nb-sdouglas

lr-notebook/


lr-notebook philosophy & features

2016-03-03

Background

I've used the following as research notebooks:

  1. traditional pen & paper notebook (July 2005–Aug 2005)
    • pros: nice to sketch things on real paper.
    • cons: writing sufficient detail is slow & tedious, resulting in daisy-chained entries "did same as yesterday (see pg 23), except changed x". single irreplaceable copy is hard to share, impossible to search.
  2. group-wide mediawiki instance: 1 user account per lab member (Aug 2005)
    • pros: everyone's work in one place
    • cons: hard to enforce organization, lacks portability, scales poorly over time
  3. institute-wide wikifarm: unlimited mediawiki instances per lab member (Mar 2006–Jul 2012)
    • pros: more similar to individual paper notebooks, more portable.
    • cons: poor group-wide indexing, search. collaboration is still hard.
  4. per-project git repo of flat files + markdown templates, with per-user branches (Sept 2012–Jan 2015)
    • pros: benefits of git version control: diffs, lots of git(hub) integrations available.
    • cons: basically a step backwards in terms of default readability. similar cons to 2 & 3. any group-wide indexing requires constant burdensome merging.

Next iteration: lr-notebook

  1. per-user Lektor project, synced to cloud storage.
    • pros: Combined benefits of per-user notebooks with automatic per-project indexing, and cloud-based storage & version control. Editing can be done via native apps, or via browser using Lektor's cms-like admin interface. Lektor server provides API access to filesystem (!) so we can now build html views of all our content, and have in-page buttons that create/edit/delete files, etc.
    • cons: TBD

So far the following is working:
Automatic indexing of daily subfolders Monthly 2-column text & thumbnail view Per-subfolder indexing of days

Not working yet: