I Combined some images I had from past presentations with images from http://gtdrawings.blogspot.com/2005_12_01_archive.html

The art of stress free productivity is the result of using a system that frees our mind to focus on the moment we are in (mind like water). I was introduced to the idea through David Allen, Getting Things Done (GTD),
which is typically thought of as a task management system, many years ago.

Basically, if you consistently do the following:

• Use ubiquitous capture (a pen and paper always in reach, or a smart phone)
• Develop a system for managing all the stuff for all the contexts of your life
• Do a weekly review

You will take advantage of the miracle of ready state, distributive cognition.

For many people exploring GTD, there is a great sense of relief in the first few weeks. Keep doing capture, process, and review for a year or more, then bask in the transcendental experience: mind like water.

This is because most people have a system they can trust for the majority, but not all of their lives. Do this for a month and feel at ease. Do a hundred reviews and that system will cover all the contexts of your life.

For more information refer to his talk at Google http://www.youtube.com/watch?v=Qo7vUdKTlhk or mp3

Either book by David Allen, Getting Things Done or Making it all work

To return to part 1

One of the downsides of Sheila… is her ability to read, though this can easily be worked around. Using character regognition (OCR) it is possible to navigate through the user interface (UI) and read the text. Using tools like Tesseract or the Python Imagin Library   (PIL) it is possible to read the screen. This allows for more human tests, such as click, File, Save As.., which if perfect for integration, which naturally evolves into future regression testing. Without this type of testing it is possible to release software with functionality that is not avaialbe in the UI.

While Bruce is still better at patern recognition, it is possible to teach Sheila to search for paterns in screen using image comparison tools in PIL or any Tiff-Diff tool, and sheila is wonderful for looking at every single pixel out of millions and detecting the slightest shift in color. In Kodak Proofing Software™ we do exactly that comparing millions if not billions of pixels of hot folder output for jobs we believe are cool or tricky.

As mentioned earlier one of the great advantages of automated testing is the ability to clone systems. Using virtual machines any number of instances can be duplicated, which is fabulous for trying hundreds of variations of very similar conditions. There are even companies that provide virtual farms, which allow tests to run in parallel using a great number of computers. Tests that once required an entire night to run could be run on a virtual farm of 100 computers in a matter of minutes. This makes it practical for developer to get immediate feedback after a code change. This makes refactoring code easy in comparison to nightly or slower test cycles.

It might require a group of monkeys an infinite amount of time to create the entire works of Shakespeare banging on Smith Corona. It might even take a group of virtual machines longer to write the Great Gatsby using random combinations that the duration of the universe or in particular the human species will practically allow; however, virtual machines can easily test 1024 commendations of the user interface without ever getting bored. The same group of monkeys or worse humans would likely cause total chaos before completing such a tedious task.

To set up automated testing a group needs to commit to a framework. A perfect choice is Python. It is open source, multiplatform, widely supported, and has a great number of libraries useful in testing. It also is highly readable, and encourages unit testing. Python can be used to drive the user interface, so the robots don’t actually need to type. Perl and AutoIt could also be used as a framework for automated testing, though AutoIt is Windows only. It is also well possible to have the developers create an API for the code being developed, though this is not required, especially when using the user interface to test.

One of the biggest challenges in automated testing is interacting with hardware. If you require testing of the system that requires hardware feedback in real time this may be challenging. For this it is likely a workaround will be required. For instance, in Kodak Proofing Software it is not possible to have a spectrophotometer read a chart that provides the same measurement data each time, since print color shifts over time. Virtual machines are also useless at moving paper. Therefore we decided to use canned data and have Python provide this using XML during the test. We still use humans to measure actual color in real time in an additonal to the automated test. So there is at least one reason to keep the humans around.

more…

I was introduced to the idea through David Allen, Getting Things Done (GTD),
which is typically thought of as a task management system.

Basically, if you consistently do the following:

• Use ubiquitous capture (a pen and paper always in reach, or a smart phone)
• Develop a system for managing all the stuff for all the contexts of your life
• Do a weekly review

You will take advantage of the miracle of ready state, distributive cognition.

For many people exploring GTD, there is a great sense of relief in the first few weeks. Keep doing capture, process, and review for a year or more, then bask in the transcendental experience: mind like water.

This is because most people have a system they can trust for the majority, but not all of their lives. Do this for a month and feel at ease. Do a hundred reviews and that system will cover all the contexts of your life.

For more information refer to his talk at Google http://www.youtube.com/watch?v=Qo7vUdKTlhk or mp3

Either book by David Allen, Getting Things Done or Making it all work

Studies show that people tasked to memorize a less significant series of numbers will make a healthier decision, and those tasked with memorizing a greater string of digits go for donuts: explaining today’s growing office waistline.