[Been occupied with packing and moving since my last post until a couple days ago, though more packing and moving needed. Hopefully I'll get on a daily posting schedule before too long.]
As with the others, this list of Twitter tweets is sorted by first name. (Hmmm...a thought just struck me. By posting these collections, am I engaging in mass retweeting? Guess not since it isn't on Twitter. But there must be a name for this...or should be, I guess.)
Alan Shalloway - people are often confusing predictability with control. if you do, you won't believe you can control your dev process. being in control is not a bad thing. being controlled by someone else is.
Alan Shalloway - the more i work w/ lrg orgs (300+ devs) the more i believe that any approach that focuses solely on the team or is bottom up is doomed. top down does not mean top down control. but rather top down enabling of working on the right things in the right way.
Alan Shalloway (series of quotes) - Issue isn't "when does agile work?", rather "how agile can i be?" This implies a continuum. Going to iterations is a break. Scrum is evolutionary in perspective but revolutionary in implementation (making it hard). Kanban is revolutionary in perspective but evolutionary in implementation. Many (most?) people in Agile community think Agile==Scrum. So when Scrum won't work they think Agile won't. At conferences people say "we're doing agile" when they mean they are learning scrum.
Andreas Boes at XP Days Germany 09 on “The New Industrialization” (via Alistair Cockburn) – “Industrialization" converted subjective process to objective process (remove individual's capabilities). Taylor reduced 'processes' to invidual activities and actions. Taylor was a micro-methodologist (studied microtechniques). Alistair Cockburn - I think modern agile ppl are doing Taylorist work: study top performers, decompose, teach to others. That's exactly what we're doing! (I disagree with [Boes] - he says programmers are black box. but nano-TDD opens that box. I think agilists are whitewashing a secret Taylorist agenda, and it worries me.) what does he mean with “separate subjectivity from individuality”? I need a translation of Taylor’s 3 ground rules. Boes - agile brings in "living process"; which also lean brings in, too. diff : Taylor = optimize before announcing the process; Lean = optimize after deploying the process. Agile brings self-organization; Lean brings repeatable processes. Agile = collective learning & knowledge capture; this makes the "objectivity" and visibility needed. What does agile mean for the average worker? we haven't thought through it (bad things to follow?) Alistair Cockburn - Boes/audience Taylor tried to diminish the human, sw tries to augment (I worry that parts-replaceable programmers goes the wrong way.
Bob MacNeal (more on being self-directing) - 1) Responsibility - given or takes responsibility for results and 2) Autonomy (free to determine, plan & schedule activities).
Chris Sterling - scrum is a learning system; modify tools, practices, & process when your team learns something.
Dave Rooney - Agile is a collection of those 'things that work' put together in a synergistic way... whole is greater than sum of parts.
David Anderson - Kaizen done right is statistically based systems analysis often using SPC. Diana Larsen - Retros done right get at diff probs than SPC - esp human issues & + deviance. Well-run retrospectives incl subjective & objective data, as relevant to retro focus, + analysis & action. David Anderson - retros tend to project level focused. higher maturity orgs will have an organization level focus across teams/projects. This is not guidance or opinion. This is field reported cases. You need to be open to ask why? and challenge your beliefs. Paul Dyson (via Rachel Davies) - Like all agile practices (or practices in general), there is a risk of 'cargo cult execution' with retrospectives.
Gerald Weinberg - Definition of Bureaucracy: Each thing is in control, but everything is out of control.
James Bach - Saying ISTQB certification makes a tester better is like saying a chefs hat makes you a good cook. I had a woman in my class who studied for two years to be a baker. Then she studied for a few days to get a tester certification. Why are donuts worth two years of school, and testing worth no training at all? Because testing is easier to fake than donuts.
Jason Yip - Crisis is required to provoke deep change only if top management has a monopoly on setting strategy.
Jean Tabaka asked about 100-130 chars to describe Kanban and some responses were: Karl Scotland - Map value stream, visualize, limit WIP, establish cadence. Reduce WIP to improve value flow & individual fulfillment. David Anderson - visualize flow, limit WIP to encourage evolutionary change towards lean outcome, high maturity culture.
Jurgen Appelo - Introverts are not shy. Introverts just prefer low-noise communication. [Added by Dave Rooney in a retweet - Introvert == shy is common misconception]
Karl Scotland - If your process is designed to expose dysfunction, what do you do when your process becomes the dysfunction?
Karl Scotland (asked) - Are retrospectives a form of Schewart/Deming Cycle? David Anderson (replied) - I find most retrospective guidance to suggest subjective, anecdotal feedback rather than objective data-based Deming style info. SEI classifies OID based on subjective, anecdotal evidence as low maturity even though OID is a ML5 process area. Deming's method would be considered high maturity OID/CAR ML5 by the SEI. Typical retros are a low maturity precursor to PDSA. meanwhile, @kjscotland and I are only reporting facts from field. High maturity kanban teams tend to drop retros as waste.
Vasco Duarte - Ppl talk a lot about business value, but they forget that for most people business value is totally subjective! (i.e. unquantifiable)
Vince Lombardi (via Jason Yip) - Perfection is not attainable, but if we chase perfection we can catch excellence.