The Frictionless Web: Gathers.us and the Reality of 2012 NoSQL

• UGtastic Archive
Transcript Verified
Watch on Vimeo Watch on Youtube
The Interviewer

Mike Hall

Interviewer, community organizer at UGtastic

The Guest

Ethan Gunderson

Co-founder of Gathers.us and ChicagoDB

The Guest

Ryan Briones

Co-founder of Gathers.us and ChicagoDB

The Conversation

Mike Hall Interviewer, community organizer at UGtastic
Hi, I'm Mike, sitting down again with Ryan Briones and Ethan Gunderson. These guys created Gathers.us, a website for managing user group information and events, and are also the founders of the ChicagoDB user group. Thanks Ryan and Ethan for sitting down again. We had a little technical snafu beforehand.
Mike Hall Interviewer, community organizer at UGtastic
So, what we discussed earlier is that Gathers.us had a pretty interesting origin, based off a need you had in creating ChicagoDB. Can you describe the inception of Gathers.us? I've used it myself, and I know GeekFest uses it; it's a great success with over a hundred sign-ups each week.
Ethan Gunderson Co-founder of Gathers.us and ChicagoDB
There were two major things that happened. One, we were too cheap to use a tool that costs money. We didn't want to pay for something that might not have legs. Second, a personal gripe of mine: most meeting websites force you to create an account just to attend an event.
Ethan Gunderson Co-founder of Gathers.us and ChicagoDB
Our big design goal was to make it dead simple: first name, last name, and email. That's all you need.
Mike Hall Interviewer, community organizer at UGtastic
Do you ever have problems with spam? I know you have a CAPTCHA out there now.
Ethan Gunderson Co-founder of Gathers.us and ChicagoDB
It hasn't happened since we introduced it.
Mike Hall Interviewer, community organizer at UGtastic
So, Gathers.us was born out of ChicagoDB. What database did you end up using for it? Considering you're a database group, that's a critical question.
Ryan Briones Co-founder of Gathers.us and ChicagoDB
We used MongoDB for no particular reason whatsoever—other than the fact that I won an argument.
Mike Hall Interviewer, community organizer at UGtastic
Did you learn anything in creating Gathers.us that you brought back to the group?
Ryan Briones Co-founder of Gathers.us and ChicagoDB
We never did anything with the group specifically on Gathers.us, but I gave two talks at MongoDB conferences about our data strategies. At the time, almost none of that was documented; everyone was just doing their own thing.
Mike Hall Interviewer, community organizer at UGtastic
So you were on the bleeding edge. You mentioned you founded ChicagoDB two or three years ago. What was the typical format of a meeting?
Ethan Gunderson Co-founder of Gathers.us and ChicagoDB
The reason we created the group was because we decided on MongoDB almost by flipping a coin. We had no idea why it would be better than any other database. Our goal was to design a group focused on deep learning—understanding the core of what makes one database better than another.
Ethan Gunderson Co-founder of Gathers.us and ChicagoDB
Chicago has great talent with both modern NoSQL and relational experience. We wanted to 'leech' off that knowledge. Every meeting had a face-forward presentation, but to start the night, we would read database white papers—Dynamo, Bigtable—to set the stage.
Mike Hall Interviewer, community organizer at UGtastic
Anything surprising you learned from those white papers?
Ryan Briones Co-founder of Gathers.us and ChicagoDB
The biggest thing for me was the realization that basically no new ideas have come out of database theory in the last 35 years. The Dynamo paper is just a group of 35-year-old ideas presented slightly differently.
Mike Hall Interviewer, community organizer at UGtastic
You also mentioned how specific concerns drove modern NoSQL, like Dynamo being the shopping cart back-end for Amazon.
Mike Hall Interviewer, community organizer at UGtastic
Covering these white papers is a large undertaking. How did you get people to actually read them month to month? Was that a challenge?
Ethan Gunderson Co-founder of Gathers.us and ChicagoDB
It was a challenge. Most of the time people would not read the paper. We didn't want to be the only voices in the group, but trying to get volunteers to lead a discussion on a topic they didn't truly understand was rough. A lot of those papers are written in an 'ivory tower' style with lots of math.
Ryan Briones Co-founder of Gathers.us and ChicagoDB
The issue we had with bringing vendors in was that they just wanted to convert you to their database. It leads to sales pitches. It's really hard to get a vendor to tell you the trade-offs you're making.
Mike Hall Interviewer, community organizer at UGtastic
How is ChicagoDB going into the future?
Ethan Gunderson Co-founder of Gathers.us and ChicagoDB
Interest has died out a bit, mostly due to specialized groups forming—Redis, Hadoop, Hadoop. People have segregated themselves by platform. I've been tossing around the idea of a general computer science white paper group instead.
Ryan Briones Co-founder of Gathers.us and ChicagoDB
I don't mind the segregation. The whole purpose was for us to learn, and I learned. That's all I care about.
Mike Hall Interviewer, community organizer at UGtastic
Well, thank you very much for sitting down, Ryan and Ethan. I don't want to keep you from your chicken.
Ethan Gunderson Co-founder of Gathers.us and ChicagoDB
Thank you.

Critical Insights

durable
"Gathers.us was built on a 'frictionless' design philosophy, intentionally avoiding user accounts to lower the barrier for community event attendance."
time bound
"The early NoSQL movement (circa 2012) was often driven by pragmatic choices and 'won arguments' rather than documented best practices, resulting in a 'bleeding edge' culture of shared data strategies."
durable
"Modern database innovations (like Dynamo and Bigtable) are largely implementations of core computer science theories established over 35 years ago."
durable
"Sustaining a 'reading group' model in a technical community is difficult because of the high cognitive load of academic white papers and the fear of presenting incomplete understanding."
durable
"Vendor-led technical talks often fail to address architectural trade-offs, leading community organizers to prefer peer-led 'white paper' deep dives for unbiased learning."