Re: Event Modeling and TDD... somewhat
|
I think event modeling is poorly named. Yes these are the main events of the system, but no these won't be the only events. Some events will exist just for the systems sake.? It also doesn't capture state.
However, I find it a very useful tool, and works with TDD well.? Capturing and discussing the main events that the system must have, and their relationship from a business perspective, is useful in deciding where to start doing TDD. It also aligns business with developers and enhances the ability to do DDD.
The trick is to realize that event mapping mainly gives you a high level architecture with the critical paths that need to be covered to produce real business value. Its not intended to cover everything, but it is better than talking about vague goals of the system.?
I.e. we want a room added event, not just "make reservations". It helps you think about what you want to do with the reservation in the fist place. (The room added event can be sent to housekeeping and added to a reservation, and also sent to maintenance to check the lights are working, without compromising privacy)
On Fri, May 22, 2020, 04:20 Jeff Langr < jeff@...> wrote:
Hi JB,
It sounds like the New Message I heard c.2013 when Fred George said
"microservices are so small you don't need to write tests for them." I
asked Fred what he thought was small; "oh, maybe 100 lines of code." I
later read a book about the tao
of them that echoed this sentiment with numerous, similar putdowns
throughout.
Those with their fingers in typical codebases know full well the
impenetrable & defective mess that one can make in just 3-5 lines of
code, and the non-obvious nuanced behaviors it's possible to build into
not much larger bits of code.
I know many of us have been developing for some time. You reminded me
years ago that, as we move toward mastery, we think our ability to
discard "forms" (stand-ups, test-everything, etc.) means that everyone
should feel comfortable with doing so. We do need "beginner's
guidelines." And I still need TDD most of the time to continue going
fast enough.
I chalk this new New Message up to perhaps the same lofty levels of
experience that Mr Dymitruk may well have.
Then again, it's hard to get past messages like " TDD
offers little if you have your design in front of you" and feel gracious
at all.
Regards,
Jeff
|
| Jeff Langr / Langr Software
Solutions, Inc.?????
|
| ?+1-719-287-4335
|
|
|
|
J. B. Rainsberger wrote on 5/21/20 5:57 PM:
Hi, folks.
Adam Dymitruk
wrote this on LinkedIn.
I don't like that he framed this as "Don't do TDD; model events,
instead", but I otherwise trust him, so I'd like to know whether you
folks have any experience mixing event modeling with TDD.
I
don't, but I can see how I'd do it, and I imagine people who practise
TDD will eventually (eventually!) see the benefits of event modeling as a
shortcut. Adam and I have made vague plans to discuss this in some
detail, but if you folks have anything to say on the topic, I'd enjoy
reading it. -- J. B. (Joe)
Rainsberger :: ::
::
|
