Dev::Blog

Paint by Numbers:

Basically, someone had the bright idea of looking at fine art collecting as an investment. Turns out that well-picked art appreciated at a rate of over 10 percent per year. There have been studies like this for all kinds of commodities (I always like the one that compared buying a case of beer, drinking the beer and returning the empties as being better than buying Nortel stock), but what’s really cool here is that they took this knowledge and developed a kind of pension plan for artists.

Artists in the plan donate 20 pieces over a 20 year period, and when they’re done, they’re left with a fund that might be worth $1.5 million or so. I don’t know that this would work with the “long tail” of artists out there (the fund looks for people who sell pieces for $5K to $10K), but it looks like an innovative safety net for those it can help.

 

I’m incredibly stoked about the latest MSDN article about registration-free COM. True, it only works on XP, and I have no environment remotely like that at work (I think we’re going to jump to XP shortly after we roll out 2000. Yes, we’re racing against Longhorn. Pity us.), but the time’s coming pretty soon. One of the biggest problems with enterprise development is the testing cycle that’s required to make sure that no application contributor does anything to interfere with any other application. With .Net and XCOPY deployment, we were close, but there’s always some legacy COM code that needs to be around, and once you’re registering objects, you lose the ability to say that you’re definitely not straying from your playground (yes, I intentionally avoided daying “sandbox”). It’s not perfect (doesn’t work with ActiveX EXEs, for example), but this will go a long way to easing DLL dependencies. Like I said, stoked.

That’s mostly from a desktop developer’s perspective, but I also need to check out side by side assemblies in WinXP and 2K.

It’s highly likely that there are simply a billion different technology groups at play at MSFT, but I’m going to go with the theory that the company’s finally recognizing that .NET isn’t the be all and end all of app development, and neither is Avalon/Indigo/whatever. VB6 retirement notwithstanding, it’s nice to see some kind of (inferred) acknowledgment of COM’s future. Let’s hope there’s a migration path that doesn’t make it the next DOS.

I made it to an MSDN “deep dive” event today on ASP.Net 2.0:

1. Microsoft has been getting some flack from some in the software community for their constantly shifting platform. I’m reminded of the late 90’s when I thought I’d teach myself some data access stuff, and I discovered ODBC, DAO, RDO, and OLEDB. I put those plans on hold and waited for the dust to settle. (I’m just now getting into ADO.NET.) Today’s summary of 2.0 gave me the impression that I was wise to wait this long, at least. There’s a lot of cool stuff available and they seem to have addressed a lot of the concerns from the earlier versions (I’m liking the little things, like the increased use of flow layout so the designer doesn’t wreck my HTML markup.)
2. Of course, with all the cool stuff we saw, it’s kind of silly to call the event a “deep dive.” The presentation was good and had lots of demos, but it really wasn’t any different from any other vendor presentation where they try to wow you with the cool gizmos but only scratch the surface of the underlying technologies. I’m fine with the format, but the name doesn’t provide much credibility.
3. Speaking of gizmos, the focus really seemed to be on reducing the amount of coding that developers have to do to make web pages. There once was a time that I wanted to do every line of code by hand. Oddly, it’s only since I started getting paid by the hour that I started seriously straying towards the land of wizards and code generators. I still need to understand what I’m pasting in so I can extend it, but my productivity with the new Microsoft tools has soared.
4. Bringing points 1 and 3 together, I suppose Microsoft’s long range plans are less important when it comes to small and medium scale web development. If a new technology comes along that requires changes to one of the application layers, that’s really all that gets obsoleted. I’m curious as to when we’ll see a point where DHTML pages get archived like COBOL programs.
5. Finally, I was confronted with a situation I never thought I’d experience as a man: at the break, the lineup for the men’s washroom was without exaggeration about 30 feet from the door. There wasn’t a lineup for the women’s washroom, since there were all of 5 women at the event, and 3 of them worked at the venue (OK, a mild exaggeration). If we want to make progress on the gender gap in IT, we should post job ads in the washrooms in bars…

I ran into a bug recently on a job importing dates from a feed of events into a MySql database - I was using DateTime.Parse in C# to bring them in, then formatting them as MySql DateTimes. Of course, what I didn’t notice until later was that the dates were from the UK, and they came in as day/month/year, while my computer was set for month/day year, so March 12 was being read (and saved) as December 3.

The fix was easy - just specify the format when you parse - and probably should have been included in all DateTime.Parse calls I’ve ever made:

dt = DateTime.Parse(myDate,
     System.Globalization.CultureInfo.CreateSpecificCulture(”en-GB”).DateTimeFormat);

This got me to thinking… Since the project’s coming down to a crunch (so what else is new), I thought I could do my own parse bug, and say we thought the month and day fields were backwards on the schedule.

Sadly, the clients are one step ahead - they made the due date April 4.

04/04

Bastards.

Dave Eggers (and interviewer) on procrastination and productivity:

OK, I’m starting to recognize that I’ve got a few too many things on the go (or at least, the things I’m doing are a little too intense), but that’s pretty much how my head works. I imagine it’s a common pattern.