• sites with a narrow specialty make material more discoverable, and thereby more useful to someone trying to find it. people find my photos through the google (why such interest in druid hill park? and fleas?), but are even more likely to come to flickr or picassa looking for images. a link tagged on del.icio.us contributes to a global pool of information on a given topic.
• i get the advantage of code from real programmers instead of my own stumbling efforts.
• i don’t have to maintain anything.
• popular services attract many people with many ideas and needs, leading to features i wouldn’t have thought of or would never have gotten around to implementing. (geotagging photos on flickr, for example.)
• it’s far more satisfying to complain about someone else when something breaks.

it hasn’t even meant giving up control. services that want to be used have to allow exportation of data to standard formats, making it easy to switch from one service to another if needed, or to go back to your own creation. i could pretty quickly write a script to make my own links page from my del.icio.us rss backup or generate my own photo gallery from the backup of my flickr photos and their metadata. it’s easy to go back if 1999 ever calls you.

my goal has been to find services which can be automatically backed up by a cron job on my always-on computer (the account at my hosting company). as a result, when i make a spreadsheet on google docs or post a photo to flickr, i know it will soon make its way into ~/backups/ and my own backup system for my files. (do i have to link here again?)

i’ve made a list of tasks, services i use for them, and backup methods for those services. some are personal tasks (tracking todo items), and some are methods of sharing information (videos on youtube). privacy controls allow many to be used for both (a calendar which reminds you of your dentist appointment and helps arrange a meeting with other people). it’s a rather exhaustive list, but i’d welcome your suggestions of other services you use or any answers you have to the questions in the comments.

task	service	backup method	comments
address book management	? (not: plaxo, google contacts)	?	i can’t understand why no one has filled this obvious need. it seems like a natural: make a site where i can record all my contact information (address, phone numbers, etc.) and the same information for people i know. as other people join the site, they can request to connect to me, and i decide how much information to share with them. as i update my phone number, they get a note that it’s been changed in their address books. run it on open standards so it can interoperate with similar sites. let it connect to my email service for “insert joe’s email address in cc:” and to my documents service for “insert joe’s mailing address in this letter”. let me export my address book to various formats and print a copy. plaxo had the right idea, then developed it halfway and made something that doesn’t seem useful in any real way. i don’t know why google still hasn’t done something with their anemic contacts system. why has no one made a reasonable online address book? or is one hiding somewhere?
articles	wordpress?	mysqldump? (i’ve never used wordpress.)	see notes below.
audio cataloging	mp3tunes? what else?	?	mp3tunes is a pretty good idea and a surprisingly good implementation, offering support for mp3, ogg, flac, etc., tools for organizing files by albums, a good browser-based player, and methods for downloading various slices of your collection, syncing them to a digital audio player, etc. if it had anyone but michael robertson behind it, i might be seriously interested. also, its syncing program is gui-only and can’t be run from cron. i’d love to have something like this which lets me organize my collection online, especially if i could share selected files with the world (ones for which i own the copyright). does anyone know of any similar offerings?
book cataloging	librarything	a www::mechanize script to grab /export-csv	a handy way to keep a list of books i want to read.
bookmarking	delicious	wget download of https://api.del.icio.us/v1/posts/all
document editing	google docs	the google docs download python script	painful after vi or emacs, but great for sharing documents, both for viewing and collaboration in editing.
email	gmail	offlineimap	i would prefer one-way syncing, and was using isync (mbsync), but have found that offlineimap runs anywhere, with less fuss. 1
financial accounting	mint	download through browser (non-automated)	i just download my transactions file each month after balancing my accounts.
goal tracking	43things	n/a	don’t feel the need for a backup of this.
graphing data	google chart api?	n/a
photo archiving	flickr	offlickr	offlickr does a thorough job of grabbing metadata, from comments and dates to sets and collections.
scheduling	google calendar	wget downloads of the .ics files for each of my calendars.
score archiving	mutopia?	a recursive fetch of everything with my name?	it could be convenient to let the good folks at mutopia keep my scores up-to-date with the latest lilypond versions. it could take care off all but the few not worth adding to such a noble project.
todo management	toodledo	wget download of my rss file
video storage	youtube	?	youtube’s a good place for sharing video for quick viewing, but is it a place for storing high-quality video? i assume they throw the original files away once they’ve converted them to .flv. does anyone have a recommendation of a place for storing full-sized videos while sharing small flash versions?

i hope there’s something useful in there.

i want to ask one last bit of advice, concerning my personal website. aside from offering bits of news about what’s happening with me to far-away friends, it’s been a place to stick things i want to keep which may be useful to other people (musical scores, hatt-baby videos, etc.). it’s the public side of my home directory. with the moves described above, a lot of this is no longer necessary. i don’t need photos here and on flickr. i don’t need scores here and on mutopia.

when i strip away everything that can be as well or better-hosted elsewhere, all that’s left are articles like this one, so i’m thinking of turning jeffcovey.net into a hub for my material spread around the web, leaving just the articles here. i’m thinking of wordpress for this. yes? pros/cons? alternatives?

i’d like to start using something new right away and gradually migrate my old material. how would you do this? my first thought is to make a cms-run new.jeffcovey.net so the links to jeffcovey.net continue to work, adding 301 redirects as i move things off jeffcovey.net, and finally resetting dns to point jeffcovey.net to the new site when finished. any better ideas?

thanks for reading! i hope you found something useful in this and are putting the ever-expanding world of online services to much better use than i am.

1. btw, a handy tip: i found offlineimap’s two-way syncing helpful in an unexpected way. i have a corporate email address with a certain company and am supposed to read email sent there. unfortunately, they won’t just forward email to my real address because the competition might steal our tps cover sheet design or something. i can’t have gmail fetch it because they only allow imap access, not pop3.

   i was having getmail fetch and forward mail to my real address, but found another solution in the fact that offlineimap can not only sync an imap account to a local maildir collection, but can also sync two imap accounts to each other. i have the webmail system at the company file everything into a folder named (just as an example) “forcesourge” and have offlineimap sync it with a “forcesourge” label in gmail. this works great because of the way gmail treats labels as imap folders. any new mail at the company is mirrored in my “forcesourge” label collection in gmail, and when i delete a message or remove the label from it there, the message is deleted on the company’s server. a cron job keeps it all together, and i don’t have to bother with checking another mail account.

i’m sitting in shelter #92 in patapso state park on a glorious cool spring day. the sunlight makes the leaves glow as the breeze tosses them, and the pair of does grazing the hillside next to us have finally wandered off. the picnic remains are back in the trunk, and glenn sits across from me, reading the paper.

it’s a simple outing, but, until recently, one that would have been hard for me to take. at the very least, i would have been sitting here anxious about getting home, instead of committed and engaged in being here. the woods would have been wasted on a mind full of all the work left undone.

the culprit, as i’ve gradually realized this year, is the computer. last year, i would have brought my laptop with me. i would have done what i could here, then hurried home and put it online. for my day in the city saturday, i would have broken my back lugging it onto the light rail, where i would have brought it out of suspension to get 15 minutes with it before i had to close it down to take it over to the subway, repeating the process all across the day. last year, i either had my fingers on my laptop, or i wasn’t working.

the problem is that a computer, especially the big computer we call the internet, is infinite. i can turn on a computer at 7:00 am, and it instantly becomes 11:30 pm, and i’m looking back on a day of youtube and wikipedia and chasing this which reminds me of that which makes you stop and wonder whatever became of those and what was i doing and why does my back ache? enough days like that, and you’re looking back on a life of wandering the electronic sinai.

i was spending long hours each day glued to a screen and brief, almost guilty ones in the physical world. the world is better and healthier, and i’ve worked for the last few months to flip the time i spend in each. i’d like to share some of the ideas and tools in my arsenal and invite your suggestions and your own experience.

how to live without a computer

millions of people today still go through lives untouched by lcd screens and laser mice, and all the bachs and shakespeares of history did reasonably good work without them, so it must be possible. is it preferable? i hit techno burnout recently and took a nine-day no-computer break to recover. what i noticed most by midweek was that the days had become much longer. they expanded and opened to me as they hadn’t for years, until they reached a natural length i’d forgotten they held, and there was plenty of time to be spontaneous, cook a good meal, listen to some music, sit on the deck awhile. reminded of what i was missing, it was painful to go back to staring at the screen and waiting for a webpage to load or an overloaded machine to start responding again. i’d like to spend as many days as i can with no computers in them.

how can a person enjoy the undeniable advantages of computers without getting sucked down the time drain? what are the ways and means of a neo-luddite life? i’ve identified two strategies. i’ll call them computer agnosticism and computer freedom.

computer agnosticism

computer agnosticism is disbelief in the doctrine of the one true computer. for a long time, my laptop was the only computer i used. it had all my files and was configured just as i wanted. it’s still my most comfortable workplace, but now any reasonably new net-connected computer is as good as any other. this gives two advantages. first, i can step out of the house anytime or travel to another city with worrying that something i’ll need is left behind. everything’s available wherever i can get online. secondly, trusting the world to provide a computer for me means getting a situation that is workable but not overly pleasant. i can sit with my laptop forever. if i’m stuck with illegible fonts, an unfavored operating system, an uncomfortable chair, and a room of boisterous library or hotel patrons, i’ll do what i need to do and get out of there. libraries are especially good, with a timer counting down and the next person on the waiting list hovering in the background.

online services are the key here. here are some tools i like for making any computer “my computer”:

toodledo

when i began implementing the ideas in getting things done, i wrote my own gtd script for managing my projects and todos. over the following years, i regularly enhanced and refined it. by the end, it had a powerful, good-looking interface for manipulating tasks. tasks could be scripts in different languages, launching complex repetitive jobs with a couple of keystrokes. it handled repeating and scheduled tasks. it attached and detached screen sessions associated with various categories of work so jobs could continue in the background while i was off to something else. it integrated with a scriptable window manager and switched to a different virtual desktop each time i moved to another work category. it managed time devoted to different projects, work categories, and other things to do, counting down remaining time on a taskbar and changing the color of all the windows when i should do something else or take a break. it was beautiful.

surprisingly, it was not only efficient, but effective. i really did get a lot of good work done through it and accomplished things that otherwise would have stalled. so why did i let it go? it was a fine way to work on my computer, but i no longer wanted to be tied to my computer. it was still the right answer, but the question changed.

i looked at several online todo list/project management applications and settled on toodledo as the most open and flexible. the development team frequently extends and bugfixes it, and i’m generally happy with it as long as i actually do the tasks on my lists instead of tweaking and pushing them around (not toodledo’s fault). one dealmaking feature was the ability to print a pocketmod-style todo list, so wherever i’m working, i can print my current open tasks, stick them in my pocket, and walk away from computers for as long as i like with a clear conscience.

gmail

the no-brainer replacement for mutt + procmail + postfix. the spam filtering alone makes it a winner. the interface is great, and when i’m on my laptop, imap and mutt let me burn through my email even more quickly.

google calendar

for a time, i synchronized this with the datebook on my palm pda. since i can’t do that from any available computer, i now print a one-month calendar on one side of a sheet of paper and a ten-day agenda on the other (the agenda includes my notes about events, addresses of places to be, etc.).

google docs

a pattern seems to be forming. this is another winner for google and the home of my spreadsheets and the documents i need to share with others.

del.icio.us

the firefox extension lets me use bookmarks as i normally would at home and have them available anywhere.

linode

(or any hosting service with shell access. linode is terrific, and i’ve been hearing good things about “nearly free speech“).

an always-online linux box gives me a way to connect to other servers as needed and otherwise fills any gaps left by web-based services. ajaxterm over ssl gives access from every library computer i’ve encountered. unison keeps everything up-to-date so that changes i make online appear on the laptop, and vice versa.

computer freedom

so now i can roam the world freely, treating computing as a public service, like water fountains or buses. that’s a big relief to my mind and my backpack-unladen spine. can i go all the way? what could enable someone who works online to spend days together completely free from computers?

some of my answers:

cron

when i moved my recurring tasks to toodledo, i looked at each and asked, “does this really need me to be present and active?”, and was amazed that 95% of the time, the answer was no. it can take some sideways thinking, but even tasks that “clearly” have to be done by you in person can be automated with a different approach and a bit of perl. i enthusiastically expanded the role of my crontab on my always-online computer, and now i just check the resulting email messages to see that all went well. any changes made to my files are pulled down by “unison -batch” cron jobs on my laptop when i’m not looking.

rss has been a lifesaver for getting out of news-checking quicksand (remember when you went to your favorite sites to see what was new? repeatedly? all day long?). i stick my google reader feeds into “monthly”, “bi-weekly”, and “weekly” folders and check them only when their time comes. for sites without feeds, a netstiff cron job lets me forget them and alerts me to anything new once a week. hpodder downloads my podcasts so they’re just there when my mp3 player runs dry.

non-instant messaging

irssi + bitlbee + screen running on my always-online machine let me engage in irc and instant messaging services on my own schedule. instead of letting them interrupt me, i just attach the screen once a day (or week) to see what’s new.

paper

let me say that again:

paper!

this is remarkable technology, flexible, portable, and widely available. as with cron, it may take a new way of thinking, but how many of the things you do on the computer could be done on paper? how many could be 95% completed on paper so that when you go online, you just have to transfer the results or execute the last steps of a paper-outlined plan? is an offline day stealing from your boss, or are you even more productive when you print your work the day before and worth it with a pen? which accomplishes more — 45 minutes work on paper + 15 minutes bringing the computer up-to-date, or two and half hours on the same task online, with side trips to check email and rearrange your netflix queue? which leaves you refreshed and ready for work the next day?

this article is an example of paper-based computer work. i’m writing it in an old college composition book, much more portable than my electronic notebook. how it will get online is discussed in the next section.

while on the topic, i should add my plug for the famous moleskine notebooks. they’re pricey, but are things of beauty, simple and elegant and a joy to write in. they fit right in a pocket, and with my toodledo list and google calendar in the expanding back pocket, one makes a fine pda (ppa?).

vpas

the virtual personal assistant marketplace is flourishing, and online helpers can help keep you offline. they can be useful on both ends: before fleeing the net, you can hand over a list of online chores. when you connect again, you can send what you did offline for processing. this article is an example. when i’m done, i’ll tear the pages from the notebook, scan them to a pdf, and email it to an assistant to type into a reply to me. a little editing and formatting, and it will be ready to post. lather, rinse, repeat, and you have an invisible staff working alongside you all your offline days. you’re taking full advantage of the net and full advantage of that table in the park, putting out more work and pulling in more fresh air.

tim ferriss discusses the details of working with a vpa. i’ve been happy with onassist.com.

phone-based services

sooner or later on a non-digital day, there’s a moment of temptation to boot the laptop just to check the weather report before leaving the house or to get directions to the coffee shop. and since you’ve sat and waited for it to come on, you may as well make it worthwhile by checking out that site you saw on a billboard. and though you definitely weren’t going to check your email today, you’re so curious to see how mike replied to… wow, you never expected him to say that. how are you and connie going to fit that into your schedule? it’s going to be on your mind all day now…

phone-based services can give you the limited information you need without exposing you to the danger of spoiling your free day with full net access. i hope they continue, and don’t fall in the face of web-capable phones. three i use regularly:

goog411 (1-800-466-4411)

goog411 provides the aural equivalent of google’s stripped-down webpage look. it’s simple and powerful and lets you quickly get the address of a business, then connects you to them to ask directions or how late they’re open. i hope google adds support for residential listings.

tellme (1-800-555-8355)

provides news on various topics. i use it to get the weather report, to decide how to dress for the day.

jott

this one’s for putting information in instead getting it out. i have it linked to my toodledo and google calendar accounts so i can call to add an item to my todo list or put an event on my calendar. it’s a good way to save some typing, though their voice recognition leaves a lot to be desired. they should license whatever google uses for goog411.

conclusion

for someone professing to write about computer-free living, i’ve spent a lot of time discussing computer-based software and services. i have two excuses:

first, many of these tools provide ways of letting work pile up out of sight and mind. email piles up in your inbox as long as you like while gmail sends replies with your phone number and the request that people call you if it’s urgent (and grandcentral separates the wheat from the subsequent chaff). articles wait in your rss reader until the time you’ve dedicated for them. just keep that laptop safely tucked away (a remote shelf in the basement works well), and life becomes less interrupted.

second, and even better, are the tools that run unattended. cron jobs run even when you’ve forgotten about them, and vpas have pleasant surprises waiting when you get around to logging in again. jott’s added a reminder to check out a wine festival next fall. services like these uphold the promise of computer automation and let you get on with real life in the real world.

there’s another solution, of course — to drop the halfway measures and just go cold turkey forever. maybe it’s worth stopping by the park office to see if they have any jobs for the summer.

it’s very simple:

1. when a call comes in, the phone should check the incoming number against its phonebook.
2. if the number is found, it should put the call through right away.
3. if the number is not in the phonebook, a recorded message should say “you have reached the cellphone of [click, followed by my voice] jeff covey [click]. if this is the person you wish to reach, please stay on the line, and i’ll connect you. if not, please hang up and redial the number you were trying to call.”
4. after about five seconds, if the caller is still on the line, the phone should start ringing.

since small phones with small number keys became commonplace, i’ve received many more “wrong number” calls than ever before, and this seems to be true for everyone i know. just delaying the incoming call for a few seconds while the person on the other end decides if this is really what she wants could noticeably reduce the number of times each day we go scrambling to find the phone for no reason.

please mention it to your friend’s sister, the one whose uncle’s ex-wife’s stepson’s cousin’s roomate’s boyfriend is the godson of the ceos of all the major phone companies.