Wednesday, February 10, 2010

If the dataset is mine, then why can't I have it?

While I'm generally the pacifist type, I've decided that I really hate the software companies that offer fake interoperability for your data. I couldn't meet their gaze if we met on a deserted street. I would feign ignorance if they came up to me with a reference query. They're dead to me.

Over the past couple of months, we've been working on a project that requires keeping track of a heterogeneous group of stuff that we are going to put through the proverbial sausage grinder to end up with something more delectable. In other words: primary source documents for research.

This sort of thing requires lining up a selection of choppers and slicers, back to back, and having the pig go in one end and the spam come out the other. If the meat won't feed its output from one slicer into the next, then it leaves a mess of raw meat on the floor. It's not pretty.

So, as a result, I've been learning a lot more about databases and database connectors, and how amazing they are for lining up the choppers and the slicers.

These are the applications where I can make prodigious numbers of dataset love children and they will love me back, all the way from the slicer to the chopper. (Did I just mix those metaphors? That's just awful!)

These applications are the ones who love me back:
  • Zotero: with an API and an ODBC connector (via sqlite) I can spend hours immersed in your elegant data hierarchies, letting them wash over me, through me, and to me.
  • Filemaker Pro: as a fickle lover who only supports SQLserver, MySQL, and Oracle via ODBC
  • MySQL: as the plain faced lass who brings the water to the entire village, day in and day out (Man, now I'm really getting into trouble with the metaphors: pigs and babies being chopped up into spam by plain faced village girls? Why not just pull out a horror movie script?)
  • MS Access: I hate to say it, but over a decade ago, when I was just a wee noodle, I first learned about databases on Access. MS Access, you are the illicit, fickle lover who has a headache most of the time, but ah, when the migraine medicine finally kicks in, you have an ease of use that is all your own. You support ODBC connections that allow me to fetch and kick around data fast and loose, and then push it elsewhere before anyone notices.
These applications are dead to me, or at least dead-end in the flow:
  • EthnoNotes: You promote yourself as having an architecture that support ease of access to the data, but for how much work it is for me to actually get and use my data for outside analysis tools, you might as well just write it to a harddrive and drop it at the international space station.
  • Devonthink: Same as before, except that it is more possible, only that in order to get at the data, you have to be able to work extensively in AppleScript. Should I really have to take two classes in AppleScripting just to be able to do what any good ODBC connector can do in 35 seconds flat?
  • nVivo: You're different. You're dead to me because you have never heard of collaborative work. You have no version control, no way to merge files, no way to share a research project with more than one person in a way that isn't pulling teeth. And I can't get the data out either.
That's it for the rant today. Check back in later for more mixed metaphors of pigs, love-children, and the unrequited polyamory of software hacks.

No comments: