Good Will Blogging
How do you like them apples...

Phase 1. Use a GTD system. It works well. Things are getting done, nothing is being forgotten, and you’re feeling less stress. Life is good in Phase 1.

Phase 2. You grow confidence in your system. So much confidence that you throw new tasks into it with giddy abandon. Adding tasks to the system starts to give you the same feeling of accomplishment as actually completing the task. (This is obviously wrong, and the first warning sign.)

Phase 3. Tasks are going into the system much faster than they’re coming out. It starts to grow out of control. The number of tasks becomes overwhelming, and you resist looking at them.

Phase 4. You try to regain control by “organising” the system. You make categories and special lists. You associate actions to projects. You invent new disciplines for yourself to “keep things organised”.

Phase 5. The work required to keep the system up-to-date starts to exceed the time spent on the actual tasks it tracks. Now you feel it’s not working. Something has to change.

Phase 6. You notice an article or an ad for a “simple”, “clean”, and/or “powerful” to-do manager. You think, “That’s what I need! This system I have now is too complicated, too hard to use.” And you switch. Maybe you figure out a way to export all your projects and actions from the old system and import them to the new. Or maybe you decide to “start fresh.”

Repeat from Phase 1.

I’ve personally been through about 7 or 8 different GTD systems. That’s 7 or 8 times through the cycle above. I know it well. And I just completed Phase 6 again this week.

As a future reference for myself, more than anything else, here are some thoughts on what makes GTD tick — what makes a “good” GTD system.

First, and most important, GTD is a process of thinking, not a system or a tool.

A couple of years ago, I heard an interview with one of the “coaches” from the David Allen company, talking about her experiences teaching GTD. She said that often people will ask what’s the best tool to keep GTD lists, or complain that they really like their Filofax, or their Outlook, or their Post-it notes. She would say, “That’s fine! I can teach you how to do GTD on Post-it notes.” Her point being that GTD is about thinking what is the next action, and relieving your brain of the burden of having to remember so much stuff.

So it really doesn’t matter what your tools are. In fact, I’ve found that the tools can actually get in the way, because they distract me from the really important job of thinking.

Second, don’t invent parts of GTD that aren’t really there.

A classic and very common example is association between projects and actions. It’s so common for GTD tools to offer a way to link projects to actions, that many people I know think that it’s part of GTD. It is not. This is like that Real Monopoly meme. Go back and read the GTD book again — it’s not in there. There is even specific advice not to try to link projects and actions. You just need a Next Actions list, a Projects list, and that’s it.

The rationale for this is that if the project is current, the action real, and you are doing your Weekly Review, you will know which actions relate to which project.

Having learned this lesson once already, I again fell into this trap while using Flow. I am now convinced that trying to link projects and actions is death to a GTD system. Having actions tucked away in project lists just keeps them out of sight, out of mind. And it hides the true size of your system. Keeping your lists short should be part of the motivation to get things done, and to not over-commit yourself.

Third, use the Someday/Maybe list aggressively.

When you find yourself with a bit of time, scanning your actions list to find something you can get done, you should be able to complete any action on the list. If you look at an action, there can only be three possible reasons why you can’t do it right now:

1. It needs a context, either a place or a person, to get done. If this is the case, you should really annotate the action with that context right away, so you don’t go through this again next time.
2. It’s not really a next action, i.e. you haven’t done enough thinking to boil down the next “physical, visible thing needed to move the situation forward.” If this is the case, do that thinking now, and replace this action with the real next action. Then do it.
3. You don’t want to do it. If you don’t feel motivated to complete the action, and it’s not because of the two reasons above, then either delete it and forget about it, or move it to Someday/Maybe. (This also applies to projects.)

The goal here is to keep your Projects list and your Next Actions list as short as possible. They should be a list of things that you are really motivated to complete. Looking at these lists should get you excited, not make you groan.

Moving something to Someday/Maybe is not throwing it away, so long as you commit yourself to reviewing the Someday/Maybe list regularly, about twice per month or so. Of course, when you do, delete things that you recognise will never happen — you don’t want that list to become overgrown and full of useless crap, either.

Fourth, no metadata, no notes, just things on lists.

Another bad habit encouraged by all these GTD tools (especially electronic) is adding all sorts of tags, due dates, priorities, notes, etc. The best GTD system I ever had was paper-based. I really encourage people to try running a non-electronic GTD system for a few months. The nice thing about paper-based systems is they resist a lot of metadata cruft. You can write down due dates, tags, priorities, etc., but it’s a lot more effort, so you don’t so much.

This is how it should be, even when using an electronic tool. The occasional due date or priority highlight is fine, but over-reliance on them dilutes their meaning, to the point of becoming a waste of time.

Fifth, don’t get too hung up on contexts.

I’ve never found it helpful to keep separate lists for each context, mostly because there would just be too many of them, and again, having a lot of lists means actions are hidden and easily forgotten.

It’s best to just keep one big list of all actions (or two, if you use “the line”), and note your contexts at the beginning of the action, e.g. “@Dad — ask about ideas for Mum’s birthday present”. It works well for both paper and electronic systems, because you either scan down the list visually, or just search for “Dad”.

And anyway, what are those contexts? For me at least, there’s @Home and @Office, and maybe @Shops, and the rest are people. And since most actions in “people” contexts can be accomplished by phone, email, or text, they can actually be done anywhere, anytime. So don’t spend a lot of time adding contexts. Add them only when not being in a context prevents you from completing an action here and now.

Finally, GTD is about getting things done.

Any time spent fiddling with the system is time not spent completing actions. Behind every minute spent “organising” the system, is a reason why you’re not motivated to complete the actions in there. Find that reason. Figure it out. Move forward.

The API for developers to interact with Siri will be in the cloud.

Siri requests are processed in the cloud, so the shortest path from Siri to apps is within iCloud.

This will require apps to be integrated with iCloud. Siri API will be a part of the iCloud API.

How this could work:

• An iPhone user speaks to Siri.
• Siri in the cloud deciphers the user’s speech, and figures out that the user wants a service from app X.
• Siri invokes the iCloud API of app X. App X responds with some text, an image, and/or a URL that their iPhone app has registered.
• iCloud returns this data to the iPhone, and Siri speaks the text, displays the image, and/or jumps to the developer’s app on the device.

What I haven’t figured out yet is how developers will register their apps to be invoked for certain phrases. Seems incredibly broad. And how will competing apps register to receive the same invocations?

It’s going to be very interesting.

I’m trying a variant of the 30/30 Work Cycle for days when I work by myself, without the distractions of meetings and random interruptions. So far, it seems to work fairly well, especially when motivation is a bit low and the work takes a lot of concentration.

My variant goes like this:

1. Set a timer (on iPhone, in my case) for 25 minutes.
2. Work on something hard until the timer goes off.
3. Take a minute to jot down notes on whatever thought you were in the middle of.
4. Stand up out of your seat, walk around, stretch, jump, whatever. Just get the blood flowing and the heart rate up.
5. Repeat from 1, choosing a very different task to the previous one.

I have noticed a few nice benefits to working like this.

First, getting up and moving my body about twice per hour seems to help a lot with the lethargy that usually comes from working at a desk all day.

Second, the context shifts keep everything feeling fresh. Motivation is a lot easier when I know I only have to concentrate on this task for 25 minutes, especially for tasks I’ve been avoiding. And, it’s amazing what can be achieved in only 25 minutes.

Third, it stops the opposite problem, where I get too engrossed in a task that is easy and/or fun, but ultimately not very important.

Fourth, when I have something I want to do, and something I know I should do, I don’t have to choose. I can do both! (Pro tip: always do the “should” task in the first 25-minute session.)

I’ve found that with this method, I can work for many hours (my record so far is 12!), without falling into lethargy or fatigue, with 50 out of every 60 minutes spent in concentrated, productive work.

I like standards. Standards are good. Except when they’re stupid.

A while back I found myself about to argue that the Japanese date format (yyyy/MM/dd) was the international standard, followed by a long spiel about why it’s better. But I thought I should check my facts first, and a quick search in Wikipedia found this: http://en.wikipedia.org/wiki/ISO_8601 The tl;dr summary is, the ISO date format is yyyy-MM-dd.

Close enough though. My argument was going to be that the year-month-day format has 2 big benefits:

• The units are in size-order. In this case, it’s big-medium-large, but the opposite order, eg. Australian (and presumably UK) format is dd/MM/yyyy, is just as valid for this criterion.
• Having the biggest unit at the left makes it useful as a sort order. (In this case, ISO and Japanese formats are optimal).

After thinking about it for a while, I found that the ISO format’s use of hyphens instead of slashes as the separator also has some benefits over the Japanese use of slashes:

• Hyphens are less distracting visually, making it easier to quickly read the component numbers.
• Slashes are not very different to 1’s, whereas hyphens are not similar to any digit.
• Slashes can’t be used in file or directory names. Hyphens can.

So far the ISO standard format sounds pretty good, right?

That’s dates, so how about times? According to ISO 8601, the standard format is hh:mm:ss. This is fine, and pretty much expected. However it does have a couple of problems:

• The colon character is generally not allowed for file or directory names.
• The colon is a little bit too close to digits in visual density, making it a little bit harder for the reader to find the component numbers.
• No singular standard for sub-second precision.

OK, so not ideal. But not stupid. So what’s stupid?

It’s the part of the standard for combining a date and a time into a date-time stamp. The ISO format is yyyy-MM-ddThh:mm:ss, eg. 2011-07-26T13:46:15.

WTF?! T???

“T” is a terrible delimiter. It has the same visual density as the digits on either side, making it harder for the reader to see where the date ends and the time begins. The example above looks like it’s made up of two parts, “2011-07” and “26T13:46:15”.

Worse, I’ve never seen “T” used as a delimiter in this way outside of the ISO format. At least with date and time, “-” and “:” have been commonly used as delimiters. So the “T” is foreign to everybody.

You know me. Faced with this anomaly in the universe, I feel compelled to appoint myself as a committee of one to fix the fracking international standard for date/time formats. Here goes.

• Date: yyyy-MM-dd
• Time: hh.mm.ss.SSS
• Datetime: yyyy-MM-dd..hh.mm.ss.SSS

(The “SSS” can be extended to as many digits as needed. The sub-second and second components are optional.)

The benefits of this format are:

• The units are in big-to-small order.
• It’s easy to locate the component parts visually, and date and time are more separated than their internal component parts.
• It can be used for file or directory names.
• As a bonus: it can be typed entirely on the ten-key keypad with one hand on a standard PC keyboard, with no chording.

The disadvantages of this format are:

• It’s one character longer than the ISO format, but I think this is an insignificant difference, and worth it for the visual clarity. A space character can be used in place of “..” in situations where a space character is permissible.
• Using dots in the time format is less common than the colon character, but I think it’s ultimately better, and not difficult to understand.

What do you think?