October 9th, 2013

IMG_5123_2

Once again, thank you to Colin for taking notes.

Topics that made the cut:

  1. Data nerds and the (un)natural world
    • Tracking data
    • Examples: Apartment prices in building, MAX Car #s, License plate photos from all 50 states
    • These are unnatural world
    • Like stargazing and philosophy
    • “boardgamegeek.com” games played database, bridge, chess, Magic the gathering, scrabble tend to be “solo” games
    • Given a data set how do you find a signal in there?
    • “Kids game”, I have a device that easy to capture, fun, wasn’t seeking data
    • Idle curiosity, liking to measure and capture
    • Similar to days before, shell collections etc, collecting comes first
    • Technology captures
    • Quantity of data
    • Journaling events in calendar, e.g. sleep in google calendar, morning routine etc, corrected as it happens and tweaks (why do we have empty spaces on calendars?)
    • becoming useful, e.g. density of out of state drivers
    • Act of collecting changing your view (vs only collecting for a reason or goal)
    • Collecting is noticing, becoming aware
    • Collecting and Mining, sometimes can’t tell you why the other is important, but doing the other gives you a new perspective
  2. How to learn more Git
    • Git is powerful
    • Easy to screw up a checkin
    • Live in fear of so many comitters
    • We should have git competency, aptitudes that everyone should have, and an advanced set of skills that someone in the group should have
    • One in the “group” vs one in the “building”?
    • Could have Git “moves” algorithm to do things safely, recipe of safe ‘moves’
    • Safety vs correctness
    • Problem is git safe stuff is right by unsafe stuff
    • Commits every 5 minutes
    • Source control complexity, much comes from the idea of “storing unfinished work” (git has stash, branch, other tools)
    • Software development is chaos, discovery, exploration – Is that inherently problematic? Not modeled properly in source control?
    • Short branch and frequent integration, a long branch is a bad branch
    • Could pick a core set of git practices that favor good behaviors (How to communicate this?)
  3. Probes for Complex Problems
    • How do we deal with problems? not sure of scope, don’t know the right way through etc.
    • How do you wrestle with them?
    • Being ok with failing
    • Mystery problems, not sure why it’s broken.. collect more data? Get a better view of what is going on?
    • Agile – Try a solution, iterate, ship frequently
    • “what do I NOT know how to do?” Tackle the unknowns first
    • Top to bottom solutions, a “spike” all the way through quickly
    • breaking problems down
    • Go for a model problem,what embodies the hard part of the problem, but is simple in other respects? (solve a smaller problem first)
    • Learn the parts, relationships
    • “I don’t think there are any problems that can’t be broken down”
    • How is a probe different than a solution? A probe is a test for a solution.
    • Your probe is based on assumptions that you make up front, you form a hypothesis and test it
    • Why are things the way they are? What would make that be wrong? Can we test that?
    • Gaming – solving real problems (e.g. the protein shape puzzle)
    • “Your problem is you don’t understand your problem” get used to it, play with it
    • “Mathematicians use “play” to mean very serious work” “and software engineers use “work” to mean very serious play”
    • Divide and conquer separates you from all the pieces, you get stuck in a reductionist mode)
  4. The Gervais principle
    • (e.g. Ricky gervais)
    • Pyramid: Losers on the bottom, clueless in the middle, sociopaths on the top
    • Blog post on this problem
    • This describes a dysfunctional organization well
    • People choosing to do what is dysfunctional because it’s comfortable
    • Feedback from middle based on politics and yes men
    • Non-delivering political environment
    • The labels are very judgmental (and humor based), but the problem is real
    • Clueless band are highly loyal much more than the other layers are
    • Tribal leadership, 5 stages of evolution, cultural evolution
    • L1: Like in prison world sucks, L2: My life sucks, L3: I’m great but everyone else sucks, L4: Were great, L5: The world is great
    • Has to do with the way leadership make tribes, or tribes of tribes
    • Instil a sense of empowerment
    • Pyramid with positive layers, building it up without falling into above is a great achievement
    • They may have a lot of problems going on, but they are forgiven for achieving great things
  5. What is the most significant change in software develop in the past 10 years?
    • Google
    • Stack overflow
    • Accessibility
    • Leveling sharing we’ve talked about forever is being delivered on (with the rise of the internet)
    • Huge repository of knowledge
    • You can almost always find someone else who has had a particular problem before, and what they tried or solved
    • Not just sharing and size of knowledge base but searching and filtering
    • We still really haven’t adjusted our model to realize
    • We don’t assess ability for new hires to rapidly acquire knowledge (habits in hiring that are hard to break)
    • Let developers decide who to hire, #1 question – Who do you follow online (e.g. where do you get info)
    • Another thing that we see, now you work in 3-6 languages at a time, because you can. (this wasn’t true before, couldn’t keep that much in your head)
    • Right sized lego blocks
    • “has this been done before, and can I find it?”
    • If I learned more I can contribute back and say how I found it, blog post on my outage, error, etc. Culture of contributing knowledge.
    • Google never confronts you with “that’s the wrong question”, searching is big skill, what terms? What question? How far back?
    • How far back do you have to go to figure out you’re doing the wrong thing?
  6. Can we run our government like wikipedia?
    • Wikipedia runs interestingly today, some problems
    • How does wikipedia run?
    • Every kind of misbehavior is understood, never destroying information, transparency and history
    • goal to inform could be solid
    • People can find the media that agree with their world view, (confirmation bias)
    • How strong is special interest influence, power seeking?
    • You don’t have a total budget for pages on wikipedia, unlike the national budget
    • Decisions made impact people directly, unlike wikipedia info, may no agree with you
    • Heart of the question: Can we have a governance system that allows transparency? Feels like not yet.
  7. How to get to ‘Done’ with an employee communication problem?
    • Recently had a long time employee leave due to reorganization of administration and shifting roles
    • Changed role of a person who had been around a long time, cause them to be unhappy
    • Mediation didn’t work
    • Most people are thinking about what they are going to say in response
    • Need to JUST LISTEN – True listening is a hard skill
    • There are things you can’t say, things that are inappropriate, business doesn’t allow the directness
    • Bringing up specific grievances instead is problematic, because they can be addressed.
    • Need to have other people who can take in the vague input and divine the direct problem
    • One of the problems – coming with focusing on “how do we get this done?” “how do we solve the problem?”
    • The thing needs to be allowed to NOT be done in order to get to done
    • When people do a job change this affects them negatively, you get a negative curve for a while, you need to help them settle out. You have to be checking in with them, face to face conversations
    • Working with introverts, have to elicit feedback (and respect it)
    • Change cycle is similar to grief cycle (“my job has just died”)