

I've had a pretty good experience with KDE Neon over the last year or so, very few problems.

With this method isn't the downside that you have to invalidate all tokens, and not just the attacker's tokens?

They would have to get a lot of patent deals in place and pay a lot of royalties, so the price point would likely be significantly higher.

Consider hiring a resume writer. The guy I hired was older, and had some very specific techniques to combat ageism (I'm still pretty young, but was good to learn about them).

In my experience, American companies tend to be overly optimistic, and Japanese companies tend to be overly cautious about timelines.

See uncertainty avoidance and long-term orientation: https://www.hofstede-insights.com/country-comparison/japan,t...

The hardware gives a better guarantee of botnet uptime though. If you want it to be powered on and attached to that network 24/7, you can't rely on their laptop/desktop.

I was wondering if it is lazily replicated to PoPs via KV gets. Would be interested in clarification on that though.

Regarding the commentators, I saw that part and took it as them having fun with the situation, not really making a statement about how the technology works. Esports commentators tend to be more jokey/entertaining than traditional sports in my experience.

A single accuracy rate is not that useful when looking at ML problems. There is always a tradeoff between false positives and false negatives depending on the threshold selected. Amazon is guilty of this in this article, saying 0% false positive rate when you set the threshold to 99%. However, it doesn't say how many faces it successfully matched, it's easy to have a 0% false positive rate if you just say "no match" for every photo.

The best practice is to show a ROC curve, which shows the tradeoffs of selecting a given threshold: https://en.wikipedia.org/wiki/Receiver_operating_characteris...

I'm using https://github.com/guregu/dynamo for the data reading/writing. In my experience, the trick with Dynamo is the schema design. I haven't had problems with the libraries/queries as long as the schema made sense for how I'd be accessing the data.

For my recent projects, I've used Go + DynamoDB for a GraphQL backend, and Next.JS with SSR on the frontend. Deploying both to lambda with Apex Up. Stripe for billing, SendGrid for transactional email.

nice work!

If you live in CA you're required to file as a foreign corporate entity, and still pay the $800 in taxes. You can't open a bank account in a CA bank without proving that you've made that filing regardless of where the corp was registered.