Doing Design

There's no point discussing design methods without explaining the plan - and notice that there is a plan! Never start a design without one.

For me, it was easy: readers told me what they wanted. Let's make no bones about this - you may think your ideas are good but you will never do better than listen to users. I had a short list of what people wanted.

• To be "prettier"
• To be able to add stuff
• It should always work
• To see new changes easily

Pretty it up

The first is probably the most common request and almost impossible to be sure of. Its rare for any two people to think changes make it better the same way.

I seriously considered using a genetic algorithm to make site changes depending on the way it is used, mainly because the new templates make it fairly easy. I even tested it before I realised that the big problem was the design would be the basis for the rest of the linked sites - and if it kept changing, the whole site would have to do the same.

As a second choice, I set up on-line polls with different design elements and asked people what they liked. I didn't ask about colour - I had decided to use colours based on our family crest. Of course, the snag with asking is that you might get answers you don't like.

And sure enough, I did!

Personally, I don't like boxes with round corners: but - by a two to one majority - users did! I also don't like separate boxes to one side of the screen: but users did, by an even larger majority. There were even ideas about what should go in each box and mostly I've stuck to them, even limiting the newsreel to genealogy in the UK, although to be perfectly frank I am not sure how much I trust Google's news filter.

Like it or not - and I still don't - I had a design plan, but users are a fickle lot, so I'll run a poll to check what you think and keep running it from time to time.

Add it up

The second was trickier - there are many ways to add to a website - but gave me the chance to use some of the experience I have.

First to consider was to use existing web space and add a database. That was a no-go since my provider is cagey about it unless you cough up a chunk of cash. I don't want to pay, so that I can avoid asking users to cover costs.

I was left with fewer free options than I liked, but one was to use a free web host.

Problems with free hosts are legion but I'll leave you to check those for yourselves - there are plenty of articles about it in the computer press. If you're happy with those - and many are, including myself - then fine.

I didn't feel it was right since I wanted control over membership and access. Frankly, it was too much like hard work! That left me one choice: a free wiki service. Part two of the plan was in place.

Keep it up

One of the first lessons I learned was not to ignore advice, so I make no apologies for using two writers for whom I have had a long and deep respect.

The first is Guy Kewney, best known to many because the BBC mistook this tall, thin, wiry, red-headed and (I hope he will not mind me mentioning) greying and balding middle-aged man and interviewed instead a young, large, and very much black person who knew damn all about computing!

The second is a man for whom I have the deepest respect - even though I've met and talked with him and found he is a lot funnier and less serious than his journalism would suggest. In the past, he used to be known as "Wavey Davey" but now seems to be known only as Dave Winder - and as yet I can't find a link to his web page, but will set one up when I do!

Both have been around the block and use different web pages to make sure their work is available. As far as I know, neither has come out publicly and said "for reliability, spread your pages across multiple websites", but it seems to be the best idea and is precisely what I chose to do.

This, then, is the third part of the plan: different services provided by different servers linked through a central page.

Feed it up

The last suggestion was to keep people informed about changes. This was the hard part and the decisive issue in choosing some of the other options. It can be hard to keep you informed and, to be frank, I didn't want the effort of doing it myself . This is, after all, a site to be used rather to take me a long time just to keep people up to date!

RSS, also called Really Simple Syndication (and, yes, I know there are other claims for what it means) was and is the obvious choice. Best of all, it leaves it up to you to choose how much, how often and in how much detail to be kept up to date. It even lets you choose which parts to be kept informed about!

This is a no-brainer. The fourth part of the plan was decided by being blindingly obvious!

Design summary

So the plan was now in place.

It would be split into parts, each dealing with one service.
Page design would be based on the blog.
Changes might be needed because of how other parts looked.

The first three parts would be a blog, a wiki and the central pages on my own server. I would design the blog and copy it to my own page. Then, set up a wiki with a similar design - and change mine to look similar to both. All I had to do now was find the right wiki and get working.

Except that at this point family members asked if I could include a forum where they could meet up and maybe a chat-room where they could talk.... I had not started and already the requirements were growing!