Monthly Archive for: ‘July, 2004’

Microsoft have acquired Lookout Software, manufacturers of a personal search-engine add-in for Outlook called Lookout.  I’ve just given it a quick spin and it’s fantastic – for anyone who’s ever spent time hunting through thousands of old emails it’s a Godsend. In a few short minutes it has indexed 21,776 outlook items dating back four years, enabling me to find things I’d long since forgotten about.  Download it here [860kb]

Jez emailed:

Yay!  MSDN magazine is now available by air mail to “international” MSDN Universal subscribers for the bargainous price of $25 dollars a year (which is hardly anything in real money).  That beats paying five quid a copy in Borders.

Update (25/02/2006) – Hey, we now have String.IsNullOrEmpty, you know!

What’s the quickest way, in C#, to test whether a string is empty?  This is a question which came to my mind yesterday.  I’ve always been fond of:

if (myString==String.Empty)

but it occured that this necessitates the creation and destruction of a second object (the empty string with which we’re comparing), so maybe it would be quicker to write:

if (myString.Length==0)

And looking through some existing code I also found examples of:

if String.Equals(myString,String.Empty)

I decided that an experiment was in order, so wrote a quick windows application to try each method in turn. I was also interested to see if there was a difference in performance where the result was true than when it was false.

Here are the results, with times in milliseconds for 50 million iterations:

So there you have it. Experimental evidence that the fastest way to test for an empty string is to test whether it’s Length property is zero (and that this also takes exactly the same length of time irrespective of whether the result is true or false).. unless anyone knows of a quicker method? Of course, since we’re only saving ourselves around 50 nanoseconds per iteration against using the slowest method shown above, this is hardly going to result in a noticeable performance gain!!

In recent months I have begun recoding the Filthy Lucre web site and associated systems, pretty much from scratch.  The database has been redeveloped, the middle tier objects are just about finished, and some previously dodgy vbscript scheduled tasks have been replaced with a much more stable Windows service.  It’s looking good.

So, now I’m at the stage where I want to redevelop (and enhance) the website, which is where most of the changes and rework will inevitably take place.  The decision I have to face is whether to stick with coding it in VS2003, or take the plunge and use VS2005.

VS2005 looks great, and has lots of features which I know will save me much time – Master Pages, for example, will make life much easier than the convoluted basepage system we currently have to code manually to achieve the same results.  I’m eager to get my hands dirty with VS2005, and use some of it’s new features in anger, on a real-world application, and I know that using it to redevelop the FLIC site would achieve this.  But on the other side of the coin, it’s obviously still beta software, so I fear I’ll get so far down the site creation process before finding that some functionality I require isn’t yet implemented.

What to do, what to do?  I think on balance I’m going to have to dive in head first and start getting used to 2005, even if it does have some quirks.  By the time I’ve finished the FLIC redevelopment the RTM version will probably be on the shelves anyway