# Continuous Integration > Usually, we think about writing projects as a whole to be made up of these four parts. First, do some research, and in the end you finish writing the text. Most of the time, the process isn’t perfectly linear, but it seems to suffice to separate the tasks and continuously move forward. > Instead, work iteratively. Research for a while, then read and take notes, then write something in your draft. This is like [continuous integration](http://en.wikipedia.org/wiki/Continuous_integration) in software development: as long as you get feedback quickly, you won’t be surprised by parts not fitting together. **The opposite would be “Big Bang” integration: you develop all the parts in sequence or in parallel and hope for the best, that is that they fit together nicely. Most of the time, the parts won’t fit**. That’s why “Big Bang” integration has such a bad reputation. It’s the same with writing projects. If you work on the final draft continuously, you’ll know how well your past efforts fit into the bigger picture. > Because you work on the final text all the time, you'll know when otherwise rewarding activities won't help you move forward. **You'll have a clear vision of your project, constantly.** > **It’s useful to keep the knowledge cycle short.** The shorter it is, the better you can judge what’s useful in the following round. We pick material based on our actual knowledge, and from increasing our domain knowledge follows that we get better at picking new material. You minimize waste by making informed decisions – waste of time and energy, that is. This is similar to the *iterative* process of **evolution** or **knowledge creation**. We start with a **conjecture** and then **evaluate** it. If we let our conjecture period last too long before evaluation, we may waste far too much time than if we had simply seen that the conjecture was destined to fail early on. We want to test our conjectures as early as possible, for this will allow us to subsequently make *better* conjectures! > Most writing projects with sufficient complexity are completed iteratively anyway. We form thoughts when we write. These thoughts are important for planning and making better decisions. That’s why I call it a cycle of _knowledge_, not a cycle of getting writing projects done: ultimately, everything we do in a writing project is about increasing our knowledge during the process and teaching the results with our writing. Consequently, it’s rational you start writing early so you’re able to improve your ability to think about the topic and make more informed decisions. The shorter we can keep the cycle the more we can avoid falling into the [Collectors Fallacy](Collectors%20Fallacy.md). --- Date: 20220606 Links to: [Learning & Creativity MOC](Learning%20&%20Creativity%20MOC.md) [Knowledge Cycle](Knowledge%20Cycle.md) Tags: #review References: * [Use a Short Knowledge Cycle to Keep Your Cool • Zettelkasten Method](https://zettelkasten.de/posts/knowledge-cycle-efficiently-organize-writing-projects/)