veeralpatel979

Netlify Forms is another alternative:

https://www.netlify.com/products/forms/

https://docs.netlify.com/forms/setup/

The killer GitHub feature for me is "browse files" at any commit in the past.

I tried out the beta version of GitHub's iOS app. I didn't find a single bug in it, and I really enjoyed using it for casually browsing code on my phone.

I would highly recommend downloading it if you sometimes take a look at GitHub repos when on your phone, which I think is most programmers.

Unfortunately I didn't use it for about fourteen days or something so my beta access was taken away :(

Just curious: will this help the economy recover, when for large parts of it (entertainment, dining, travel, etc) the reason it's closed is due to the virus and not due to lack of money?

I'm a fan of Plague Inc on iOS and have unlocked all the pathogens...what's the difference between Plague Inc and Plague Inc Evolved?

Is the latter worth buying and playing?

Also: if there's a design doc out there for Plague Inc or a similar game, I'd love to see it.

I chatted with Markus de Shon, one of the creators of riskquant, a few days ago. I would also check out Ryan McGeehan's writings on this at scrty.io.

I'd be happy to discuss this topic with anyone; my email address is in my profile.

The point I was trying to make is that there's many scripts which are easy to write as shell scripts, so developers will write them as shell scripts.

But they are time-consuming to write in mainstream languages, so developers often won't do it if that's their only choice.

Functional core, imperative shell is how this is commonly described!

https://www.destroyallsoftware.com/screencasts/catalog/funct...

https://news.ycombinator.com/item?id=18043058

You're right, maybe the tool or script would be more maintainable if it was written in the company's main programming language.

But also, a developer may not even write it if he or she feels pressured to do this.

Here's a thought experiment: say you're seeding data into your local database server using the `mysql` CLI.

- Would you write and check in shell script to do this, so coworkers can use it? Probably yes.

- Would you write a Python script to do this? Probably no.

- Would you write a Java or C program to do this? No way.

Great work! Is being self-hosted something small teams are interested in, in your view?

For me, if I was part of a small team, I would choose a hosted service everytime so I don't need to do any ops work.

> there's no official Rails graphql solution yet either

I use https://github.com/rmosolgo/graphql-ruby!

Link to the GitHub repo: https://github.com/Netflix-Skunkworks/riskquant

I've used Netlify as a "Heroku for static sites".

I deploy my React apps and landing pages there and don't need to worry about the underlying compute, or load balancing, or anything else. Thankfully I didn't do any frontend work before Netlify was available!

I don't really get all the hype around the Jamstack (https://jamstack.org/), though.

Netlify's hypothesis seems to be that if you deploy websites on CDNs, instead of on web servers, then you'll get better performance.

Also, because it's just a static site, there's no backend or database, so you get better security.

I accept this hypothesis, but it seems like you still need a backend. Netlify is promoting functions as a service [1] so you can avoid having web servers for the backend too, but I'm a little skeptical you get the flexibility to build applications with dependencies or non-trivial business logic or design with this approach.

If anyone's tried a FaaS and this is incorrect, please describe your experience!

[1] https://www.netlify.com/products/functions/

Scale offers labeling as a service. Datasaur is an interface that companies can buy for their own labeling personnel, if I understand correctly.

I liked the Practical TLA+ book:

https://www.apress.com/gp/book/9781484238288

> I'm not into all this YC stuff, but does the YC20 in the title means that someone has funded this idea?

Yes, Y Combinator (ycombinator.com) is a top startup accelerator (and the owner of the site you are currently on).

They fund batches of 150+ startups twice a year. W20 means they were in the first batch of this year.

Strange to say but I actually don't mind being woken up due to someone else's alarm. It lets me experience the pleasure of knowing I get to go back to sleep :)

Fair point. Typically I'm on the side of the free market and less government regulation, but this system breaks down if consumers have no choice but to use a small set of companies, who all have an interest in doing the same thing, and creating a new competitor is difficult.

Great post!

I went ahead and read all your posts in the series:

https://posthog.com/blog/before-the-yc-application/

https://posthog.com/blog/the-yc-interview/

https://posthog.com/blog/moving-to-sf/

https://posthog.com/blog/pivot-to-posthog/

https://posthog.com/blog/after-the-hn-launch/

I'd recommend them to anyone who liked the linked post.

Something I read in another HN thread was that while $200M may not be much compared to T-Mobile's annual revenue, it's probably a lot compared to how much ISPs earn from selling user data.

This doesn't factor in the ongoing PR cost of headlines like "FCC weighing fining ISP X", "FCC fines ISP X", "ISP X appeals fine", "ISP X loses appeal", etc.

Still, I agree the fine could have been harsher.

I would also add learning to evaluate your idea(s) objectively.

When you think you've thought of a great idea, by definition, you're biased toward it.

If you then start building it, the sunk cost makes you even more resistant to abandoning it.

In my experience, just "building the product" can take a lot longer than you expect. I've spent several months building out what were mostly CRUD apps.

Why? I didn't realize it at the time, but I was working part time, I had to figure out how to do many things for the first time, I had to keep myself motivated, and I had to do development but also design and PM work.

If I was working on something new, I probably would:

1. think of a problem to work on

2. talk to a lot of users following the Mom Test (keeping in mind your product has many personas of users) to validate the problem

3. think of a lot of ideas to solve the problem

4. somehow identify what idea to prototype first and what the feature set should be

5. prototype it in Figma, show it to users, improve the prototype until customers are asking "when can I buy this?"

I haven't done this yet. If anyone has any suggestions for getting from set of ideas -> a prototype users want, I would love to hear them!

I'm also a huge fan of the Mom Test!

In your specific customer interview, I would probe deeper. For most real problems people have, they've hacked together a makeshift solution. It may work, but it doesn't mean it works well. It doesn't mean the problem is solved.

I would ask "how is it like working with this larger company?". If you have any specific pain points working with larger companies that you've seen, I'd ask "how do you deal with pain point X?"

I doubt someone would say something is their biggest problem if they have a great solution for it!

I wonder why they chose the name Autopilot!

The name seems to imply something the tech does something that it's not, though in all fairness to Tesla, they are very clear about the fact that Autopilot does require an attentive human at the wheel.

I think having a "complicated" type system like Haskell's is a good thing, because it lets you model data in your program accurately, and more importantly, lets you offload work from your brain to the compiler.

One question: How well do languages like Haskell work for code with side effects, though? Say I want to write a CRUD web API. How easy is it to work with network requests and database operations, for example, in Haskell?

The article seems to imply that if BBC launched its own video streaming service when Netflix was still sending DVDs, it would've become the next Netflix.

I have to disagree. The execution is what really matters, not the idea.

- Could BBC have hired the talent to build and scale a credible competitor?

- Could BBC have popularized the model of on-demand, subscription based TV?

- Could BBC have acquired and created content people want to watch?

If the answer to these questions is yes (and the regulations in question aren't a problem), I would highly encourage BBC to launch a video streaming service today!

Thanks for your comment! But I don't think all legacy will ever get phased out.

Today's code will become tomorrow's legacy.

Hey - sorry your startup didn't work out. If you decide to try again, I'm sure it will be easier :)

And thanks for open sourcing your tech! I see some legitimate critiques in the comments here - "it's not production-ready but you claim it is", "you shouldn't have built your own e-commerce solution" -- but even if no one uses your open source project, people can learn from it.

Why makes you saw it's a DDoS attack?

Can confirm that it's back up now.

Thanks Nat! I would also recommend writing a blog post announcing Discussions and its features as soon as you get the chance.

Reading this HN thread, I got the impression GitHub actually did just copy/paste Issues into a new tab, as a commenter mentioned, a belief that was backed up by a cursory look at Next.js's Discussions.

Now I see that's not fully the case.

GitHub bought Spectrum (spectrum.chat) in 2018:

https://spectrum.chat/spectrum/general/spectrum-is-joining-g...

But I would caution against integrating too many tools into GitHub. As a user, I like that GitHub errors on the side of saying no. You don't want to end up with a bloated product.