I couldn't agree more. Graphite is awesome, but the state of its packaging is atrocious. That's probably the biggest barrier to its' widespread adoption, actually. I spent 2 days making my own Graphite packages and automating the install via puppet, which was worthwhile, but shouldn't be necessary.
Running Samba exposed to a public network isn't the only issue here. Even on an internal network, this would allow file/directory permissions to be bypassed.
lxc is great for something lightweight that is built into the kernel now. If you need something heavier or with more advanced features, use KVM or Xen.
I love ideas that can shift a paradigm. Adding the idea of "impact investing" as a missing stage between philanthropy and traditional investments, seems like it could do just that.
Sometimes the simplest "technology" can make the biggest difference in the lives of millions. Another link about it that has a few more details, including that it generates 55-60W of light, 10 months out of the year.
I used to work at WorldLingo (http://www.worldlingo.com), who offers a translation API service. They're one of the few machine translation API services that are publicly available at reasonable costs. I'm happy to answer questions about the translation industry as a whole, shoot me an email if you're interested.
Are you using IPS panels? They seem to work much better in portrait than TN panels do. (I have 3x 24" IPS panels in a portrait-landscape-portrait configuration at work, 2x 24" TN panels in landscape-landscape at home.)
A few things I initially thought of, that seemingly came out of nowhere and changed the world, were -- the steam engine, lightbulb, automobile, radio, television, zippers/velcro, plastic/steel, various aspects of medicine (antibiotics, x-rays, etc).
Use AWS until you have a reasonable grasp of your dataset and real requirements. Then buy whatever provides the best bang for your buck, in terms of servers. That will probably mean getting 6 mid-range servers, rather than the three servers with the absolute fastest CPU/most memory available. Use either RedHat (CentOS) or Debian, and you'll almost certainly be using Hadoop. Dell servers are fine, although you can sometimes save significantly by going with something like Supermicro servers from Newegg. In terms of cost, you'll want to order the bulk of your servers' memory from a third-party, not have it included in the build.
One strategy would be to initially continue to offer it as a "private beta", while allowing X invites to each existing user to expand the beta userbase.
Contribute to your 401k through your employer up to the match level, and open a Roth 401k to contribute an additional $5k/year there. Build up 6-12 months of emergency fund savings. Max your traditional 401k to $16,500 per year. Invest in "target year" low cost/fee mutual funds to get started.
For monitoring, Pingdom has a basic free account with uptime/response time monitoring and email notifications. If you host your nameservers and email (Google Apps?) elsewhere, losing a web host isn't quite so bad. I also unfortunately just went through this with my personal VPS a few days ago, with Hazenet disappearing off the face of the planet.
Big data analysis is currently hot, and will almost certainly only get hotter. You can apply big data analysis in many areas, but only a few have been tapped so far. What used to be $CRAZY_EXPENSIVE amounts of computing power can be leveraged to provide new insights into existing fields. In a word: Hadoop. Combine that with your NoSQL tech of choice (HBase, Cassandra, Riak, etc).
The reaction timer was interesting -- I managed to get 0.16s in one attempt out of ~15, with most attempts ranging from .25-.35. It reminded me of exercises they had us do in driver's ed in high school to show how reaction time+braking distance was affected by alcohol.
This would be awesome if you could combine the list with availability of <xxx>io.com -- to know when you could get both the shorter <xxx>.io and the traditional .com
Co-founders living together often makes sense economically, but otherwise I would think having the ability to spend some time apart would be good - occasionally/regularly/as needed.
Burning through four developers in six months, with the latest starting at the beginning of January and leaving at the end of February would seem to indicate a significant problem internally.