Visible Certainty - The Verifiable.com Blog

Apr 13 2009

What a long strange trip it’s been (part deux)

Hi Everyone,

The last time I blogged about what a long strange trip it’s been was July 23, 2008, when we launched the public beta.  We have been working like crazy, not just on bugs and features, but on the whole production environment. You may have read our blog entry: Public launch of Verifiable.com!  And wondered what is this thing you call Verifiable? :-)

Since everyone is pushing for transparency from our startup culture to our President, I guess I should do my bit as well…

So using my handy dandy time machine I will take you all back to the end of July last year:
1. we were visiblecertainty.com
2. running out of the amazon cloud.
3. We didn’t have joining, new variables, or small multiples.
4. We had an old 2001ish type layout and interface
5. We opened our public beta to see how all you folks liked it.

We discovered that:
1. No one could spell visible certainty with any consistency (“visual certainty?”,  ”virtual certainty?”, “visable certainty?”). It would have been funny if it hadn’t driven us so crazy.
2. The amazon cloud is expensive and for our application somewhat underpowered (hey, don’t hate. we aren’t the only ones who say so)
3. People like joining, new variables, and small multiples
4. Engineers are not the best designers so we hired some experts
5. With the explosion of data availability and the many demands for transparency, our mission is becoming more and more vital 
6. Our rich media UI is really really hard to test

So we decided to:
1. create a new product name:
Boy was this a long sweaty ordeal.  Our spec was that we had to have an available domain (a dot com not a dot net or a dot whatever), needed to evoke the spirit of our corporate mission, and be somewhat whimsical.  For your enjoyment or our humiliation here were the runner ups: dataink.com, cogister.com, datadunk.com, datasofa.com, datatribune.com, presentfully.com, charti.st, datasoup.com. I still have a deep fondness for datasofa and datasoup both of which were suggested by my 3 year old daughter who to my momentary irritation “interrupted” an adult conversation to tell me something crucial and timely.  In the end, we decided to give up whimsy for a name that once we chose it, seemed to shout out to us to work harder and get this product done.  On the other hand, I think my daughter is still slightly pissed that we didn’t go with her suggestions.  Anybody out there want a slightly used domain name?

2. Move out of the cloud:
Boy was this a long tough grind (are you seeing a trend here?).  Our spec was extensibility and expandability and not too much money.  I’ve never had the deep felt need to live near the colocation facility.  In my first business, the computer (a half-refrigerator sized SGI box- remember those) ran out of a closet near Mae-East (does that even exist anymore) in Vienna Virginia and I didn’t go visit it for years at a time. So we chose Wisconsin as the unlikely place to look for colocation. If you must know, we chose big company over small (feel a little bad about that) and HP over Sun (don’t feel bad at all about that). Then we embarked on the task of inventing a whole new infrastructure for creating and fielding instances (only Peter and JY could have done it with such aplomb).

3. Lot’s of flex coding:
I’ve started writing GUI code in the last 4 months, something I have avoided my whole professional life. And for the record, I have no idea how JD and Kris managed to write all this code.  It is a wonder to me.  And I am very thankful that I fought the rich media user interface for all those years, oh my god, is it a lot of lines of code and a lot of work and sooooooo hard to test.  But, here we are and I think that these guys have really done a remarkable job.

4. Hire a design firm:
We hired (actually JY hired) a design firm to figure out how to make our site easy to use.  I have always been a put a couple of buttons on the page and put clear labels on them kind of guy, but JY was so adamant that I figured he either had to be right or would bring a gun to work if I didn’t agree (admittedly when everyone works from home bringing a gun to work is a lot less of an issue for co-workers).  But I love his wife and kids, so I agreed to the plan.  Frankly, after seeing how great it looks, I must admit that I was an incredible jerk for dragging my feet.  A shout out to Juhan Sonin (of MIT <—- yes aren’t we coool) for heading up the design work.  The strangest part of the process was when the project manager of the design team introduced us to Juhan and it turned out that JY and he had gone to Lexington High School together.  Of course, the design is just the first step and would have gathered dust without  Kris and JY’s brilliant implementation.

One last thing: this company is filed with adults; adults with lots of pulls and pushes on their time.  I want to remark at the impossibility of raising a young family, taking care of your relationship with a spouse, and doing this kind of work.  If you get to work early you miss the kids eating breakfast, you miss walking them to school.  If you stay at work late you miss having dinner with the family. This is all a little different if you work from home, but there is enough heartache there  that it might even be harder.  For instance, you are working on a particularly thorny coding issue and your  small child walks in wanting to get in your lap and type a little.  What do you say to that?  Or, imagine that you have a spouse with a real job that comes home in the late afternoon expecting you to be available and present.  Yes, it is impossible.  Yet somehow all these guys have managed to do it and do it really well.

That’s it for now,
I have some press releases to write,
Some flex code to break,
Stuart

Comments (View)
blog comments powered by Disqus
Page 1 of 1
I want my StackExchange