[vbernat]

On the other hand, publishing publicly will force you to fix every little detail.

For example, the most popular article on my blog is about the TCP "time wait" state on Linux. I knew for years how it worked and wanted for people to stop blindly use some sysctls. Therefore, I thought, what's best than a blog post to spread some awareness on this? But, for every sentence, I had to be sure that this was true. So, I checked, I tested. It took me a lot of time. I would never have done that if this was just for my drawer. My drawer never sets random sysctls and my drawer doesn't know better than me, nor it would shame me for telling incorrect stuff.

It's like publishing as an open source project: you have to be more rigorous in what you publish. There is some balance to find.

[jasonthevillain]

I agree.

The open source analogy is actually pretty good. Some projects you know you how to go about right away. But when you're not sure, start anyway; rough drafts are like prototypes. Dropping a few (not all) can be part of the process.