Agile to Waterfall 3. My Brain Hurts!

Episode 76 - 15 Mar 2017

Here's the story so far:

  • Act one: An iterative process evolved into a waterfall process. Efficiency improved.
  • Act two: the waterfall process worked well. Except when it didn't: all too often, there were last minute panics.

Is there a third way? - More efficient than the iterative process? - More predictable than the waterfall process?

And, ideally, more _ _ _ ?

Let's find out.


FILM CREDITS Directed by Milos Forman
Writing Credits Peter Shaffer Copyright 2002 - Warner Brothers

This is one of my favourite movie moments.

This is F. Murray Abraham - you may recognise him from Homeland -on his way to an Oscar-winning performance as Salieri.

Salieri was a composer, who had the great misfortune of being a contemporary of young upstart Wolfgang Amadeus Mozart.

Salieri is about to get his first glimpse of Mozart's manuscripts.

"It was actually this beyond belief.

This first and only copy showed no correction of any kind.

He'd simply written down music already completed in his head.

Page after page. As if taking dictation".

I know what you're thinking.

Mozart's process is PURE WATERFALL!

Quick recap

Last time we were looking at my waterfall process.

How it looked on a good week. And how it looked on a bad week.

The process I use now - the process that I used to make this episode - differs from this in TWO ways.

The Script

I didn't have to have too many "car crash" weeks to realise that I had to be super-careful not to put things in the script that would be too difficult - or time consuming to make visuals-wise

This is easier to say than to do.

It meant that while writing I had to "compose" the whole thing in my head.

What was effortless for Mozart, was exhausting for me.

Too much "cognitive load".

It was getting me down. It was kind of making me kind of sad

It wasn't working, so I changed it.

Optimisation No 1

Instead trying to whole thing of writing the whole thing, I might write a bit, then mess around with the visuals for a while.

ESPECIALLY if it was something complicated. Like the boot sequence.

There are at least three benefits of working thing way:

  1. It moves some of the work from over here... to over to here.
  2. It helps to make the audio and the visuals more more integrated.
  3. It has Reduced my Cognitive Load.

This has been big for me.

Instead of being tortured, I'm much more chilled.

Let's move on to optimisation number two.

Optimisation No 2

I'd love to report that my new process over here was an instant cure for this pile over here.

But I was still having to get up at 4am on Tuesday and Wednesday mornings.

I need my beauty sleep. I'd rather be in my bed!

Until the day I looked at the pile.... in a different way.

What if I could bring order to this mess? Maybe I could move the pieces around a bit?

I realised that I could look at this process - not as a linear series of tasks - but a series of a layers.

The layers

Here's how the "layers" approach works for a video episode.

This clip - the one your watching now - is just text. It's quick and easy to produce.

So when the script is done, the audio is complete, the first thing I do is to create a set of visuals that is, err... not very visual.

It's 90% text.

My aim is to get to "complete" (albeit rather dull)

And then I go on to the next step:I edit the Audio and the Visuals - very roughly - into a video.

And once I've done that. I can breathe a sigh of relief.

At this point (and at any point after this) I could - if necessary - create the finished video in less than an hour.

As Suren said, it's my Minimum Viable Product.

You've never seen one of these MVPs. And I hope you never will. Because I'd like my videos to be a little bit better than that.

In order to make a half-decent video, I need to layer on the, err layers.

The more time I have, the more layers I'll add, and the better the video will be.

I hope you can see what a big deal this is:

  • Before, the work was a constant, and I had find the time.
  • Now, the work adjusts to fit the available time.

And that to me has made all the difference.

Current Process

My current process looks something like this (see video).

Not nearly as iterative as my original process, but a good bit more iterative than my Waterfall process.

I've traded efficiency for effectiveness. For predictability. For fun.