A new line of work
Not too many of you know what I actually do for a living.
I write tools with which other people in the company do what we call "command line data mining".
This is simply a process of chewing through terabytes of data to produce useful information; anything from "what images are most popular" to "if you enjoy this site, you might also enjoy these other sites".
As you might guess, command line data mining has a fairly long ramp-up time, as you learn the various tools and techniques. There has long been talk of creating a large document explaining all the ins and outs, both to get new people up to speed quickly, and as a reference for those who mostly know the ropes.
After a long and lackadaisical search for someone to write this thing, no one was ever found that had both the talent and the availability. Eventually my desire to have this document exceeded my desire not to be the one to write it.
So now I'm a writer, I guess. At least I don't have the tech writer's usual problem of getting access to the developers, since I'm usually available whenever I need to talk to me.
I've certainly enjoyed learning a new set of tools. I'm writing it in docbook, which is an XML schema. You mark up your text just so, and then other tools hammer on it to create a web site or a PDF document or whatever.
As I try to document things, I see where there are holes or inconsistencies in the tools, so I document what the tool ought to be, and then jump over and change the tool to match.
It's hard to stay on task, as I'd rather be doing pretty much anything else, but life's hard all over, I suppose. I'm trying to enlist people at work to feign interest, to keep me motivated day to day.
Stay tuned for the further adventures of Tech Writer Man
I write tools with which other people in the company do what we call "command line data mining".
This is simply a process of chewing through terabytes of data to produce useful information; anything from "what images are most popular" to "if you enjoy this site, you might also enjoy these other sites".
As you might guess, command line data mining has a fairly long ramp-up time, as you learn the various tools and techniques. There has long been talk of creating a large document explaining all the ins and outs, both to get new people up to speed quickly, and as a reference for those who mostly know the ropes.
After a long and lackadaisical search for someone to write this thing, no one was ever found that had both the talent and the availability. Eventually my desire to have this document exceeded my desire not to be the one to write it.
So now I'm a writer, I guess. At least I don't have the tech writer's usual problem of getting access to the developers, since I'm usually available whenever I need to talk to me.
I've certainly enjoyed learning a new set of tools. I'm writing it in docbook, which is an XML schema. You mark up your text just so, and then other tools hammer on it to create a web site or a PDF document or whatever.
As I try to document things, I see where there are holes or inconsistencies in the tools, so I document what the tool ought to be, and then jump over and change the tool to match.
It's hard to stay on task, as I'd rather be doing pretty much anything else, but life's hard all over, I suppose. I'm trying to enlist people at work to feign interest, to keep me motivated day to day.
Stay tuned for the further adventures of Tech Writer Man