October 24, 2004 23:28 | Confession / Features / Technology

Process through adaption

Continuing on my current theme of "data management", let me share with you an example from my own experience.

I mentioned a few weeks back how overwhelmed I was by the size of my MP3 collection.

McLuhanists, and others surely, will often cite how disruptive any new medium can be in our lives, viscerally even, until we learn to "deal with it". (What constitutes a "medium" is ascertainable by applying the McLuhan Tetrads, something I very much want to explain in english sometime.) Also, it seems we are in a situation where we are creating mediums faster and faster and with greater and greater disruptive effects, at a pace we are essentially unable to cope with. But I digress. But it is relevant! Anyways...

So, my MP3 collection.

I can identify three distinct phases in my archival methods for MP3s, each corresponding roughly with changes in available technology. I say roughly because while yes new technologies had evident impact, later on there was not only more MP3s around but also more technologies (apps, protocols, methods) and each had different absorption rates in my digestion of them. Heeee!

Phase 1 - Discovery (or Willy Nilly)
Oh my god. This is SO cool.
Going back to my old archive CDs I see that starting late 1998, I start having a folder called "mp3" into which I literally would just dump mp3 files. The files would be named with the track and artist name. ID2-3 tags were barely present and if so, hardly used really. Each of my archival CDs from that period was named after a character from whatever book I was reading and I figured that was plenty of a mnemonic device to remember that "Archive CD Jacopo Belbo" had a copy of "The Dukes of Hazard" theme song. We're talking maybe a dozen or two mp3s per archive CD. That worked for about a year.

Phase 2 - Hunter Gatherer (I start building a home)
Whoah, I need to put this stuff somewhere I can find it.
Napster hits like a hurricane. We all spend hours and days hunting for individual tracks, making little piles, renaming files and scrubbing ID3 tags. Every now and then a friend drops by with a full album ripped directly from a CD. Blessed be thy name. Alphabetized folders containing /artist_name/album_name/tracks*.mp3. I haven't added to that archive in over a year and it sits at about 17Gig. (The previously mentioned blog entry was about my malaise of trying to figure out how to incorporate all my new stuff into my old scheme.)

Phase 3 - We're all in this together (The Village)
Hey, do you have X? I can send you Y... it's sorta like Z.
This is where I am right now. All the mp3s I've amassed over the last year (36Gig - notice the increase in volume) sit in an a hierachy which begins with a folder named for *where I got it from*.
Notice I still keep trying to keep everything neat in artist/album folders. Thing is, my work is greatly reduced now, because I either get them already like that, or the ID3 tag information, which was entered in a distributed fashion by all the other mp3 sharers in my social network, is correct and I can easily rename stuff programmatically if I need to.

So again, my malaise from a few weeks back was because I was trying to shoehorn my new reality into my old system. Realizing the ground has shifted allows me to just accept that I need to stick with this new system. For now.

So what's next?
Well, it hasn't started yet, but I imagine something like last.fm figures. Information like "who" recommended a track and "why" and "how" will supplement the "where from".

More freeform and shared taginess is something I'm also very interested in seeing applied here... if 200 listeners "categorise" track X or artist Y as "indie electro laptop rap rock", who am I to argue? And why would I? I'll have a way of finding it easily.

Sidenote: I am one of those iTunes users who has selected to disable that software's "Keep Music Folder organized" feature, because I like doing organization on the filesystem level. Again, the malaise comes from being in a situation where that is becoming less and less feasible. I will have to give up eventually I suspect, but not until the management applications catch up to the shifted ground themselves. In this case, iTunes would need to be tied into my (future, networked) AddressBook... ;)


tons of info touches upon the need to store and to find when one needs critical info: the fbi visited my brother yesterday and had all sort of questions to ask me, they did not meet me at the time they had arranged which is making me worry.


2- philippe shink

fascinating article by Anthony Hempell. I very much enjoyed it. You wouldn't happen to have any more interesting articles on coms. theory lying about in your bookmarks would you?

[man this preview function is disconcerting]

I can identify with these phases you've observed. But as far as I can tell you're making your organisation overly complex by thinking about the metadata in the filesystem. You need to standardise the filesystem in a simple manner (I use Root-Genre/Artist/Album/) or not worry about it at all and employ the metadata capabilities of iTunes to make sense of it. At phasee 2 I swapped to iTunes from Audion - and spent entire days (no, weeks) adding tags (genres particularly, for browsing) to my collection. Throughout phase 3 I stuck with this and now presumbaly at phase 4, having ripped all my CDs and aquired many others (1200 albums) I know where everything is. That said I'm now throwing stuff out that I could actually keep, for the sake of better organisation.

If you wanted to use the origin (who, where) of a track you could add it as a 'grouping' to those tracks. You could actually compare the grouping feature to Flickr's tags. You'd then use smart playlists to select tracks matching a particular 'grouping'. I don't use this myself as I'm happy with genres, and anyway grouping all my tracks would be a nightmare beyond considering...

In terms of networked music organisation, iTunes' own music sharing is a glimpse of how things may be, but Goombah is probably closer. But in reality we're more likely to see iTunes store features make into the organisational scheme.