The sempai - kohai relationship

Ok, I do admit that I intended to do my next blog about the nature of problem domains, but I read some interesting stuff on Alex' blog about the importance of feedback (http://alegu.blogspot.com/2006/02/getting-better.html)

Interesting enough, the japanese have a culture of so-called sempai-kohai relationships used to educate the younger generation. Kohai and sempai could be translated as a master-and-apprentice relationship, where the sempai (master) actually acts like an in-between between a fatherfigure and an teacher.

Now, what makes this special is the fact that the sempai will be expected to educate not only in an intellectual sense, but to build and install personallity within the kohai.

Coming back to Alex Blog, the method of doing so is by continously feeding back and reflecting on things past. In fact, I do believe that one of the reasons (not the only one) this method is (or was) very successful is that the transcending element is permanent improvement through very short feedback cycles.

It will be interesting to see what we could achieve by feeding back !

On defined and empirical processes

Textbooks like Craig Larman's "Agile and iterative development" state that "A defined process has many predefined and ordered activities to be followed during development [...] Empirical processes [...] are based frequent measurement and dynamic response to variable events.

Now, since agile and iterative processes are definitely part of the latter, one must not assume that there is no such thing as a defined workflow that describes the development. In fact, structures are in place like in every other process; workflows are defined. The biggest difference is that you evaluate what you have achieved and try to learn from these results.

It is therefore, as opposed to defined processes, that empirical processes focus on feedback and metrics that have been collected during the past iterations. Forecasts are based not on estimates alone, but on empirical data that harden the estimates.

So far, so good. The interesting part however, is what makes a defined process a defined process as opposed to an empirical process.

Since neither defined nor empirical processes do advocate a certain way of doing things, the difference must be in the way the solution of the problem will be achieved.

For instance, all processes in tool manufacturing are defined processes: They work perfectly well, since they have been proven multiple times. Therefore, nobody would need empirical processes here, the project can be overseen from the very beginning to the end.

Interesting enough, the greatest strength of empirical processes, the integration of feedback and the possiblity to actively manage the project by incorporating this feedback, are not needed.

In tool manufacturing, this is due to the fact that the feedback has been incorporated after the project's end, the new project has adopted these changes and so on.

It seems that even defined processes are actually empirical processes - they did incorporate the feedback, but not while the project was running but in the aftermath! So, interesting enough, the domain in which we apply the process is actually one of the crucial factors of the success of the process.

I guess this will be worth a different blog entry - on the nature of problem domains? Let's see where this goes ....

The Lard Game

Here's something funny which I came across while chatting to an english colleague of mine. It's actually some sort of a pub game, best played in an pub ....

Okay, the rule is quite simple: Just think of a songtitle that has the word 'Love' in it and replace it with 'Lard'

Example:

Jennifer Lopez: My Lard don't cost a thing
The Beatles: All you need is lard

And of course, credit where credits are due: Some contributions from other players:

TSU: The lard song
OTE: Lard is a battlefield.

If anybody thinks this is funny - I will make it a point to post the most hillarious title in future blogs.

Have fun !

Finally ...

it's happening to me ... Seems that it is no longer inevitable to share some thoughts on the web. At this point I have no idea where this will be going, but then again, it would be boring, wouldn't it ?