~/talk

Agile Software Architecture

slides:none on file

Argues architecture is a usable map, not just a list of hard-to-change decisions, and agile is a risk‑reduction strategy with short feedback loops. Recommends walking skeleton and spikes to prove architectural risks, C4/context maps and sequence diagrams for diagrams, and bounded contexts and quality tradeoffs to partition work. Gives a concrete example: a Huddle file service using a task queue over EBS buffers to meet scaling and performance needs (20% time work; queue implementation proves robust in production).

AI-written summary · about

  1. 0:00Start
  2. 3:49Db Migration
  3. 8:55Three Key Principles of Architecture
  4. 10:04Diagrams
  5. 18:11Physical View
  6. 18:53Visualizing Software
  7. 22:54Quality Attribute
  8. 23:16Functional Requirement
  9. 24:52Performance
  10. 28:58Bounded Context from Domain-Driven Design
  11. 35:39The Agile Manifesto
  12. 46:38We Talked about the Key Architectural Risk of that Project It's Really about Scaling that but in a Performant Way We Have To Have Basically a Buffer Area Somewhere in a Ebs Volume Where Effectively We Can Put the Files We Had To Be Performant Enough that those Buffer Doesn't Become Overwhelmed and Q2 Much Best of Enough Av Processes Running Essentially at Once in Order To Basically Meet the Demand Everything about Scaling Up that's Our Key Risk Right some Stuff like the Http Endpoint Which You Already Have Is Not Accurate Don't Worry about That Do You Address the Key Risks in Your Architecture You Say What What Could Go Wrong What's Difficult Why Don't We Really Understand
  13. 53:01You Have Somebody To Oversee What All these Teams Are Doing Making Sure that They Are Adhering to the Plan and He's Kind Of Sit outside the Individual Team Structure When You Get Up to About 30 or More People You Tend To Embed an Architect in each Team and They Act as a Horizontal Team so They Essentially Bring the Architects of Concerns into the Team from across the Group You Um People Talking about Time about Self-Organizing Teams and They Say We Don't Need an Architect or a Self-Organizing Team Right We All the Happy Band Are Working Together We Share Architectural Responsibilities Uh I Rarely Ever Seen that Really Work I Think There Are Usually One of Three Circumstances
  14. 54:16Because There May Not Be Very Good at some of the Aspects of You Know Communication Working with Others Leadership Politics and the Third Situation Is Which Is Even Worse in a Way Is When the Team Has a Couple of Individuals Who Are Effectively at Loggerheads over Basically the Direction Technically that We Should Be Taking and the Team either Makes Breaks Compromises That Please Nobody or Doesn't Get Anywhere so Most Cases That Tend To Think It Tends To Be Easier To Appoint Somebody into the Role because It Clarifies the Situation for Everybody and Reduces the Amount of Politics Involved and Lets You Choose Somebody Who Doesn't Isn't Doesn't Just Take the Role because They Have Mad Coding Skills but because They Actually Can Deal with Human Beings for Example There's a Key Part of Architecture
  15. 56:51You Have Enabled Them To Put Aside that Piece of Complexity for Now and Keep Designed the Which the System They Can Work with Their Level Experience Delegate Designer Authority Rights When We Do a Quite a Lot You Can Be the Archetype for the Next Project I'Ll Mentor You and Help You Do It Nothing Is I Think Is Better for at Your Pet in Your Role and another Guy Do It Is He's Usually Moaning about It and You Give Him He Say to Me Ah Great Not Learning about It Do It Right and They'Ll after a While They'Ll Say I Don't Want Do It Anymore and Designed Together this the One I Always Tell My Team Is that You CanNot Impose an Architecture on a Tee Don't Come to the Document and Say Here Is the Architecture I'Ve Been Away to the Mountain
  16. 58:01Here Are the Key Cube Challenges Were Facing this Project Go Away and Sit There for Four Five Ten Minutes Scratch Out Your Idea on a Piece of Paper and some Post-It Notes It's Important this People Time on Their Own Have a Think then What You Do Is You Say Okay Everything Up and Present Their Idea and You Have To Remind Them There Are no Bad Ideas because I'Ve Seen some of the Things That People Think Is My Crazy Idea Everyone Goes Wow that's a Brilliant Solution To Probably Wouldn't of Thought of Right When the Person Presents Their Idea You Do Not Critique You at that Stage
~/search
type to search 6,249 conferences, 138,987 speakers and 185,172 talks
~/help
cmdctrl k
search every conference, speaker and series
/
focus the filter search
h
home
c
conferences
s
speakers
e
series
t
light / dark — the status bar button does the same
?
this help
esc
close this, then clear every filter — or just the search, while you are typing in it

Ignored while you are typing. Pages without a filter rail have no search to focus.