I've been looking into expanding my hosting resources the past 2 days and I've come to the conclusion that, in the long run, it's much cheaper to own your own machines, except for one tiny detail and that's the upfront capital cost (well that and admin/management costs).  But read on, as I think I have a solution to that very problem.

That solution is a internet/hosting/colocation co-op.  Now, I'm not the first to come to this conclusion, there have been others (such as http://www.hatters.org.uk/ and http://www.onlinepolicy.org/about/cccpagreement.shtml) and I think it makes a lot of sense.

For the next 12 months, I have a hosting budget of about $1800 and for that amount I can get a (small) bunch of VPS' somewhere or 1 decent enough dedicated server.  But I'd really like (at least) two separate machines for HA and safety.

For about the same money, I'll show how we can have 4 VPS' with 2-4G Ram and .5 - 1 VCPU each spread across two physically separate machines.  The capital cost (one time  - coop share price?) would range from ~$600-$1100 and monthly dues from ~$15-$65 depending on the number of people invovled (really it's shares in the co-op ie someone could buy two shares).

So what's the ideal number of people?  It looks like 6 would work the best for economy of scale and costs ($800-$900 upfront and about $25-$30 month), but I've also worked on (not detailed) numbers for 2, 4 and 12 people, but 4 and 12 share arrangements give up RAM and/or VCPU's as the tradeoff for lower costs. 

So, is anyone else interested in saving money on their startup's hosting needs?  Let's chat about how it would work.

I am happy to announce that my summer side project called EPS (Encrypting Proxy Service, a descriptive, but horribly boring name - suggestions welcome) is now ready for early beta testers.

EPS is a simple web proxy (written in Javascript using node.js) that sits betwee your app and S3.  It encrypts files being uploaded to S3 using AES256 and decrypts them when they are downloaded. It should work with any existing s3 clients (assuming they allow you to specify the host and post for the s3 service). It is intended to solve the problem of 'enryption at rest'.

Why?

In the last year or so I've had two clients that expressed the need to store files in s3 encrypted.  I suspect there are more and in the spirit of the Lean Startup I've built the core software and I am throwing it out there to garner feedback and to see if the need I perceive actualy exists.

How you can help

 The code is very young and while it's been tested and seems to function quite well, it needs some people to help test it out in more real world situations (I don't recommend it for production use yet though!).

So sign up for a beta account on the home page and we'll get you set up as soon as possible.

I know the biggest challenge we face is not technical, but one of trust, so please feel free to email me with your questions or concerns and I will do my best to answer them.

I was just reading Danny's VC is not broken - Anymore. Digital Media’s new funding models post and it's prompted me to write since I've been thinking about this kind of thing for a while too.

I'm a big fan of bootstrapping.  I've been bootstrapping Pennyminder for over 3.5 years (and I think I just might have finally gotten the product right with the upcoming release of 3.0 ;-). 

Raising capital is nearly a full time job and with the cost of building and launching 'Minimum Viable Products' dropping all the time, it makes sense to build out your prototype and launch it without outside capital.  This is the easy part (for me anyways).  The hard part is turning the prototype into a running business.

This usally takes some money for marketing and advertising.  At this point the company might do a friends and family and/or an angel round, but these two traditional sources of funds are outside the sweet spot, F&F too low and Angels a bit too high.  

I think that sweet spot is somewhere between 50-150K, enough to survive 6-8 months and validate that there's a market for your product (did I mention I'm a  fan of Bootup Labs? I think they have the right level of funding). 

If there is a market, you can probably grow it organically or you can decide to take a Series A round and go big fast, but you don't have to spend a lot of money upfront to find out.

More important than the money though, is the mentoring, contacts and network effects that come with groups like Bootup Labs, Y Combinator and Tech Stars.  Projects get press and exposure because they are part of those programs and that's nearly invaluable.

At Sourdough Labs we're trying something a bit different, more like that of Brightspark Studios, in that we want to build lots of simplist/stupidest things that will possibly work and let darwinism select which ones survive and those that don't, funding this with client work.  I think that by helping to build and launch many startups, both our own as well as clients, I can generate a network effect of my own.

My name is Vince Hodges and I am the founder of Sourdough Labs. This past year I launched three sites, Chipped.In, (relaunched) Pennyminder and SharableThings and I am working on another two that will launch later this fall as well as plans for the third version of Pennyminder (more on that in later post). I've gotten very good (and fast!) at building and launching web applications.  I would like to help you build yours.

I came across this post from Andrew Hyde (From TechStars) regarding the similarities between the Gold Rush and startups and what we can learn from them:  10 Lessons for Startups from the GoldRush. If the 1990's were the equivalent of the California gold rush, then now (roughly speaking) would be the Klondike gold rush. 

During both rushes, prospectors would carry a little sourdough starter in a pouch so they could make bread and eat.  During the Klondike Rush, old experienced prospectors were called 'Sourdoughs'.  Of course,  sourdough is also about bread and using a little bit from each loaf to start the next one.  All of these are a reasonable metaphor for Sourdough Labs and what I want to accomplish with it.

The mission of Sourdough Labs is to help build and launch new internet and mobile startup applications. For now I am a team of just one, but that will change. Check out our bootstrappers plan and then call or email me to talk about staking your claim.