Why software architects fail: and what to do about it - Stefan Tilkov | Craft 2019

Поділитися
Вставка
  • Опубліковано 9 лис 2024

КОМЕНТАРІ • 19

  • @SanthoshDivakar
    @SanthoshDivakar 4 роки тому +32

    Quick read through of the topics the speaker addressed. Apologies if I have missed something important , so please watch the video anyway
    6:27 (A) Over-Generalization Drive - Turning to generic solution for all the problems
    6:51 Phases in a Developers Life
    1. Enthusiastic Developer
    2. Disillusioned Developer
    3. Enthusiastic Architect
    - KISS, YAGNI, Lean, MVP, Story Focus
    4. Disillusioned Architect
    5. Wise Architect ( Answer to all questions: It depends )
    10:52 (B) Domain Allergy - Treating Domain as a nuisance
    - 90% The generic machine solution + 10% domain
    - The 10% controlled by some configuration
    - Customer wants simplicity not a confusing configurations
    - Benefit of choices already made
    14:11 (C) Obsessive Specialization Disorder - Opposite of A. Every problem is unique
    - Print list of most frequent words example
    - Believing everything needs to be a perfect match
    17:18 (D) Unhealthy Competition Attraction - We are so smart , that we wont do simple things
    - benefits of complexity > challenging work, interesting and new experience, self-esteem, community, barrier of entry, job security
    19:20 (E) Analysis Paralysis - Taking longer to evaluate than to do it.
    - Vendor Selection example
    21:39 (F) Innovation Addiction
    - Conference Fashion
    - Choose boring technology , that helps.
    - Stop looking at google, Facebook, amazon , the biggies
    - Choosing two coins of change at every new project ( ex: only UI and DB will change )
    26:12 (G) Severe Tunnelling Fixation. - Wanting to use the tech which I know of. Enforcing architectural approach that clashes with framework , libraries or tools
    - Desktop web apps example
    - Using fashionable technology because it is popular
    29:32 (H) Asset Addiction - Becoming attached to a particular tool/framework, that every problem seems to be solved by that.
    31:02 (I) Exaggerated Risk Aversion - Sticking with Horrible tools.
    - This is how we do things around here ( OMG ! )
    - After a while the new person gets used to it.
    - Confusing “easy” with “simple”. The hard thing is to do the simple thing
    - Rick Hickey “Simple made Easy” talk
    35:05 (J) Impact Dissonance - Becoming too detached from the actual system.
    - Believing everything has to be approved by you
    - What Architects wants to do
    1. Shape Strategy
    2. Explore Technologies
    3. Make Important Decisions
    4. Mentor Developers
    - What others think Architects do
    1. Slow Down Development
    2. Define Annoying Rules
    3. Never listen to developers
    4. Pick the wrong tools
    - What architects should actually do
    1. Act as salesperson (30%)
    2. “Try” to be involved (35%)
    3. Defend Architecture (30%)
    4. Do technical Stuff (5%)
    41:45 An Architects Success Formula
    1. Dogma and Rules (10%)
    2. Experience (20%)
    3. Pragmatism (20%)
    4. Flexibility (10%)
    5. Minimalism (10%)
    6. Trends and future needs (10%)
    7. Experiments and PoC(10%)
    8. Hands-on participation (10%) - coding, but not on critical path.
    9. Vendor Advice (0%)
    46:05 Q&A

  • @Zhuinden
    @Zhuinden 4 роки тому +28

    8:14 "I can actually solve THIS problem ONCE with a GENERIC solution! My life is gonna be so much better! That's when you turn into the Enthusiastic Architect. This is a really critical phase of your life as a developer, as this is when you can do the Worst things to other people, and systems around you, and really ruin people's lives and make them miserable"
    My 6-year-ago-past-self feels personally attacked (and it should be)

    • @mr-boo
      @mr-boo 4 роки тому

      5 years ago for me. Happy to have suffered myself the consequences of those decisions to know how stupid, stupid, stupid that can be :)

  • @nathanhughes8354
    @nathanhughes8354 4 роки тому +37

    Actual content doesn’t start til 4:30 in.

    • @wety789
      @wety789 2 роки тому

      Thank you!

  • @kalmarnagyandras
    @kalmarnagyandras Рік тому +1

    Video actually starts at 4:28

  • @PetrGladkikh
    @PetrGladkikh 4 роки тому +3

    Can confirm going through all 5 stages. Only to find that no one actually needs my craft. And saying "I had warned you about this at least 1 year ago but you did not listen" does not bring as much Schadenfreude as earlier, so stage 6 is "fuck that, sky wont fail anyway" :)

  • @ruslikaici
    @ruslikaici 3 роки тому

    probably a good example for "you have to work more for the end result to be simpler" would be a new physics theory that replaces the old one that had many special cases: for example the equation for planets movements compared to way they were computed in the antic times

  • @audriuskaunas5999
    @audriuskaunas5999 4 роки тому +1

    Thanks, Stefan for the truth.

  • @teresaanderson7384
    @teresaanderson7384 2 роки тому

    nice

  • @Mik1604
    @Mik1604 Рік тому +1

    This is mostly nonsense. At some point in your architecture career you learn that is has very little to do with software and everything with people.

  • @edgeeffect
    @edgeeffect Рік тому

    Good grief that intro was one of the most annoying thing I've ever seen.

  • @greg1486
    @greg1486 2 роки тому

    Long story short: Just make the big ball of mud they told you not to make, because it probably won't actually become that.

  • @osten_petersson
    @osten_petersson 3 роки тому

    only ppl that would play a ten walls song nowadays are IT people - completely clueless...

  • @vvp8590
    @vvp8590 4 роки тому +3

    Cringy into.