There's no such thing as a "technical co-founder". There are founders and there are employees. Usually but not always, one or more of the founders is capable of building the first salable version of the product. A hoary cliche among startups is the company that has matured to the point where its dev team removes commit privs from the founders, which should give you an idea of what the bar is here.

Algorithms, database design, version control, and deployment are all important things (design patterns much less so), but they don't matter a whit compared to building something that someone will pay money for.

Stop procrastinating and get to work.

Read the Lean Startup. Then build something. Who cares if you don't have the answers to problems you don't yet have? Startups are businesses first. If you make a product people want, even if it's badly engineered, that's a much better situation than being an expert hacker with an awesomely engineered product that nobody wants. If you have traction, you can hire people better than you to fill in the gaps.

Besides, nobody gives a shit about algorithms. ;)

As a long time programmer, I've learnt the mantra "somebody has already solved this, let's go find their solution." That applies to pretty much any technical or non-technical issue that you need to solve. So much of what we need to know is written up somewhere on the internet, leaving us to focus on the 1% of the problem that we alone can solve, and differentiates our value proposition from all the other alternatives.

My short answer would be "know enough to be dangerous, and surround yourself with smarter people".

Recommended reading - http://www.sacred-texts.com/nth/tgr/tgr10.htm

Start from 'Henry Ford was called "an ignorant pacifist."'

Lots of good feedback already, but one thing I'll add is that you'll also need to be a hell of a recruiter, which can eventually also take a bit of the pressure off (assuming you're surrounding yourself with people smarter/better than you)

Actually I am in the same position you are in. I think the best way to tackle this problem is to see if you can come up with a minimum viable product(MVP as they say) to validate your idea. You can do it either by developing yourself or hiring a programmer(Of course you have spend some money on it). If you see the product idea works and you have some traction, then you can always raise a bit of seed funding and hire some coders to do the development work.

But I think product/software design is only half of the battle. The other half is to find customers, convincing them to try your product and ultimately make them to pay for it.

From my experience and from many successful technical founders, or co-founders, you're either not qualified or just qualified for it. Even if you're advanced in specific areas of programming/computing, being a technical co-founder means doing everything related to the product (I'm guessing it's software). Deployment, database, web app, version control, services, server ops, etc... is a lot of stuff to learn and become proficient with it.

Don't be worried about the sheer overhead of skills or knowledge required. Startups is all about learning, maybe just as much as executing.

Yep, I totally agree. From my (a technical co-founder) point of view - the JFDI approach is a master factor here. If you don't know how to do something - that's good! Take the second step and get to know how you can do it, then double check how to do it right and the just do it. Plus: take the responsibility for what you've done.

Before I started building my first web app in Rails, I didn't even know how to write Rails! I had only previously been a front-end developer, but I can say I learned a TON. Am I a Ruby on Rails guru? Nope, but I know how to traverse the stack with the knowledge I do have and get as much done as possible.

I think the easiest way to figure it out is to try and make your software in your spare time. If you can make progress every day, you'll be fine. An objective measurement of your ability only becomes critical if you quit your job to work on your startup fulltime and have a hard deadline. Don't put yourself in that position.

You're good enough, no need to sure-up.

Anyone who even has "design patterns" as a blip on his radar can crank out all the code required for most startups' MVP. It will take you longer than it would take someone else, and that's okay. It won't always be pretty, and that's okay, too.

Good luck!

Better to launch and see if people wants the product than being wasting effort for nothing. As it is stated on the Lean Startup book, any non-validated fact is a waster of time.

The best hackers i've meet there were, "doers".

>How do I know I'm a good enough hacker to start a startup?

You try it and see.

nobody is qualified to be a co-founder unless they've done it before.

it's a test that you pass by doing.