I can’t believe three weeks has gone already and what do we have to show for it. Quite a bit really, but not much in other ways.
It’s really interesting to reflect back on the thoughts I wrote for the end of Week Two. Self-doubt creeping in — check, here in spades this week. The daunting size of the challenge? Check, it gets bigger the more we learn about it. Re-focus for a hands on version? Yes, did that too.
One thing I’ve already come to value is our Monday mornings. We spend a good couple of hours talking and thinking strategically about what we are up to, planning the week and the milestones. It’s a great time to do it actually because on the weekends, we usually manage to step away from the computer for a bit and reflect, so come Monday we are full of questions and ideas. It’s good to tackle these and get on with things.
This weekend the self-doubt appears to be back again. There are two crucial questions — will the Engine work and even if it works, does it do anything people really want? Of course the answer during the week when you’re wheeling and dealing, coding like a demon and focussing full-time on it is “absolutely”, but at 3AM in the morning a small door opens up in the mind and you really wonder what you’re doing and why.
Last week, the Monday session attempted to tackle this by taking the Darebin milestone and re-focussing it towards something that people could “get their hands on”. We’ve decided to build an automated Twitter List bot — it uses HiveMind to work out who demonstrates the expertise you want to build your list around and maintains it automatically. I know this sounds a bit of a tangent, but it does deliver a few things we think are important.
- Proves the engine in a production like environment.
- Provides a hands on demo customers can actually see it working.
- Lets us access the taxonomies that people build to see how the actually use the Engine.
- Done right, should lead people back to our site — it’s broadly targetted, but we know a lot of E2.0 types are on Twitter and may be interested in HiveMind.
Interestingly a lot of people are excited about this idea so it will be fun to see where it goes.
We also met with a lot of people this week and did a few social events as well. It was exciting to sit down and talk with people about HiveMind. The analogy on where we are at however is it’s somewhat like a car salesman talking to people about a cool engine (which is what we’ve got), unfortunately most of the customers buy cars (which we are still trying to decide exactly what car we put our engine into). That said, we’ve now got one customer willing to hot rod HiveMind himself and another who’s taking us to meet their developers soon. Maybe you can sell an engine without a car?
It’s this point that leads to the self-doubt however — engines are hard work, but to be blunt, the car (in this case the website that sits around the outside) is even harder and needs a lot of things (security, rights, permissions etc.) which aren’t really core to the engine at all. We’ve asked around a lot and got a lot of validation on the vision, at some point there is a leap of faith needed to just build the car. Still for now we are continuing the market research and trials a bit further.
Perhaps the biggest highlight of the week was the blog article on NextWebAU http://thenextweb.com/au/2009/12/03/hivemind-revolutionary-expertise-discovery-engine-enters-closed-beta that was a significant milestone and one that drove around 40 beta sign-ups since it was released. We certainly saw a bit of interest from around the globe. It drives the need for us to get our own look and feel now which has been on the to do list bit we haven’t hit the go button yet.
Alex and I have fallen naturally into some different roles and I think are working very well together. He’s the gun engineer and churns out code like crazy, constantly re-factoring the engine and improving it and the things around it. To let him do this, I seem to be getting the “curly” questions — given I have to research pretty much everything I do on the technical side anyway (my coding abilities being a lot weaker than Alex’s, although I’m improving) I may as well research the problems that are going to hold us up. This has good and bad sides. I feel like I’m contributing, but if you give me a technical problem, I get like a dog with a bone. I can’t give it up and get sucked down the rabbit hole.
Let’s just say the problem I hit this week was solved with detailed reading of at least two Internet protocol specifications, a HTTP packet sniffer, reverse engineering calls to another site, a LOT of reading and eventually re-writing the third-party library I’d picked up from scratch myself. Boy did I feel good when I nailed it this morning though!
Stress (I think it was the bugs) got a bit out of hand, although I think the cold contributed too, I woke at 3AM on Friday and worked from 4AM till 9.30ish, then again on Saturday, woke at 1AM and managed to stay in bed until 6AM, but slept very poorly. Tonight however, it’s time for me to hit the sack and having squashed the bug, I feel great. A bit of exercise yesterday and today really helped clear my head too.
I wrote this on Sunday evening so Alex could read this as well (he only checks feeds once a week — he’s very good at staying focussed), but I think I might have it a bit late, he’s also more disciplined about going to bed earlier!
- Kicking off the Alphington sprint and feeling like we might soon have a product around the engine that people can interact with.
- Meetings with lots of great people who have all contributed a lot to our thinking and knowledge.
- First commitment to test the engine in a live customer environment.
- Press release and the surge in beta signups.
- Don’t talk to me about Open Auth protocols (OAuth).
- Don’t talk to me about HTTP Verbs and bad library implementations.
- Being awake at 3AM two mornings in a row.
- Wasting too much time — not “letting things go” and moving on quick enough some times.
- Lack of time to do everything we’d like to do as well as we’d like to do it.
If I could change on thing / goal this week?
- Focus on the bare minimum to get the Twendly / Alphington release out the door and live to customers.