I’ve been dark for about a week whilst I’ve been delivering two training courses here in Canberra. As a result I haven’t updated my blog much and tonight (well – last night, its past 1AM now) I spent quite a bit of time catching up on it all.

There is lots of stuff that I had casually read but just didn’t have the time to reply to, but I’ve managed to get through most of it. I’ve got some stuff that I want to read and post on tomorrow and then hopefully on Sunday I’ll have some time for my pet projects and to get stuck into preparation for my Code Camp Oz sessions.

P.S. We now have about 270 people registered for Code Camp Oz, when I get a chance I’ll do a City-by-City breakdown of where people are coming from.

I thought this post by Johannes Ernst was spot on and really eloquently put some thoughts that I’ve been having for about a year or so.

A big part of the problem I think is that we have a lot of really smart people bogged down building the same business application over and over again to support the same out of date business practice.

Where are the unique applications for our talents. I actually think that the next world changing technologies aren’t going to be necessarily derived from someone who works in the technology industry – but you can bet that some of us are going to have to write the software that supports it.

I say not necessarily because it might just happen that the next visionary does work in IT – but thats probably not where the vision came from, although you can bet that they were into sci-fi or something like that.

AJAX Fan Fare

April 8, 2005

Dare Obasanjo makes a good point about the hype that is rattling around the blogosphere re: AJAX, but also SOA and REST. I find the whole AJAX thing amusing because I have seen some of our customers doing some truely amazing things around buiding compelling and responsive web applications.

The issue to date has been around tool support for the approach that AJAX promotes – I’m not really in a position to say whether AJAX solves that problem, maybe it does.

I think that the client callback feature in ASP.NET 2.0 might be a better abstraction for you if you are coming from the .NET side of the fence.

Brain Drugs

April 8, 2005

An interesting piece over on the Computational Complexity blog about a theorhetical situation where academics/researches take drugs to improve their performance. Lance draws the obvious parallel to drug testing in sports and asks the obvious question – “why should professional athletes be held to a different standard than professional academics?”.

Everyone likes to think that you can have one rule for everyone, but the reality is thats just not the case. In this instance most people working in sciences are ultimately working for human advancement, although I am sure its a case of three steps forward one step back from time to time.

In athletics the game is very much about competition with ourselves, and in the spirit of competition its best not to use drugs. Although – if everyone had access to the same drugs then would that be OK?

Dunno. I guess the answer depends on how much we care about the long term health of our athletes? Do we not care about the long term health of our academics and researchers?

 

How do we build?

April 8, 2005

Scott Hanselman posted up a good piece on his blog about how they layout their source tree. He mentions quite a few good tools but gets down into specifics about how they lay things out. My builder managers toolkit pretty much includes these things:

  • CVS or SVN today, Hatteras tomorrow.
  • NAnt today, MSbuild tomorrow.
  • FxCop today and tomorrow.
  • Draco.NET today, Team System tomorrow.
  • MbUnit today, built-in VS2005 tomorrow.

Now when it comes to laying out my filesystem – I do this and it just seems to work really nicely.

  • ModuleDirectory\
    • Module.build
    • Module.FxCop
    • Source\
      • SolutionDirectory\
        • ProjectDirectory\
        • ProjectDirectory\
        • ProjectDirectory\
    • Documentation\
    • Tools\
      • NAnt\ (copy of NAnt)
      • MbUnit\ (copy of MbUnit)
      • Build\ (custom NAnt tasks)
      • FxCop\ (copy of FxCop)
    • Templates\
      • AssemblyVersionInfo.cs
    • Keys\
      • Organisation.snk

Of course, this is just one module in CVS – it fits into a much larger picture.

This article on Wired News about Windmills in the Sky is facinating. The concept seems crazy but strangely plusable (I had to check the date on the post). The Sky Wind Power site answered alot of my questions/concerns.

A few weekends ago I started on an ambitious project to remove as much paper from my life as possible, starting with the pile of paper that needs to be filed that is building up next to my desk – I was inspired in a big way by this video up on Channel 9.

To date my wife has been doing an admirable job taking care of filing activites but its time to give her a break – and make that information more accessible at the same time.

Looking at the tools I have at my disposal, there is a scanner. Microsoft Office and a bucket load of storage that can be easily backed up to DVD. So I started scanning, and on that first weekend I scanned a hundred or so pages using the Microsoft Document Imaging program.

The nice thing about this is that it performs OCR on the document text and embeds it as meta-data into the TIFF or MDI file formats, then when placed on the file system, the scanned documents can be keyword searched, so if I wanted to find out when I called a particular phone number I just type it in and bingo, it gives me a list of all the phone bills that have calls to that number in them.

As I started working through this material I determined a few things (in no particular order):

  1. I need a scanner with a paper feeder.
  2. The out end of the scanner needs to be attached to a shredder.
  3. Tagging is required when OCR doesn’t work.
  4. Where do I store this crap?

Thats why my eyes lighted up when I read this post. I think we are close to a time where we can outsource our personal records management to some third party that we TRUST. Now – I suspect this is a game that both Microsoft and Google want to be in, MSN “Spaces” should be a big hint.

The problem is, I don’t know if I trust either organisation to protect my information forever and a day. I haven’t been burnt personally but I know a few people that haven’t been happy with MSN changing the policy on what services are free and what services cost a fee.

And despite the general consensus I feel that as Google grows its “not evil” moniker will change to something more inline with shareholder values (if they wanted to change the world they shouldn’t have floated).

What I really want a solution that integrates well with the tools that I use today – and tomorrow, so the underlying storage model needs to be flexible (WinFS, if it ever arrives will be a compelling client-side version of what I want), but that should only be a cache – I want my information accessible anywhere and on any device.

I want to know how much it costs today, and how much it will cost tomorrow, I’m not stupid enough to believe that this service will be free, I’d be willing to pay a small monthy fee (indexed with inflation) for the rest of my life to:

  1. Have the information available anywhere.
  2. Have it searchable.
  3. Be able to do bulk extractions and inserts (to move providers).
  4. Never run out of space – even if I decide to copy the entire contents of a DVD into it every month or so.
  5. Index everything including video, audio and text.

Number three is kind of interesting I think. Since this thing is going to record my life, its going to need to record my work information as well – e-mails, documents, everything. However, you’ll find that in most employment agreements you need to surrender corporate information if you change jobs. So I need to be able to say give me all my work related stuff and export it to this media.

Anyway – its certainly an interesting area, and something that I think could take computing to the next level in terms of utility.