Howto organize your scientific projects?
My daily academic job requires me to organize several projects, some with the goal to write and publish a paper, other to provide a small contribution to someone else's project, working on a small code snippet and so on. Sometimes project are connect, or completely disjoint. For some projects, I need to keep track of sub-projects and sub-goals, i.e. some project are large and running long enough (longer than several months) that I need to introduce milestones and divide my studies into smaller pieces. It can even happen that I do not know beforehand where my project ideas will lead me whether to a before set goal or something else; it may fail. The latter case may seem still interesting, e.g. for producing a negative result paper or just a note on the arXiv. However, you will seldom find negative results very welcome in the community, although they still exists and are important, they rarely generate any fuss.
A supportive environment helps us to maintain our concentration on a single task without our mind wandering away with other task nagging at us. I spend most of my time using my terminal and/or Emacs, the latter offers the obvious choice of the Org-mode. Org-mode presents itself with the following description:
Org mode is for keeping notes, maintaining to-do lists, planning projects, authoring documents, computational notebooks, literate programming and more — in a fast and effective plain text system.
Org-mode provides everything I could imagine for a organization setup. And, I organize myself within Emacs without the need of concentration-loosing search for a window on my desktop. But, emptor caveat, the Org-mode developers and the community are writing splendid manuals, tutorial, guides and so on, however, as far as I have read those, I never found a comfortable setting for myself. Therefore, I try to find myself a suitable way to work with Org-mode's agenda.
Let me start with a question: Why should I provide an extended note on my personal workflow? Because in the first place, the text exists already in a kind of limbo, I write my ideas down to get an overview and to connect different ideas and concepts to each other, thus, I can publish the refined notes. Second, although a plethora of good organizational tools exist, I always struggled to use those tools. Although those tools are working for someone, they do not work for me, because they do not reflect my workflows and my desired workflow, which I've been optimizing since I started university.1
We always forgot how much a good design helps us. In general, we do not recognize good design, we just take it for granted and use it happily. However, we directly notice the absence of a good design. Before I started to dive into the details of a new Org-mode configuration, I wanted to really think about the design of my workflows and lay them out a priori. And, because I am a lazy writer, I wanted to record my ideas in a coherent format. However, the following notes represent the outline I found after a lather, rinse and repeat marathon, so don't be ashamed if you do not come up with your setup in an hour or a day.
Before I start let me outline my current setup: I use the formidable Spacemacs. It already comes with Org-mode (and some other great org-related packages). Org-mode is very versatile and has a lot of configuration places and those come mostly with reasonable settings. Therefore, I try to stay as close as possible to the vanilla settings of org-mode for my setup with regard to Spacemacs. For a better understanding of the inner workings of Org-mode, I recommend to skim over the manual - mentally marking points of interest - and to know where to look for the explicit configuration definitions, either directly in the Org-mode source code or in the documentation.
For the first, I will update the note in a recurring manner; keep an eye on it…
Workflows change over time, for a good or bad reason, comme ci comme ça. But I need a tool, which I can refine at every step with knowledge and experience of my current status. So, most tools try to impress their workflow (at a given point) on us, but I do not expect myself to keep faithful to that one workflow …