This week, I resubmitted our manuscript to PLoS Biology in reply to reviewer comments. We'll see how it goes, but I'm ecstatic to no longer have that work and deadline hanging over my head. Publication is onerous.
I have learned a few lessons -
limit the core team to as small a group as is feasible who can all work together well. Overly high maintenance and clashing personalities can be difficult to appease, especially when in principle, consensus is needed to move forward.
As a corollary to (1), avoid work situations where you and your boss have incompatible perspectives.
It's helpful - even as first author, doing the core work - to have a senior advocate who can assist with navigating the publication process
Structure your project in a way to maximize reproducibility; organize everything so that it could run on a new system with the code and data downloaded. Refactoring code is unpleasant.
Thoroughly document all your experiments, but keep this documentation separate from the final version of the analyses you implemented.
An interesting way to summarize each day or longer period of my life (since 2002, say) would be the collection of Wikipedia pages I visited that day or period. Unfortunately, my records are not comprehensive, but perhaps the NSA could help me out.