Agile Zone is brought to you in partnership with:

I am a programmer and architect (the kind that writes code) with a focus on testing and open source; I maintain the PHPUnit_Selenium project. I believe programming is one of the hardest and most beautiful jobs in the world. Giorgio is a DZone MVB and is not an employee of DZone and has posted 638 posts at DZone. You can read more from them at their website. View Full User Profile

Lean Tools: the Last Responsible Moment

05.09.2012
| 5207 views |
  • submit to reddit

Options Thinking lead us to invest time and money in delaying decisions to a time where we know the most about it; the extreme application of the Decide as late as possible principle is the concept of Last Responsible Moment, the optimal point of the trade-off between the available time for a decision and the need to complete a story or a task.

The last responsible moment is the instant in which the cost of the delay of a decision surpasses the benefit of delay; or the moment when failing to take a decision eliminates an important alternative. For example, failing to provide a public HTTP API may make you lose an important customer, forcing you to publish an unfinished work.

Tactics

Mary Poppendiesk describes several tactics for delaying decisions until the last responsible moment:

  • share partial design information, before it is freezed or released. The irreversible decisions, like freezing an api, are made later after feedback has been gathered; at the same time, the rest of the team can start to work with it.
  • improve the response time for new stories. If you want to make a decision later, you still will have to respect the deadline. The faster you are, the later you can take important decisions. The adjectives lean and agile usually connotates lightweight approaches where decisions can be taken later for maximum flexibility.
  • absorb changes by delaying the commitments to particular implementations, tools, and libraries. Modularization, interfaces, configuration parameters and any kind of abstraction are welcome investments in any case where there is the possibility of change in the future. By the way, the *no extra features* XP mantra recognizes that simple design, which minimizes duplication and moving parts, is the best response to the need for evolution.

Real Options

The Real Option (still the financial option metaphor) concept motivates Agile practices as for their ability to improve our options for deciding at the last responsible moment. For example, tests give us more options for a design by preserving its ability to change; and pairing give us more options for who should develop a feature, as knowledge of that particular part of the code base is spread across the team instead of being concentrated in a few people.

It's all about risk management. Delaying decisions lets us able to make them in conditions of less uncertainty, when we can only know more about the domain and the project.

Criticism

Alistair Cockburn criticizes the concept of last responsible moment for several reasons.

First, since the characterization as a single instant is not so close to reality. Cost and benefits of a decisions are soft functions that vary continuously, so it's difficult to think of a precise moment where a decision must be taken. In most cases, the *moment* spans for days.

Second, this concept is not actionable, in the sense that you don't know the point in time where it will take place until after it has passed. Knowing that there is a deadline for a decision is different from knowing it with absolute precision.

Finally, Cockburn views it as simple not good advice as trade-offs between cost and benefits should only apply to critical decisions, like a database with an high cost or lock-in, or the hardware architecture of the application.

From the Extreme Programming point of view, it is correct to delay commitment to the last responsible moment, but not to overengineer a system to postpone every possible design decision. Choices like the programming language to write code in must be taken at the start of the project; the set of classes and methods should be kept minimal as long as duplication is eliminated.

After all, this is a series on tools and it's up to us to pick up the right tool in the right context. The last responsible moment makes sense for decisions which are costly to change, but everything that can be rolledback thanks to encapsulation and information hiding is already abstracted away enough. In fact, iterative development is based on starting with a large set of assumptions and removing them one by one according to priority, evolving the code towards a more general picture.

For example I have no problem hardcoding business rules, database drivers choices inside Repositories (but not credentials of course), and web application routes. As long as I can go back to the code in the future, they are not final decision; instead, I try to reserve the delaying of commitment to published interfaces and HTTP APIs...

Conclusions

We have learned to try to postpone decisions which are not immediately required, and even to invest in finding solutions for postponing some of them even when they should ordinarily be taken at the present time. The last responsible moment is a concept not to be taken literally, but when applied to difficult design and business decisions sets a goal for gathering all the needed information to take a choice when the time comes.

Don't worry about what you can still change: worry about what will be carved in stone and delay the related decision as long as it does not damage you.

Published at DZone with permission of Giorgio Sironi, author and DZone MVB.

(Note: Opinions expressed in this article and its replies are the opinions of their respective authors and not those of DZone, Inc.)