the corner office : tech blog

A fun diversion whilst trying to solve a Project Euler problem.

SO: Fastest way to determine if an integer's square root is an integer

Which also led to the history of the magical inverse square root in Quake3 code: Origin of Quake3's Fast InvSqrt()

Hrm: InnoDB File-Per-Table Tablespaces

I spoke too soon. Turns out my comment spam filter isn't as effective as it once was. Cue some cleaning up in the backing db and locking old blog posts. Let's see if that helps.

I like having a blog, but there are times when I find myself questioning the amount of time I choose to spend on it.

I'm doing what I threatened to do, even though I'm sure there a million other pet projects I'd rather be faffing with. I'm rewriting my blog. The servlet app is going, and in its place something which does the static file thing. Except that I like having comments support and what I currently have works relatively well at keeping auto-spammers away (despite being a noddy solution). So I want to keep that bit as a simple servlet app. And that means dynamic updates of pages on a server, which means it's static pages but a process taking care of things.

I wanted to look something up on my blog. I knew I'd blogged about it. Google didn't believe me. I revisited my Google web tools page, and over the year, the number of pages indexed went down, and is now coming back up. But coming back up, faster than I'm blogging pages. So it's finding pages which it had forgotten about. Bah.

I could generate a sitemap. But before I do that I should get cracking on something I wanted to do a while back, and convert the blog to something that's file system based (perhaps just a servlet to handle comments).

I still can't find the page I was looking for.

• Greg Young: CQRS, Task Based UIs, Event Sourcing agh!

• Martin Fowler: CQRS

CQRS is a an approach (I won't say pattern) I've liked and used in numerous ways for a long time, I hadn't realised it had a name.

I have entered the world of C# with a vengeance. I'm writing a C# client for my old metalib app.

All the data's in H2, and I wondered whether I should play around with some other db (or NoSQL db), but in the end, decided to take a stab at doing what H2 suggests, and run H2 with .Net. I discovered IKVM.NET, and installed it, and managed to fire up the H2 server running IKVM (not strictly needed, but nice to know it works, woohoo!) and then built myself a dll, and copied some code to see if it would work, and it does (woohoo!!).

I'm happy and impressed. What other fun stuff can I do with IKVM?

Since the little "Get Windows 10" icon now hogging a small bit of my screen real estate opens an application which then tells me my laptop doesn't support Windows 10, perhaps you could consider doing the compatibility check in the background AND STOP SHOWING ME THE DAMNED ICON IN THE FIRST PLACE.

Kthx.

I've been wanting a personal wiki for ages. I've half-started a couple of Deep Research forays in the past, but this was the first time I got as far as paying attention to TiddlyWiki. This time around I had neither the time nor the inclination for Deep Research, and just as well. The usual "Righto, I Must Now Research The Hell Out Of Personal Wikis" instead turned into me thinking sod it, people say it's great, just make sure it can export OK and give it a bash. And that was that. Really nifty software.