All I talk about is my research

So I'm going over some class reviews and I see this sort of comment a few times: "All he talks about is his research/past jobs/work experience." I'm teaching a class that, when it's time to talk about database design, or web design, or e-commerce, or Management Information Systems, for instance, you'd better believe I'm going to mention something about how to do design on the job, in The Real World. Then. Only Then. When it fits in. Yes, I may discuss things Not In the Textbook, but that means that anyone sitting in the classroom will have to pay attention, lift a pencil and write it down. That's college.

Research. I'm researching transportation, personal information management, task management, and workflows. It involves databases, XML, and implementing some generic "task processing language" in a real language, something I'm picturing to be a two-step transformation with XSLT. It's interesting stuff. I discuss almost nothing about it, except at a very simplistic level about "current research", since I'm reading journal articles published this year on the topic, and the textbook is a decade behind the times on some of this stuff. It's current. Knowing this stuff will be a job requirement within a decade, useful for people who will be graduating and hitting the job market in about 3-4 years or so, as they still have time to learn the stuff that will keep them employed.

I've worked in IT since 1990. I've had a few jobs, and I've been actively involved in the issues of my industry. All this has happened before...

And I like how "research" is regarded. The Media talks about "cutting out extras like research to focus on teaching". First of all, universities live on the overhead from research grants- faculty research is a major revenue source. So cutting that means that other, non-funded-research departments don't get subsidized anymore. Second, cutting research means the end of graduate and PhD programs- these are supposed to be research-oriented degrees, though they've often devolved into supercharged undergraduate degrees.

Innovation is part research, part creativity, part just plain hard work. The research part isn't optional. The undergraduate degree does not have an adequate research component. And without scientific and technological innovation, our society is not going to survive the messes we're in. Like it or not, research is an essential part of academic life as well- I'd more question faculty who are not involved in some research.

Personal Area Network Information Management Stuff

Am grading papers amidst piles of small devices, thinking of a lecture about Personal Area Networking I just gave. On the table is a bunch of SD cards in a bag, and other scattered devices. Sometime in the two weeks before heading to Vancouver, I've got to think about what to say about that, too. Sometime after dinner I'll go back to my Task data structure, and the literature about a Task Processing Language. Only, the Task Processing Language has to optimize on time and location. What a mess that will be.

Paper done- time to move on finally!

It's done- I finally submitted the revised conference paper. I'm too brain-dead now to start anything new. Pathetic: I'm too tired/lazy to type in the citation information for my own paper on Zotero. Maybe by morning they'll post it for me.

My wife is watching Changeling on DVD, and from what I'm overhearing, I'm not up to trying to pick up the last 20 minutes. (I'll wait until it's done, then watch Community on DVR, desperately hoping it'll be funny this week.) Actually can't type much, nor have the mental energy to web surf.

But this drawing has been staring at me for quite a while. It's been up since last Spring I think. TB drew it, with the title "Gingerbread Man Running Through a Blizzard In The Forest".

I've got nothing to follow that.

Browser is the new email?

Reading Domain Shame and thinking, what do browsers and email have in common?
I've just finished sifting through a virtual mountain of academic journal articles on PDF for my frequently-mentioned conference next month. A bunch of articles not used discuss tasks and email, in that email is the "lowest common denominator" for managing your life- 0r in other words, anything you can use to replace email will almost by definition be an improvement.

But, isn't that true for the browser? Anything that can replace the browser is a better way to read stuff. Cases in point- the reemergence of browser plugins, feeds, etc. In the 90's there was a flurry of app development that used "browser windows" as a plugin within another application, to leverage the benefits of HTML as presentation without its limitations. Maybe this is more an argument against HTML as data format. But the growth of RSS, iCal and non-browser data formats should back this point up.

What if you could separate Facebook as data store from Facebook as application? The latter sucks. The former is kind of interesting. Double ditto for Craigslist.

The core argument is the rise of mash-ups, and the need to integrate locally stored info with whatever you do on the browser. The problem is that browsers make this really tough- unless you store all your documents and other info in the same place. "Cross-domain" scripting is a no-no because it would make hacking you so much easier.

You can build browser applications combining different kinds of personal data, if it's all stored in the same domain. Nowadays, that domain merging all your personal information and hosting lots of productivity tools is one that, fittingly, starts with the letters "Goo".

Time to rethink this whole browser thing? If we go back to data feeds and local applications that do the grabbing for us, rather than having the browser do it all for us, we don't need to be covered in so much "goo". Anything that can replace the browser as it stands today would be a huge improvement.

Last minute scribbling for first conference paper

Just about done with a bunch of revisions for my first conference paper- the ASIS&T one in Vancouver this November. I spent much of my free time over the past month or so fixing it, while trying to rebuild my dissertation proposal. I got stuck a bunch of times, finally reaching the realization that I was writing the wrong paper, and loaded that text into another set of documents- funny when you get stuck working on one paper while having something else entirely different you want to say- or when having nothing at all to say.
Zotero has been exceptionally useful, when it doesn't screw up. It's been great to quickly keyword search through the hundreds of PDF files I have. On the other hand, it seems as though a few citations aren't what I remember using, and a couple don't make any sense at all to me. But that may have been just my hasty editing, which will take a careful examination before the final deadline for revisions this Friday.

It's annoying how difficult it is to make a halfway decent graphic in Word. I had to import something in Visio finally. Feels like a real hack.

On some level, writing a literature review is kind of fun- or at least it is when you finally get a handle on the research, and once and for all figure out what the gaps are. Before that, it's a bit frustrating to make sense of it.
Zotero has been good for toggling PDF files, managing the metadata (tags, keywords, and citation stuff), though the critical stuff is still missing, or at least the stuff I'd like to see. I'd like it to represent a conference as a distinct entity, which in turn contains multiple papers. Ditto for books- which are often just collections of papers and bound as a volume (just as with a conference.) In both cases, the separate papers are likely to have only a passing resemblance, not to mention a series of authors which may repeat or reshuffle from one conference to another.

Tracking references would be really useful- that some papers are referred to by other papers. There's no way to do that, but loading that connection would be extremely useful. You can mark another paper as "related" but the direction of that relation is lost, unless you adopt a policy of only relating a paper to a "reference" that is. True, you could hack it and assume that a reference to an "older" paper implies a citation direction, but you'd still lose the ability to "relate" papers from the same conference, for instance. This really needs to be cleaned up, and it only takes the ability to attach an attribute, "relation type", to the relation record attached to an article.

I'm done with the current draft, emailing it to a few people for comments / feedback, and hopefully can submit this thing a bit early for a change. I bought my plane tickets today and made hotel reservations, realizing that three weeks from today, I'll be on my way home again.