I have a good memory, but some of the fields I'm working in (NLP, Semantic Web, Information Visualization, Machine Learning) are quickly expanding.
While I do have a good method of keeping track of good research (I only follow the best researchers and research groups on Twitter, Scholar, conferences, journals, etc), sometimes perhaps it's better to save all the interesting papers, add notes to them and put them in context and relationships with other papers (if a paper frequently cites another paper and is a sequel to that paper, perhaps it's better to read those 2 papers together, for example). Essentially the things I want to do are these:
- keep records of the workshops, classes or tutorials I attended, but also video recordings of presentations I have never seen in real life
- annotate PDFs, extract metadata from them and generate bibtex bibliographies
- keep tutorials in pdf + video format synced (where possible)
- keep everything in context - organized in folders, tagged,
cross-referenced if possible (if an article is cited in a paper and I already have it in my library, I want to be able to access it fast)
- I want to be able to search about various keywords/tags regardless of the location where the actual files are stored and OS.
I would preferably only sync a huge zotero (or similar) folder using Dropbox, and papers/tutorials should be in sub-folders named after their research field, but also tagged so that I can find them regardless of the folder.
EDIT: I have paid Dropbox, but apparently finding free WebDAV is an issue (therefore that would mean an additional 5-10 euros per month). As far as code goes I use GitHub and GitLab.
After seeing your answers, it occured to me that I'm thinking about it too much. Perhaps the best idea is this: the whole workflow should be a webapp - this way it will perform almost similarly on all operating systems, as even if there are differences between browsers, they are not so big as the differences between OSes.