Quite interesting!

At first, I didn't understand why you only wanted a 200x200 image. I expected it to cut a big picture up into smaller squares and lay them out in a table.

But I was wrong! :)

It actually creates a 200 row by 200 column table and changes the background of each cell to match the image's color at that pixel.

I'd recommend improving the description of the service on the site. It took me a couple tries to get it to work (I used too big of a picture), but finally got it to work. Glad I did.

Thanks for sharing.

What are the roles of each person? Is the younger person the hacker and the older person the bizdev? Or vice versa?

If the older person is the bizdev and has a lot of domain expertise, then it's a HUGE advantage!

The problem I think you are trying to solve isn't much of a problem.

I've been developing with CMS applications for 12+ years and have never once ran into any of the issues you described above (failed upgrades, hackings, site-jackings, buggy plugins) to such a degree that I wanted/needed to completely switch the CMS platform.

I'd urge you to NOT build a product for developers. I know it's tempting because the ideas we (developers) come up with follow the 'scratch your itch' advice. I wrote a blog post about this: http://www.travisdoes.com/you-itch-sucks-scratch-somebody-el...

I'd highly recommend you focus your efforts on building a B2B SaaS product. There are millions of great opportunities to build boring, but profitable niche products.

If you need help finding some ideas, shoot me an email.

First of all... good blog post. Money quote: "Your universe is small and very hard to sell to."

Secondly, I hope I am not misunderstanding your comment, "... completely switch the CMS platform." While it may not be clear in the slide, this is not a new CMS. It is a backup and security service for popular CMSes.

Secondly, I am not aiming this service at other developers. It's partially omitted in the slide but the market segment is administrators of sites that receive >15K uniques per month.

Administrators != Developers. 15K+ uniques per month implies traffic is high enough to monetize which means uptime is critical.

Does that alter your opinion? Or generally, are there any parameters on the slide that could be tweaked to reach a threshold of viability in your opinion?

Next time... :)

Make sure all co-founders' equity is vested.

Don't do a 50-50 split. One person needs to have a majority stake. Otherwise, you spend too much effort being a democracy instead of a startup.

It's funny how obvious some things are in hindsight. I never once thought there'd be an issue with a 50/50 split; but yes, no one person has power of veto. Interesting about vesting equity though, I'd not heard of that before (naive much?).

There's a twist to the tale though: we haven't incorporated a company yet. No shares have been split; we've just got a verbal agreement that we'd divide them up that way. If I wanted to do a Zuckerberg, I could quite easily register the company and put myself on as the sole director.

If you give 40% to a designer, how much are you going to give to your biz dev person (when you need one in three months when your beta is done)?

If you both use equity to give to the biz dev and the biz dev requires more than 20% to retain, you just lost majority of your business. :(

40% is way too high for a designer.

Also, make damn sure you vest any equity you give out.

Remember, you can only give it away once, so do so sparingly.

This raises the question of whether you need a co-founder or a hire. It depends on the business.

IMO "40% is way too high for a designer" is too general. It depends on how significant a part of the leadership team they're going to be. Since you say you want them to "help out on the front end" - I'd lean towards guessing that you might need a hire, not a co-founder.

Also a good point. I wasn't too specific but I was thinking more someone with previous startup experience and also at least a little bit business savvy.

Good points, thanks. I guess I was just trying to figure out what is common since this is my first time thinking about finding a co-founder or two.

Car-as-a-service. Nice.

Build and install a scaled out application. Multiple AWS instances in a load balanced configuration. Use small instances to purposely bottleneck the available resource pool.

Use something like Blitz.io to beat the hell out of the application. This is your baseline

Use performance monitoring and optimization tools to improve the architecture and application. Go through each layer of the application to see where the biggest gains can be found. Also look/test for data consistency.

Rinse and repeat with blitz.io.

Also, test how well the architecture responds to various types of server failures.

And make sure you test a viable backup process as well.

Lastly, once you have a nice, performant architecture, increase system resources both up (more resources per server) and out (more servers) to see how well the system actually scales. :)

    Build and install a scaled out application. Multiple AWS instances in a load balanced configuration. Use small instances to purposely bottleneck the available resource pool.

But where do you learn how to install a load balancer? I know I can google that particular example. But as with learning anything new, the big problem is when you don't know what you don't know.

Agreed.

Based on lots of experience dealing with off-shore developers, the problem has always been communication.

Once a company figures out how to properly communicate the right level of detail, outsourcing becomes a significant force.

The opportunity gap of $15/hr versus $50/hr is too great to ignore.

Nice work.

Glad to see somebody implementing a solution to pretty up QR codes.

Only potential issue is the contrast between the foreground and background needs to be enough to properly scan.

Way ahead of you! We test it server-side to check if it scans, and if it doesn't, we alert the user. If the user doesn't care about the successful scan, then they can go right ahead and use it anyway.

Check out the specials listed on WebHostingTalk.com.

I use unmanaged dedicated servers.

Going price for a dedicated Quad Core i7, 12 GB of ECC RAM, 4x2TB is under $150/mo. That's a wicked cheap price for those specs.

Obviously if you are going from $10/mo hosting, you won't bite the bullet on a $150/mo server, but it's the only price point I know.