Wednesday, September 15, 2004 - Posts
A Spam Calculator from CommTouch . . . interesting methodology, but you have to consider the source!
I think I need to add Carlos Aguilar Mares to my list of computer-related “people I'd like to buy a beer.” I'm evaluating some reporting/charting options and his .Net WebChart control may be just what I needed. Yes, SQL Reporting Services is very nice, but it may be overkill for my particular task. Plus, I'm not real happy with using IFrames or other page parsing methods to get the charts out of Reporting Services1. I may be missing something (I'm far from a Reporting Services Expert), but CarlosAg's web control is light weight and ready to roll. The sample code mostly works on the site, and he even includes a step by step tutorial for the less adventurous users, so it was easy to get up and running.
For the record, my computer-related “people I'd like to buy a beer“ list includes:
- Bob Reselman (of CodingSlave fame); when I caught up with him at TechEd he was abstaining for a variety of reasons. He's still on the list.
- Anybody involved with WeProgram.Net; I think the one thing our user group misses is more social stuff -- including a beer or two. We had a nerd dinner cooked up a long while back, but it fizzled.
- Brendan Tompkins (since he has yet to attend a WeProgram.Net event, he doesn't fall under item 2). We haven't caught up since June.
- John Robbins from Wintellect -- as Darrell is fond of pointing out my non-sexual nerd crush on Johnnie R, and no list such as this is complete without it.
1 To clinch my decision, the customer isn't 100% certain on Sql Server as the back-end so I can't bet the farm on having a SQL Server available to report with.
Peter takes issue with FireFox, but I have to defend it. I think a little modification to the ClientTarget on the page would remedy this particular situation (although it might not -- I need to test it).
More generally, FireFox is compliant to a fault while IE lets developers get away with non-standards based HTML. This was par for the course back in 1999, but is it still acceptable in 2004? We should be beyond the browser-detection nonsense of the 90s, but we're only part way there.
I feel Peter's pain, though, as I notice many websites render oddly in FireFox -- although none that my organization works with as we've started designing to the same standards as FireFox (and occassionally having to bend over backwards for IE compatibility). Personally, I don't mind finding these “FireFox Failures” -- it tells me a little about the owners of the site and their sophistication and/or attention to detail. Of course, I do end up using IE in some cases to get at the internet content (Microsoft sites in particular).
At least with .Net the whole browser detection thing has gotten better, even if IE really hasn't.
Paul Laudeman and Darrell Norton did a fine job on their hands-on WeProgram.Net presentation last evening. They covered some of the “N” family of tools like:
- NUnit
- Darrel did this, of course, as he's the 4th resource (out of 6, I think) listed on the NUnit.org website (http://www.nunit.org/resources.html). He's really earned a reputation as an NUnit authority, so we were fortunate to have him!
- NCover
- Darrel did this one too; I messed around with NCover a while back, but I think it was the sourceforge NCover. The GotDotNet Workspaces NCover tool is preferred, they tell me, so I think I'll revisit the code coverage tool with this NCover instance. I think NCover is a great tool and will impress the pants off the customer when we deliver the coverage analysis along with our bill! Seriously, it's analysis like that can be a sign of professionalism and real craftsmanship.
- On a side note, I don't think the presentation mentioned NCoverViewer that, as advertised, ties into Visual Studio .Net.
- NAnt
- Paul handled this area as he's the NAnt expert on his project team. We had minor trouble with the version of NAnt (we installed a newer release than Paul was expecting), but everyone got the gist of it.
In a 1-2 hour presentation, it's tough to do everything justice but I think they did great. They left us wanting more (like NDoc, NMock, the list goes on . . .). These hands-on sessions are very popular and everyone seems to get a lot out of them . . . it takes more effort from the presenter(s), but the audience appreciates it. To sum up:
PowerPoint = OK
but
PowerPoint + Samples + Hands_on_time = GREAT
On a final note, I should point out that I paired up with local Bridge sensation Brian Lamprecht during the hands-on exercises and we played with FxCop during the NUnit material (we already have NUnit in the blood). FxCop was on the agenda, but we didn't have time to cover it so maybe next time. Anyway, FxCop could've done itself a favor by making extensibility a tad easier -- maybe with an FxCop project template included with the install or a more terse API. You see, we were messing around with creating our own FxCop rules and realized it's a bit of a pain to implement the interfaces because of how many methods there are; the future of FxCop shows some promise in this regard, and there are some sample rules on line, but it's still more difficult than it needs to be.
Anyway, thanks Paul and Darrel for a great session!