Merlin’s weekly podcast with Dan Benjamin. We talk about creativity, independence, and making things you love.
What text files do you use?
Mitch Wagner | Jan 29 2008
I started keeping text files of ideas a year or two ago, but the system quickly collapsed due to its own complexity. I am a journalist and a blogger, and so I started out with three files. -- blog ideas and article ideas. I also had a file called "inbox" for random thoughts, most of which would get turned into GTD next actions. The first difficulty I encountered was that it wasn't always clear, up front, what's going to turn out to be a blog, and what will be an article. Back then, I went by gut feeling, now I think I have some good thumb rules -- but either way, this decision should not be made at this stage of the process. Then I said to myself, "I really ought to group similar ideas together, because they're likely to all end up in the same article or blog." For instance, I'm a Second Life enthusiast, and I'm working up a list-type blog post or article: "N Easy Things Second Life Can Do To Make Itself More Useful And Attractive" So I really ought to group all those ideas into a separate file. So I started keeping separate files for separate projects. Separate ideas for separate contexts, too -- for example, I'm one of those people who gets only limited time with his boss, so I had a whole list with the filename, "@Tom." Quickly, I had a half-dozen lists, then a dozen, and eventually the whole thing got too hairy and I had to give it up. But then I heard Merlin's talk at Macworld, and he mentioned, in passing, while making another point, an "ideas" file. And I thought to myself, "One file for EVERY idea. That's the ticket!" Just open Quicksilver whenever I have an idea for something, invoke the append-to command, append the idea to the "ideas" file, and then move on. Read through the file and organize occasionally. Very much in the spirit of the "trusted system" in GTD. Only now I've opened a second file -- I've started a Facebook group for InformationWeek (the publication I work for), and I'm using the "Post" command to post links to selected articles. I like to do that once a day. When I see an article during the day that should be promoted, I append it to the "promo" group, and I plan to check that group every morning. I put next actions in OmniFocus. It's usually pretty easy right upfront to tell what's an "idea" and what's a "next action." Or it seems that way to me. Which leads to the question:What sorts of lists and plain text files do you keep? 44 Comments
POSTED IN:
 one file; two files; ledgerSubmitted by Anonymous Coward on January 31, 2008 - 12:38am.
i am a big fan of collecting everything into a single file. (actually i like two files at work where i prefer to have my personal ideas separated from the get go). a single file really makes sense to me as a collection tool because it's more than an inbox. as an inbox, the file must definitely be processed (and whatever difficulties exist in doing that are an independent problem). however, unlike an inbox, even the things unprocessed can be useful. a single file has the added advantage that a particular item can be retrieved non-sequentially (by a search) even before processing takes place. this is important because processing and the weekly review can lag desired action in any system. this may seem sloppy, and you're part right. my biggest gripe with this system is that trouble begins to brew because the system loses its clean edges. the single file becomes a reference, a mini context/action list, and a collection device. this can work for a while, but i think one ends up creating sections, moving stuff around, and growing the file so large it becomes cumbersome and full of junk. it's far too large a project to solve on a weekly basis and i am not enough of a believer to think that habits can solve this problem. instead, i think the one-file system can be improved, with a little bit of magic. ... but i have not yet done it. my idea: everything that enters the single file should be recorded like a ledger--a permanent record that serves as later reference. simultaneously, the file should have a dual nature, so that portions may be deleted and removed/checked off, expanded, processed into the real system: contexts/actions/projects/20k/... the kicker is that portions unprocessed by a certain time should become stale, and "purged" from the file. it's not gone (everything is forever in the ledger anyway), but it's moved out of the file to a dated unprocessed section. these sections were not important enough to be converted, so they're moved out of the way automatically. i think this could be done in plain text, and a program could just delineate "thought sections" by delimiters or a couple EOLN markers. each section would be in ledger and time stamped, and sections that "come due to expire" can appear on a daily report, and be deferred once before no longer archived and no longer tracked. independent of the ledger & automated clearing, i do find myself creating new "percolation" text files after i recognize a theme of a number of related ideas that emerge, but have not manifested themselves into an action. (examples: quotes, words i'm collecting for some reason; resources on a topic (e.g., for a book i might write); music-to-obtain queue; possible gifts for others; shit i want). these lists are not different than david allen suggests, but i didn't know where to put them until i framed them as "percolation collections" (meaning buckets without specific actions yet). they differ from the main input file in that they are collected thematically (once a theme is evident). » POSTED IN:
|
|
| EXPLORE 43Folders | THE GOOD STUFF |
