From e88606ec1901b94634747537c829333ba7002f5e Mon Sep 17 00:00:00 2001 From: alex Date: Sat, 21 Feb 2026 13:32:08 +0100 Subject: Move cliff's notes to blog website --- blog/content/notes/cliffs/peopleware.gmi | 272 +++++++++++++++++++++++++++++++ 1 file changed, 272 insertions(+) create mode 100644 blog/content/notes/cliffs/peopleware.gmi (limited to 'blog/content/notes/cliffs/peopleware.gmi') diff --git a/blog/content/notes/cliffs/peopleware.gmi b/blog/content/notes/cliffs/peopleware.gmi new file mode 100644 index 00000000..d7dee8de --- /dev/null +++ b/blog/content/notes/cliffs/peopleware.gmi @@ -0,0 +1,272 @@ +# Peopleware + +## I. Managing the human resource + +People are different from software. + +### 1. Somewhere today, a project is failing + +* 15% of all projects deliver nothing. +* 25% for projects >25 work/years +* Not for technical reasons, "politics" => sociology + +### 2. Make a cheeseburger, sell a cheeseburger + +* Errors should be encouraged +* A project objective is to be ended. Therefore, a project is never steady. Therefore, a project is always changing and there is no steady state +* Need to think more about "why" this task needs to be done rather than how the task must be done + +### 3. Vienna waits for you + +* Spanish Theory Management: increase productivity by extracting more work for the $ +* Mechanizing development, lowering quality, standardizing procedure reduces enjoyment of work + +### 4. Quality-if time permits + +* Self-esteem makes people emotional +* Self-esteem is tied to the quality of our work +* Deadlines conflict with quality +* Manager: Market wants time-to-market over quality +* Builders: want to match their past best achieved quality, more than what the market wants +* But quality is a means to productivity + +### 5. Parkinson's law revisited + +* "Work expands to fill the time allocated for it" +* Parkinson was a humorist +* Motivated people do not want to work forever in the same task +* The team can motivate people better than the manager +* Productivity by task estimator. No estimate > Systems analyst (unbiased expert) > Programmer > Programmer + supervisor > Supervisor +* Bureaucratic work does expand + +### 6. Laetrile + +People are desperate to increase productivity, fall to the seven sirens, seven false hopes of software management + +* Missed something obvious: no +* Others are succeeding, you are getting outdated, not using the right programming language, need more automation: technical gains affect just a small part of the total effort +* Need to get to the bottom of the backlog: bottom of the backlog is worthless +* Workers need more pressure + +## II. The office environment + +It's hard to increase productivity, but easy to decrease it + +### 7. The Furniture Police + +Optimizing for cost, and uniformity is not productive + +### 8. You never get anything done around here between 9 and 5 + +Top performers work in quieter, more private, with less interruption, bigger spaces + +### 9. Saving money on space + +Cost of workplace is a small past of cost of worker + +### Intermezzo. Productivity measurement and unidentified flying objects + +* Gilb's Law: Anything you need to quantify can be measured in some way that is superior to not measuring it at all +* Individual productivity should only be measured by the invidivual + +### 10. Brain time versus body time + +Interruptions are expensive + +### 11. The telephone + +Ensure people attend their email with reasonable frequency (3/day) to allow prioritizing non-interrupting communication + +### 12. Bring back the door + +People work better in quiet environments + +### 13. Taking umbrella steps + +* Developers should design the working environment +* Windows +* Provide outdoor space, public space + +## III. The right people + +Get the right people, make them happy, let them work + +### 14. The Hornblower factor + +* Difficult to improve people, choose well +* Appearances << capabilities +* Do not hire for uniformity in the company +* No dress codes + +### 15. Hiring a juggler + +* Interviews are about performing, not talking +* Portfolios +* Aptitude tests are not for hiring, they are for self-assessment +* Audition on topic related to work selected by the candidate + +### 16. Happy to be here + +* Turnover is expensive and leads to short term planning => needs quick promotions, leads to inexperienced people doing the building +* Company moves are the worst +* Good companies *retrain* + +### 17. The self-healing system + +* Humans can improvise, automated process cannot +* Big M Methodologies automate: a) No improvisation, so must grow to cover all cases b) Lots of documents +* Big M Methodologies take responsibilities away from people into the Methodology +* Big M Methodologies lead to malicious compliance- follow the Methodology even if it has bad outcomes +* Convergence of methods is good, easier to onboard, etc. +* Achieve convergence of methods by training, tooling and peer review, without forcing a Methodology +* Hawthorne Effect: people perform better when trying new approaches +* Do new things on every project to benefit from the Hawthorne Effect +* But have a 10-page max. standard + +## IV. Growing productive teams + +Teams working as one on a challenge are the objective. Help the team form + +### 18. The whole is greater than the sum of the parts + +* Jell: a jelled team is more than the sum of its members. Jelled teams enjoy the work +* Jelled teams have a common objective, low turnover, strong sense of identity, feel elite, join ownership of product, enjoyment + +### 19. The black team + +* The black team tested other teams code. +* The black team outlived the original members +* Identity: dressed in black, some evil mustaches, mystique + +### 20. Teamicide + +You can't make a team jell, but you can prevent it from jellying: + +* Defensive management: preventing people from making mistakes. If the team cannot do the job, they cannot do the job. +* Bureaucracy +* Physical separation +* Fragmentation of people's time: the team must be together most of the time +* Quality reduction of the product: quality jells a team +* Phony deadlines +* Clique control (preventing the team for working together in further projects) + +### 21. A spaghetti dinner + +* Small successes lead to bigger successes +* Perform small projects, demos, etc. + +### 22. Open kimono + +* Trust the team +* Get them out of the office +* Let skunkworks projects happen +* Let people choose their peers and project +* Natural authority by being competent + +### 23. Chemistry for team formation + +Some organizations have environments that favor team formation + +Managers do not seem busy nor manage a lot, they maintain the chemistry + +Chemistry building: + +* Cult of quality +* I told her I loved her when I married her. Provide closure to each task. Small tasks for frequent closure +* The Elite Team. Allow and grant uniqueness. +* On not breaking up the yankees. +* A network model of team behavior. Managers are not part of the team. Occasional leaders inside the team +* Selections from a Chinese menu. Do not have a uniform team + +## V. It's supposed to be fun to work here + +### 24. Chaos and order + +Constructive reintroduction of small amounts of disorder: + +* Pilot projects. All projects as pilots, but limit experimentation +* War games +* Brainstorming +* Provocative training experiences +* Training, trips, conferences, celebrations, and retreats. + +### 25. Free electrons + +Some people should be left to work at what they want + +### 26. Holgar Dansk + +A "sleeping giant" can oppose any bad change + +## VI. Son of Peopleware + +### 27. Teamicide revisited + +* Those damn posters. Motivational posters tell obvious things people already know. It is demeaning +* Overtime: An unanticipated side effect. If someone is exent of overtime, it is even more damaging + +### 28. Competition + +Internal competition inhibits jell, Prevents internal coaching. Can come from: + +* Annual salary or merit reviews +* Management by objectives +* Praise of certain workers for extraordinary accomplishment +* Awards, prizes, bonuses tied to performance +* Performance measurement in almost any form + +Musical ensembles are better metaphors of good development teams than sport teams. Individual sport teams members can have differing valoration from the rest of the team + +### 29. Process improvement programs + +* Standardized interfaces are good, standardized processes are not +* Goal is a good product, not building it efficiently +* Good products are risky projects, process improvement avoids risky projects +* Better teams do more complex projects, more risk + +### 30. Making change possible + +People fear change + +Degrees of fear to change: + +* Blindly loyal (ask no questions) +* Believers but questioners: skeptics (show me), passive observers (what's in it for me?), opposed (fear of change), opposed (fear of loss of power), militantly opposed (will undermine and destroy) + +Blindly loyal can abandon a change for a newer one. Only Believers but questioners can be allies to a change. Work with them to make change successful + + +elebrate the old system + +Phases of change + +* Introduce foreign element/catalyst +* Chaos +* Transforming idea (finding the "correct training"/correct way to adopt change) +* Practice & Integration +* New status quo + +People need to feel safe for change, there should be room for some failure + +### 31. Human capital + +* Money spent of people is only lost if they leave +* Replacing someone is expensive + +### 32. Organizational learning + +* Organizations can only learn if people stay for a long time +* Organizations learn when middle management works together without competition and without reporting to upper management + +### 33. The ultimate management sin is... + +Wasting people's time: + +* Being late for meetings, blocking meetings, inviting people who don't need to be there +* Status reporting meetings +* Early overstaffing (and leads to fragmenting time of people) + +### 34. The making of community + +* Aristotelian politics is building communities, extending ethics to a group +* Creatin -- cgit v1.2.3