So, first off, be mindful about what’s likely to happen to a folder’s contents hours, days, months, or years from now.
- When and under what conditions will the need for this information reappear in your life?
- How will you find this information when you need it? Is it stored where you’ll first look?
- Does this information exist in solitude or are there other bits of information, tasks, and appointments that should always stay with it?
- Can you combine a few folders into one? Or would this benefit from being broken out as its own thing?
- Does the naming, granularity, and location of this information support its likely future purpose? Or does it just reflect today’s anxieties and your weird semantic peccadillos?
Above all, try to envision the future moment at which this information will become useful and necessary again, and make sure your filing and piling support that scenario and lead quickly to any needed actions [which should be scripted anyway, but whatever].
So, second, and more to the point of email and physical “pending” folders, I think it’s useful to think of all the information in your world in terms of potential activity. Remember that demonstration from 8th grade science? The bow drawn back represents potential energy, and the arrow in flight is kinetic energy. Don’t get stuck thinking that kinetic action is the only game in town, and definitely and don’t let your byzantine folder system lull you into missing all the action potential currently unmined in your files.
The danger of too much foldering in your email program, in particular, should be self-evident. The more folders you have, the more thinking you have to do on both ends of information and action management: you have to first ruminate on the “right place” to put that email and then you’ll again have to recall where that right place was once you need it again. Is there a way you could just convert it to an action right now and be done with it forever? And is an email folder actually the best place for you to store a particular piece of information that you’ll need again someday? Only you know, chief.
For me, these folder structures just get simpler and simpler all the time; 90% of my email work now goes from “In” to either “Respond/Action” or “Archive.” What else is there to maintain? Do you really need five levels of time- or project-based archiving when you have a modern search-friendly program like Mail.app or Google Desktop? Maybe. Again, it’s your call.
Bottom line: ensure that all the folders, buckets, nets, and boxes in your life exist to support action above all else. The short-term buzz of “getting something out of your way” will fade quickly and is way offset by the future hassle of having to dig it out of your crazy nested system later on. Organize to act, not the other way around.How to organize your files during your research career -- bibliography dir, 'done' dir, scripts/hacks, a dir for things to learn, todo, log, etc. Mail yourself things to do. Every paper and directory has its own directory. Use version control management tools. You might want to use general unix conventions like ~/bin, ~/lib, ~/pkg, ~/tmp. Be sure to keep a portfolio of your work so that you can take it with you whenever you go to talk with people- papers, publications, screenshots, code snippits that you are particularly proud of, stats, photographs, schematics, etc. You already have this on your website, of course, right?
I think it's -partially anyway- about the fact that none of the steps you take to increase productivity are automated? Whenever a path changes, your shortcuts lose their value. Unless you update them manually.
my home hierarchy often moves , but I follow two rules I found to be critical for an efficient hierarchy:
1. no more that 8 folders at any point in the hierarchy
2. no more than a 3 level deep hierarchy
what's more I use two other 'tips'
a. files can be stored at any level (not only leaves)
b. files can be stored in multiple places (via hardlinks)
those numbers are not 'cat /dev/random', they are key numbers in human brain memory model: the instant memory (memory that stores data seen in a blink of an eye, literally) can memorize about 8 objects, and the brain has a hard time keeping up with deeper than 3 levels of hierarchy without resorting to time-costly abstractions (~context switching)
this way I can navigate faster, as I can recall the hierarchy of a thing very quickly and as a whole, and in each level I can see and "compute" all folders at once and not resort to some linear or dichotomic search navigation.
also with 8 folders, graphical navigation (in a file browser) can be greatly improved with visual clues: nautilus emblems make navigation a snap.
what's more, spatial mode with 3 levels max is really useful and much less cluttered than some people think.
notes:
- it's awesome how people overlook tip (a.) and end up only putting files in leaves of the tree, often creating dubious leaf subdirs for that purpose (and having a hard time figuring a name), when the file could well enough fit in the parent.
- (b.) gives me the advantage of directories behaving like both a tree hierarchy and tags. indeed what I do is some kind of tree sorting where in my memory I associate a file with a tag, and that tag belongs to a known, capped hierarchy. so my search has a cost of max 1+3*log(8)/n, n being the numbers of tags the file is associated with.
- the home versioning idea is indeed a good one. it somehow reminds me of the wanna-be TimeMachine of Leopard, except that you have total control over it. with my hierarchy, I rarely have a need to 'go back': what doesn't belong to the tree is outdated, and gets moved in a 'archives' accumulating directory. now that I think of it, this 'archives' dir is just like the .git dir, except it's manually managed. also I was thinking of having the home files hardlinked into archives (on a 'commit'), so that I can just delete a file and it would still be on 'archives', while not taking anymore place when not yet deleted. I'll have to think of it and read those articles above.