There are mutterings abroad that Windows – nay, even the “personal computer” model itself – may not be the productivity enhancers that we signed on for.

Have a look at these links .. or if you know all this, and are wondering what it has to do with data analytics, just skip to the next section.

Please note .. this is not just another “bash Microsoft” rant .. it is a somewhat serious first look at how we can manage (personal) computing complexity, what we can do and how we can work our way out this “fine mess (we) have got ourselves into”. (I could, and maybe will, write a parallel commentary on the maze of modern statistical computing… some day).

Windows Words of Doom (John Dvorak)

When using Microsoft Windows, there are three words I dread to hear: Preparing To Copy. You see, these three little words are wrapped up in a mess that’s been around pretty much since the first versions of Windows, and although I haven’t used Vista enough to know for sure, the ongoing copy chaos seems like some sick joke that must be absolutely hilarious to Microsoft insiders.

…the seven idiotic Windows glitches that make it so difficult to move files from point A to point B.

1. The Time-Fluctuation Phenomenon
2. The Delayed-Query Syndrome
3. The Long-Filename Anomaly
4. The Drive-Full Abort
5. The Multiple-Folder-Slowdown Paradox
6. The Yes-to-All Baffler
7. The Cannot-Erase-File Gotcha

OK, there are some partial solutions to this particular problem here
Preparing to Copy? NO — Robocopy!

re robocopy

and you might like to look at my previous post on refactoring the API.

Of course, you will note that the Robocopy “solution” has been around a while now, but Microsoft has not told the average Joe about it .. it is in the nature of witchcraft, after all, to keep certain things secret or only available to the initiates who know where to look. Which comment is not meant to denigrate the fine engineers at Microsoft, but simply to point out that what to communicate and how and when and to whom to communicate are non trivial issues in a computing landscape of seemingly almost infinite complexity : why such apparently intrinsic complexity? a flaw in the model?

This Digitally Mundane Life (John Dvorak)

How many hours per day do you waste using technology? Probably too many

I wonder if anyone will ever do a study of how technology has actually harmed the quality of life of the average person who has fully adopted the digital lifestyle. This isn’t to say that the word processor and the information Web site haven’t benefited us all in lots of ways. I’m talking about the fact that, with the computer in particular, you have to take the good with the bad—and it is starting to look mostly bad.

And by bad, I mean time-wasting. Seriously time-wasting. Do we really need to send e-mail correspondence to the same person sometimes three or four times in one day just to say that you agree with something? And of course to even find this correspondence we have to wade though an inbox full of e-mail junk. People are spending hours on e-mail daily. HOURS! Instead of being a great convenience, it is a great burden.At that’s just the tip of the time-wasting iceberg. What about Web surfing or getting involved in threaded debates? There goes the day.

Blogs get a mention too.

OK. What he has said is “obvious” to all of us. What is not so obvious is how to get some measures of the problem for our own personal use, how to apply some analytics.

Why CSS bugs me(John Dvorak)

As we move into the age of Vista, multimedia’s domination on the desktop, and Web sites controlled by cascading style sheets running under improved browsers, when will someone wake up and figure out that none of this stuff works at all?! The current PC platform is so close to being permanently broken that I’m stunned that people aren’t already up in arms. Everyone should be sued for false advertising.

As most readers know, I’m a blogger. I’m in the process of redesigning the Dvorak Uncensored weblog, which means playing with its cascading style sheets, or CSS. The first time I heard of the cascading style sheet was in December of 1996, when the World Wide Web Consortium announced CSS1, telling the world that Microsoft, Adobe, Netscape and others were “among” the consortium members that would be adding support.

The idea behind CSS is a good one. With HTML, Web pages can become monstrosities of content and formatting information. To change the look of a site, you have to sift through the content to redo a lot of detailed information. It’s painful. CSS was designed to separate the content from the formatting, so that when you want to change your site’s look and feel, you simply change the formatting information.

CSS’s real benefit was that the layout not only could be changed easily but also could become dynamic: The content is stored in a database and presented as necessary, with instant updates. With dynamic content, it’s possible for 100 people to go to the same Web site and get 100 different versions.

Here is where a great idea begins to fall apart. And it does so progressively, worsening over time as “improvements” are made.

OK, it is clear he is not a CSS expert. But he has a point .. CSS IS ARCANE.

I have spent hundreds, almost thousands of hours with CSS, and it does work well if you spend the time to become an expert. But in all my time with it I have never been able to understand why it was necessary to develop a new language just for CSS .. it could, after all, have been done with XML.

XML is no model of clarity either (and don’t even get me started on DTD’s and XSD’s and XSLT) but it is at least a standard of sorts and we do not need language/specification dialect proliferation.

Context Fragility

Briefly, the PC presents such a complex environment that it is fragile, and a particular source of this fragility is vulnerability to program and component upgrades (and their mostly unforeseeable consequences). Upgrading or installing a new version of one of the modern IDEs (Integrated Development Environments) can result in a cascade of upgrade requirements for the installed parts – days r weeks of work, or even sheer impossibility.

Everybody knows it, but nobody’s talking

OK, none of this is news to those of us who have been working for years with that mixture of witchcraft and arcana known as the Windows API .. we know that we are working in a realm where the time required to program the simplest task is grossly unpredictable, and there is a good chance that what appears straightforward on the surface will lead us into Kafkaesque depths of layer upon layer of seemingly arbitrary “rules” and “operating system dependencies”.

What is new, and interesting, is that these ugly truths are now being “outed” and talked about in the popular press.

How can we analyze this? – and what to do about it?. The need for desktop analytics

Well, we cannot get a measure of the problem without data. We cannot simply survey people and ask them what they think. We could, but with dubious quality, ask them to log their experiences – but apart from the obvious shortcomings, such a design is inefficient in that it will capture relatively few of the rare but serious events.

What we need is some close recording of what a person is doing, and what apps they are using.

Websites have logs recording which pages were visited, how long people stayed there – this is the basis of web analytics.

So I propose that we venture into “desktop analytics” – log which apps are being used, and for how long.

Of course that does not answer the real question, does not measure “productivity” but it is a start and maybe a suitable proxy.

And it is not an easy thing to do .. it involves installing system-wide keyboard and mouse hooks, and monitoring application switching. And the records would be unlabelled.. that is, we would not know why the user was doing what he was, or what the extraordinary time spent in a particular activity was related to : the problem can be partially solved by asking for labels for unusual events.