We use ZenPayroll at Clearbit and couldn't be happier! Such a well thought out experience all round.

Here's hoping they'll start selling health and office insurance too so we can drop our other providers.

Dokku's definitely a young project and has its teething problems. We've been using a fork of it called Dokku-alt which is getting a bit more love these days - I'd give that a shot.

https://github.com/dokku-alt/dokku-alt

It's absolutely true that Heroku is a lot more than what Dokku, Docker & Vagrant provide, especially when it comes to auto-scaling. Heroku is a good choice for many - but it's also good to have alternatives. For us (at Clearbit), it made sense to build our own infrastructure for cost reasons, especially since we're using EC2's spot instances.

To your points though:

* Dokku uses Heroku's buildpacks and auto-detects your stack when you git push

* Dokku-alt supports collaboration & user-permissions

* Logging is automatically handled (dokku logs app-name)

* Support for Memcache/Redis is baked in (dokku redis:create app-name)

* Postgres is also supported by Dokku, but I personally go with AWS RDS.

> For us (at Clearbit), it made sense to build our own infrastructure for cost reasons, especially since we're using EC2's spot instances.

> * Postgres is also supported by Dokku, but I personally go with AWS RDS.

That's funny - I've costed it, and it's cheaper to run equal sized instances of a DB node than to run RDS, plus you get way more control over the tuning and implementation of your database layer.

Thanks for the recommendation Sylvain! Yes, we've been busy improving our data over the last month. Give it another shot - I think you'll be pleasantly surprised.

Afaik, Balanced don't have debit card payouts publicly available, they only have a landing page which is 'accepting applications'.

It actually works in our test API for anyone. Check it:

Make a card:

    $curl https://api.balancedpayments.com/cards \
      -u ak-test-2qEM0Znvd8LIVbZ01LbFHvHgab4fkNr3c: \
      -d name="Johannes Bach" \
      -d number=4342561111111118 \
      -d expiration_month=05 \
      -d expiration_year=2017

The 'credit' link in the response is that comes back is '/cards/CC2mEWYlYvod2E1w3TJHOiGh/credits'. So we push:

    $ curl https://api.balancedpayments.com/cards/CC2mEWYlYvod2E1w3TJHOiGh/credits \
      -u ak-test-2qEM0Znvd8LIVbZ01LbFHvHgab4fkNr3c: \
      -d amount=10000

We will turn it on _in production_ for people in the beta period. Some customers have already integrated it in test!

(Balanced co-founder here)

You're right.

Our API has been built, but we're giving private beta access to anyone that made prepayments via our crowdfunding campaign [1] first. Private beta will launch next week [2]. We want to make sure the performance (instant payouts & high coverage) is on par with our standards before a public release, which will happen later this summer.

[1] https://balanced.crowdhoster.com/let-s-push-to-debit-cards [2] https://www.balancedpayments.com/push-to-card

I'm am extremely excited about Chain.com, and the other tools being build up around the Bitcoin ecosystem. The power of giving developers simple tools and building blocks is often underestimated - but it often results in radical innovations.

A case in point is Stripe's API, and the ecosystem that has been subsequently built around it. Yes, it's possible to send CC auths/settlements and ACH transactions yourself - but the barrier to entry stifles innovation. Often the most interesting innovations in tech occur when there's an innovation in tooling.

I'm particularly interesting in Chain.com though, because I think currency is only the tip of the iceberg when it comes to Blockchains. I wouldn't be surprised if we see a number of other use-cases for Blockchains soon. [1]

[1] - https://www.igvita.com/2014/05/05/minimum-viable-block-chain...

It sounds like you may be eligible for the O1 visa.

I thought of exploring that, but TBH, my understanding is that you have to be a "publishing machine" to even stand the chance of being considered and unfortunately I haven't been as productive as I had wished during my PhD in terms of number of papers. Maybe I should investigate this further, but the original article also seems to indicate that an O1 is a long shot.

Hi, I'm the co-founder of Sourcing.io, and the engineer who built it. Stoked that we're on HN, and get the chance to respond to the engineering and hacker community.

Our introduction blog post says a bit more about our mission.

http://blog.sourcing.io/hello-world

Believe it or not, I hate 'spam' recruiting emails as much as the next engineer. This is how I feel about them:

http://blog.alexmaccaw.com/recruiting-emails

We're not here to increase the spam problem - quite the opposite. We've put a number of safe guards in place against this kind of abuse:

* We don't allow big recruitment agencies known for spamming on the system.

* Our system is geared around referral based recruiting - finding people already connected to your company. For example, promoting candidates who follow one of your engineers on Twitter. This is the type of recruiting we live for and want to promote.

That said, there are definitely some things we should (and will) do.

* Allow developers to see, edit and update their own profiles

* An instant opt-out button

Let me know if there's anything else I can answer.

I appreciate this idea leveraging data from Twitter and GitHub and truly believe that's the better way to find professional engineers which are associated with your community somehow. What about intelligent algorithms to evaluate and to display engineer skills (points or school grades)?

We actually originally started generating a 'score' for engineers, and it's still part of our API. However, we dropped it from the interface since it was confusing a lot of people, and the score tended to reflect how popular an engineer was, rather than how good they were. As you can imagine, figuring out the latter is a very complex problem. For now, we've removed it.

That's true. It's nearly impossible to evaluate accurately how good an engineer was, but for me personally it seems to be useful to display similar informations to companies. Something like "geek cred" (coderwall.com) or open-source track record, because building many projects is one of the key-factors of experience. Coderwall published that they work with statistical methods like central tendency. https://coderwall.com/blog/2012-02-05-the-companies-id-want-...

Looks good, and I like the mistakes and solutions section.

However, it's pretty much all covered by Stripe's open source jQuery.payment, which is also agnostic to your payment gateway.

https://github.com/stripe/jquery.payment

(Disclaimer - I built it.)

The presentation of Creditcard.js is great in terms of styling, and I'm sure the practical example goes a long way in selling something like this to non-developer decision makers. This is probably the market they are aiming for.

Personally, I would much rather use the MIT-licensed jQuery.payment by Stripe. I think $300 or even $149 is prohibitively expensive, and would prefer a script that is actively used and reviewed in public by many developers.

Thanks Alex!

Agreed. I wouldn't be opposed to paying $150 for a form that would take me a while to build, but there's already a lot of good open-source solutions for card inputs.

"prohibitively expensive" ?

Surely if you're a business of any merit – making even $500/month – $300 for a javascript library that'll remove bumps in the checkout process is worth it?

I fail to see why this cannot co-exist alongside the Stripe alternative library. Heck, just taking the stripe library and offering commercial grade support, and charging $300 seems like a pretty good business.

>I fail to see why this cannot co-exist alongside the Stripe alternative library.

It certainly can, anyone is more than welcome to pay for something done by open source software available for free.

>Surely if you're a business of any merit

Just because you're a business doesn't mean you should pay for things that you can do cheaper.

The point is that a decent web developer could recreate that form in 20 minutes with jquery.payment and a little bit of CSS, so why should you pay $300 for the same thing?

Can you explain how commercial-grade support for a library that hasn't been tested by many people is better than a community of people who have run and tested a library? What exactly are they going to do for the support?

I'd also argue that a library developed and used by Stripe is pretty commercial grade too.

Try checking out -- they're are still plenty of bumps in their checkout process...

Thinking price point, 300 is about a couple of hour's worth of work which they save your developers. Not a big deal.

And you have a similar UI widget at

https://stripe.com/docs/checkout

Lol, that looks exactly the same to me!

It's pretty hard to make the most minimal form look any different.

From the initial docs of jQuery.payment, it looks like a lower-level tool -- the OP provides the promise of "you just drop it in, and you get these things, designed for you." jQuery.payment looks like "Here's a toolset you can use to do whatever you want!"

But I don't want to spend time figuring out what I want and putting it together. I also don't want to spend $150/year on it though. :)

If you built a higher-level abstraction on top of jQuery.payment, that made the good design choices for the developer, so the developer could just 'insert form here'....

You guys at Stripe are amazing. I love you guys.

Or Saigon - had a great time there.

http://blog.alexmaccaw.com/traveling-writing-and-programming

I'm actually in Saigon right now! Do you have any recommendations for spaces to work? I am trying SaigonHUB on Monday.

Hmm, I was usually working out of my hotel room or cafe's along Phạm Ngũ Lão street.

Vietnam and cambodia are just shit compared to thai.