I was reading some older posts on Robert Scoble's weblog, and noticed one about naming categories of applications. I agree that it "doesn't matter what you call it". It's either useful or it isn't. It reminded me of a discussion I once had about the different types of heavy metal.

I've just found out that I've got a new job! I'm leaving my current position because there has been very little C# (or any other sort of) programming work this year, and I've been on SCADA projects instead. The new job is only 15 minutes from home, and I'll be working in a barn conversion overlooking the pleasant countryside views. Lovely.

In contrast to my current job, they use NUnit for unit testing, there is a team of software testers (one more point in the Joel test). They didn't get me to write any code for the interview, but I did take a personality test, and an aptitude test (which I'll write about soon).

I'm really looking forward to starting in 4 weeks time. I've been at my current job for six and a half years, and I will greatly miss some of the people here. But still...

Last night my wife asked me to do a minor spot of DIY. This involved screwing a piece of wood to the bottom of the stairgate, so Emma, our ten-month-old, wouldn't crawl underneath it (the stairgate was originally fitted to defend against a toddler). I looked at the stairgate, and saw that there were already two screws on the bottom, holding the two sections of gate together. Easy. I'll unscrew those, put the piece of wood in place, and screw the screws in again. Once I'd marked and drilled the wood I realised that they weren't screws, they were bolts, but never mind, it would still work.

When putting the bolts back in, it transpired that they were measured to fit only the two sections of gate, and no extra wood, so they didn't actually meet in the middle any more. Oh well, it just needs a couple of holes drilled underneath the bolt holes, and I'll use screws to hold the wood in place.

Somehow, when putting in the second screw, the screw broke in half, flush with the wood. There was no way to get it back out, so I drilled a second hole. Whilst drilling, the bit broke, and got stuck in the hole, so I had to drill a third hole with a different size bit.

It's not pretty, but Emma can't crawl underneath. It took much longer than the 5 minutes Julia asked of me, but I'm not gifted in DIY. The whole episode reminded me of Eric Sink's Iceberg Sneak-In article, where a simple request for a new software feature turns out to be a whole lot more work than it appears.

I had the results of my personality test today, and I am stunned by its accuracy. I'm not very self-aware, so some of the observations were new to me, although once I'd read them I heartily agreed.

Being a software engineer in a small team, I was encouraged by phrases like "Mr. Harriyott needs sincere appreciation and a feeling of being an integrated member of of the team", and "Although he is naturally logical and systematic in his dealings, he emphasises these characteristics when under pressure". These sound like useful qualities to have in my job.

There were a couple of things I'm going to work on, namely stubbornness and inflexibility. I suppose the testers are going to make the report sound as positive as possible, but I would benefit from reading it critically as well as egotistically. I found the test really useful to understand how I react, and I will try to choose situations that suit my personality in future.

Six years ago, I started work at a small software house in East Sussex (England). I wrote an intranet, using ASP and SQL Server, which provided links to procedures and information about how to get things done. Everything was in one place. The purpose of the intranet was to be useful to the staff, and make everyone's job slightly easier. It achieved this aim, and was well used.

Three years ago, we were bought by a large engineering consultancy, which has a really big intranet, with thousands of pages of useful information, procedures, internal advertising and so on. Content is added and updated by individuals across the organisation (including me) . There is a search facility that returns dozens of links. There are local areas for different divisions. It does everything.

My intranet was (quite rightly) decommissioned. All my resource booking pages were replaced by Exchange Server, and our procedures changed to fit the existing big company methods (which can all be found on the intranet).

However, the staff couldn't easily find anything. The search facility returned too many links, most of which weren't relevant. When somebody did find the correct procedure, they might have emailed the link around. Some people would keep the link, and others wouldn't. The next time the procedure was needed, the intranet and the email folders were searched. It took longer to get things done than before.

Last week I created a "shortcuts" page, containing links to the pages that we spend ages looking for. The links are grouped into task-based sections, without the need for searching or storing links in email folders. When anyone in the department needs another link added, I will add it. Everything will be in one place. The purpose of this page is to be useful to the staff, and make everyone's job slightly easier. I've come full circle.

I bought my son Toby a new bed from Ikea, and assembled it one evening earlier this week. Thinking about my previous adventures with flat-packed furniture, I opened a bottle of wine and put on some mellow jazz beforehand, in an attempt to limit the stress.

On the whole, everything was as it should be; all the components were present, the instructions were in English and had plenty of pictures, and an hour later I felt like a skilled carpenter. There was only irritation; the bolt-holes for joining the legs, frame and head-board were slightly too small, so I had to line up three pieces of wood and bash the bolt through with a hammer. It worked, but not at exactly 90 degrees (and it did create some sawdust). However, I'm sure Ikea put a lot of time into usability testing, and have already ironed out several problems that I didn't have to deal with.

There seemed to be four options for buying a bed:

1. Buy some planks of wood and some screws, and make it from scratch
2. Search for a second-hand bed
3. Buy a self-assembly bed
4. Buy a brand new bed, and have it delivered and installed

I have sorted the list in ascending order of cost, which also happens to be descending order of time investment. Most people have either plenty of time and little money, or lots of money and no time. A few lucky people (e.g. aristocrats) have money and time in abundance, and a few unlucky people (e.g. new parents) have little of either, but mostly people have one or the other

Software is like buying a bed. I use some expensive but time-saving software, and some dirty freeware applications (like my spam filter) that need constant tinkering with, and guessing what the buttons do. I remember when notepad had non-standard shortcut keys, so Ctrl-S didn't actually save the file, and Ctrl-F didn't produce the search box. It was frustrating, but it saved me buying a proper editor.

Writing software is the same. Usability takes time to write, and time to test, and makes development more expensive. Quoting from Ikea's "how we're different" page:

It's not easy combining good design and good function with the right quality, at an affordable price. But then we aren't known for taking the easy way out. Designing a desk that costs a fortune is easy. But designing something that's affordable to many - only the truly talented designer and product developer can do that.