Feed on
Posts
Comments

An Inviting Puzzle

As you know, the lottery learning activity is one of my favorites. It combines modeling, situating, inviting, releasing, and even a little sorting — for most teams.

A commenter wrote recently about a situation where a team just isn’t ready for the lottery:

I have several members in my team who feel very threatened about any kind of review, and get very defensive when its suggested. For example, a while ago during a version control system usage discussion, I suggested that we might want to do ad-hoc peer reviews (for readability, not correctness) as a requirement for code promotions.

This met with a lot of hostility, a lot of which was due to egos being crushed in the past (people being told they were wrong or stupid, or being criticised).

Yikes!

Two Quick Coaching Lessons

There are two quick lessons just in this problem statement.

First: never underestimate the potency of prior fuck-ups, especially those marked by by weak or indifferent interpersonals. Coaches must work hard to bring a ‘newness’ to their ideas, and to their own personality, too, so that they’re not perceived in any way as “same as before”.

Second: building agile teams is always custom. Have no doubt, there’s no coaching action anywhere that is a guaranteed winner. Lottery learning has served me wondrously, but just from the short description above, I agree with Joshua that it’s the wrong choice for this team at this time.

Me, Having Uninviting Time

The Missing Invitation

It’s absolutely clear that my old buddy lottery learning is the wrong place to start with this team. It’s not only not inviting, but is maybe even actively un-inviting.

Joshua gave the good suggestion that maybe he should do a non-lottery learning: pick himself or some other gung-ho person on the team, project that code and talk it over.

That’s a great idea. Certainly the beginning of the inviting pillar is to realize that scary = unpleasant = harder to inculcate.

For best results make sure this chosen person is a laugher. Do you have a pro, including possibly yourself, who can get them all laughing at code together?

Also, I’d certainly make it a two-hour lunch+learning if I could. Food eases everything.

Devising Exciting And Pleasant Activities

There are at least two other responses which may seem more inviting to the team.

Study-Group Lunch is just what it says. A two-hour lunch and learn, on consecutive chapters of a book. If you want to keep the lottery aspect, do one at the *end* of a session, to choose the one who’ll present the next chapter.

What book? Almost anything less than 10 years old and having a lot of code and code advice in it.

DoubleDawgDare sessions are also fun, if you’re a strong presenter. Show them an untouchable file getting step-by-step into a proper frame. Arrange it so you can ask questions like “what’s my next move?”, as well as taking some ‘optional’ refactoring, where you ask them whether it makes things better, worse, or just the same.

But you don’t really need to hear my list of activities. Instead, let’s get some fresh blood in here:

Readers!

Tell Us Your *Inviting* Idea

For Approaching Internal Code Quality!


6 Responses to “An Inviting Puzzle”

  1. Anonymous says:

    See the dynamics laid out at http://agileinaflash.blogspot.com/2010/03/personal-objections-to-agile-process.html

    This is a good discussion and application. It is important to realize that the best of us are still, in some regards, broken children. Not to be condescending (all in the same boat) but to realize that past pains are closer than they may appear.

  2. Tim Ottinger says:

    See the dynamics laid out at http://agileinaflash.blogspot.com/2010/03/personal-objections-to-agile-process.html

    This is a good discussion and application. It is important to realize that the best of us are still, in some regards, broken children. Not to be condescending (all in the same boat) but to realize that past pains are closer than they may appear.

  3. OK, I’ll bite: a long time ago I worked in a COBOL shop that had pretty good code (we did ginormous-scale data processing for 911 telecom dispatching, so when you choke on a chicken bone and dial 911, the ambulance knows where to go), but no institutional code review.

    As a dedicated software tester, NOT a dev, I would routinely run new code in the debugger. A lot of my bug reports would look like “on line 751 of FOO_LIB, the variable BAR_STUFF is set to 10 chars but the UI allows a 20-char input, thus causing truncation of critical data”.

    I got to know that code base inside and out. In fact we had a lot of outsiders reading the code: the build guy saw all the compiler errors, the testers and the sysadmins all read the code in the debugger.

    If you have dedicated testers who know how to read code, they tend to be awfully good at submitting constructive criticism that allows for improvement and not rejection; their regular day job depends on doing just that.

    • GeePawHill says:

      Chris… Another huge point in favor of “development teams” becoming fully inclusive. — Hill

    • Joshua Lewis says:

      I’m nitpicking, but in my original comment I mentioned the reviews we want to do are for readability (i.e. maintainability) not correctness. I imagine/expect the type of issues you’re describing are more correctness or logic issues.

Leave a Reply

AWSOM Powered