#+Title: On Scrumm

It is of good taste these day to critique Scrumm.
Here are my 2cents

* Personal Experience with Scrum:
** Discovering
- Fear and counter arguments
** Efficiency
- Time lost in meeting
** People tensions
- instead of tamming it made everything worse.
* Management: The root of all evil?
The root of the problem is between the developpers and the managers.
- Manager: I want a great product, I want to finish it fast, I want my customers to love it
- Developer: I want a great product, I want to finish it fast, I want customers to love it

What could go wrong?

- What Manager means: _I want a great product_:
  I want to sell it!
- What Developer means: _I want a great product_:
  I want it to use the last technology, with the last code organization/quality trends

- Manager: _I want to finish it fast_:
  I don't want to listen to technical discussion,
  this looks like something easy to do.
  It should be in my hand in few weeks.
- Developer: _I want to finish it fast_:
  I want to keep the code clean to be able to add new changes fast
  with confidence (without breaking anything)
  this certainly means, testing + testing environment + proofs ...
 
- Manager: _I want customer to love it_:
  They should buy more and more. The product should be useful.
- Developer:  _I want customer to love it_:
  User should enjoy using it. The product should be simple, clean, natural, beautiful.

The two meaning are'nt completely opposite.
Still they are quite different.

* What could we do about it?

How to solve the problem?
- Spoiler: _you can't_.

The root of all evil is "it looks easy, do it".
Proof, a guy made the same thing in PHP in 2003, or I saw the same shit in Flash around 1998.

You have two choices:

1. Use tools to finish your work fast, but the cost is very to maintain and modify.
2. Use tools that enforce quality, you'll have a starting cost.

Error not to do:
- Manager ask for something, you use your l337 H4X0R cape and you show him
  something in less than 10 minutes.
- Manager believe that everything should be as fast
- Discover what you did was just a terrible hack, and take 2 days to finish it
  correctly.
- Manager doesn't really understand why he _saw_ it right now, and has to wait
  many days to really have it in production.