A List Apart has two articles about web design. The first is about use cases, to make sure you are building the right application. The second is about how to manage scope. So one shows how to build enough, and the other shows how not to build too much.

Both are good reads, especially if you are a web developer, like I am.

Carl Franklin, the host of DotNetRocks and Mondays is going to be starting a new show called the Daily Commute. I'll listen. It took me a long time to get into Monday's, but now that I have, I think it's hilarious. There's something about crude humor that I like.

The Daily Commute will be the Slashdot of the podcast world - he even says News for nerds in his description! He says it's something new, but not new. It's a podcast (is that really old news now?), but it's new because it's externally focused - not a personal lovefest.

He says it won't be available for a week or two. That gives me time to catch up on my other podcasts. And 15-20 minutes? That'll work. I can listen to that on the way in, and still have 10-15 minutes for Bob and Tom!

He's looking for a lot of listeners, so he can make these three shows a full time gig. He's also asking for our help to get the word out!

Here you go, Carl. Here's my part! Good luck. I hope it works out. I know I'm looking forward to it!

Eric Sink has chapter 7 of his Source Control How-to up, and it's about branching. It's not really a how-to in this part, but a why-to and a whynot-to. It's a very good read, as usual, and gives some very practical reasons to branch, and then offers words of caution about branching. Next time, he'll be talking about merging. I'm looking forward to that one.

In the mean time, go read about Nelly and Eddie.

This week, I hit 500 subscriptions in Newsgator. Part of the reason is that I have started subscribing to individual feeds at the asp.net blogs. Why? Because I noticed I was always gettting 25 posts whenever Newsgator synched with them. That means I'm most likely missing some posts. That's fine, but not if those posts are from some of the people I want to track. So I started subscribing to individual feeds. I use the aggregated feed to find other individual feeds to subscribe to.

I get around 1500-2000 new posts per day, and so far, I've been able to keep up just fine. It takes roughly an hour or so to wade through them, and, to be honest, I don't read everything. Like Scoble's said before, you can scan the headline and determine if it's worth reading or not (hint, make sure you have good post titles). Most are either good for a glance, or I just skip it altogether. If I start seeing things repeat, in most cases, I can find a link back to the original post. The blogosphere is an oddly small community when it comes to that.

Anyway, I'm half way to Scobleville, at least on the reading side.

Has anyone else used MyBlogLog before? I heard about them today and decided to try them out.

What do they do? They track outbound clicks. That means if you click on a link (from my site only - no feed stats), they track it, and then tell me where people are going. That's pretty cool. It's easy to see referrers, but not quite so easy to see where people are going.

I find this interesting because I mainly provide links out. I don't write a ton of original content. I watch what others are posting about, and link to them, providing my (useful) insights. Or something like that.

I'm going to test it on my main blog for a while. So if you want to help, head to the website, and click on a few links!

Jeff, over on the asp.net blogs, asks how do you persist enumerations. So far, he has two comments. One does it by the numeric value, and one does it by the text value.

We do it by numeric value, but I like the reason for doing it by text - Steve says he does it because reporting is better with a readable field.

We actually have an interesting way to do most fields. We don't use Enumerations that much, because we want those properties to be tied to a lookup table. But we want to be able to code against those lookup values in certain cases. So what we did (good or bad, I'm still not sure) is to use a constants namespace, with classes for each lookup table. Each class has a constant for every value in the table. Now, these lookup tables don't change. Yes, I know the perils of saying something doesn't change, but we also have a Codesmith template that can regenerate the constants.

Now we get quite a bit of benefit from this. We get nicely named integers, since the constants are named based off of the description field in the lookup table. We get SQL constraint checks for those values, since we have foriegn keys to the lookup tables. And we get something we can code against when needed.

We still use enumerations for some things, but thier usefulness has been diminished by our use of constants.

I recently got an email from someone asking why an ASP.NET application wouldn't work for him. I asked for some more details, and here's what I got:

It's a Windows 2003 Enterprise Edition server with plenty of RAM and a pretty fast CPU. It's running as a Virtual Server, with IIS 6 and the .NET 1.1 Framework. He get's a "Server Application Unavailable" error when he navigates to the application.

It's a fresh install. I did some quick searching, and saw that others are having the same problem. It works on Windows 2003 Standard, but not Enterprise. Unfortunately, I have yet to see a solution.

Anyone experience this issue? How did you fix it?

Brian built a custom action for the Tourney Pool Manager installer, and every day that he talked to me about it, he kept asking me if I knew how to debug an installer. I told him I didn't, but other day, I came across this post by Josh Robinson. He explains that by adding this line:

System.Diagnostics.Debugger.Break();

you can debug the custom action. Very cool!

Brian has been actively getting the word out about the Tourney Pool Manager. Hes' doing a good job. There are many free press release services, and he's tapped into a few of them. You can see a couple examples here and here and here.

On another note, Brian is petitioning WebHost4Life to add the Tourney Pool Manager as one of the applications that can be installed automatically. If that's something that you'd be interested in, you can email WebHost4Life and let them know. That would have made it even easier for Alex Lowe to install it!

I'm a little late with an update, but it looks like it's already over. No one picked the champion, and no one picked either of the two guys in the championship to be there. So the pool was over after the quarterfinals today!

The year was marked by many upsets. The winner of the pool will end up with 70 points. As a comparison, the winner of one of the NCAA pools last year, using the same scoring, had 127 points. Ouch.

I would like to thank everyone who entered. We really appreciate. You gave us a great way to run a pool and see how it will work when the real deal comes around. We found a few issues (nothing major - more on that another day), and they are all fixed.

So, now that you've seen the process, what are we doing wrong? What could we do better?