Category Archives: web

The conciseness of the funny characters

Hacker news brought me the delights of this post on the point that the Chinese language can express more in fewer characters then English. Something I can vouch for which makes translation from English to other languages painful for me.

Anyway, the scary bit, Twitter doesn’t allow 140 Unicode characters !?!

Look, if you say character, you should mean you support any Unicode character! And yes, that may in fact be 560 bytes, but come on, English is not the only language in the world.


Not Facebook

Well I think it’s now become apparent that Facebook has been a wonderfully successful attempt at consensual data mining. Chris, you called it. I’ll say that here and now. It’s also made apparent the decision I’ve mused over.

I’d also like to point out that while the argument is based around the idea of a conversation having two end-points and each having a copy, the idea of the middle having a copy has classically been referred to as: “the Man in the Middle“. Would you like the mail/postman photo-copying all your mail ?

Graceful degradation

For those who don’t know what this means you can go with:

Graceful Degradation- The ability of a system to enter an operational yet degraded state after the occurrence of a sequence of specified multiple faults. 

Although there are a number of other definitions, which all pretty much say the same.

This should be rule #1 for any website design, and the reason this is on my rant list today is that now I use NoScript, the amount of websites that just simply don’t work when their (overly copious) JavaScript is disabled is too high!

When did JavaScript become content ? I don’t buy the “if you turn it off, you don’t get the content” maxim. The Client Side code should enhance the content, it is not the content. The Web 2.0 disaster has some made it “ok” to make JavaScript the main attraction or simply hide the fact there is no content.

If you don’t have JavaScript, your site should still work. However your website uses HTML and you’re going to need to explain things if it works in only one browser…

Now hang on there…

Today I read about the statement from the CEO of Virgin Media saying something which we all have thought that ISPs held to be universally true anyway:

In an interview with the Royal Television Society’s Television magazine, far from covering up their intentions, Virgin Media’s new incoming CEO Neil Berkett – who joined the Virgin Media Board just a few days ago – has launched an attack on the ideas and principles behind net neutrality.

“This net neutrality thing is a load of bollocks,” he said, adding that Virgin is already in the process of doing deals to speed up the traffic of certain media providers.

The more scary thing was that whilst looking through some blog posts on this that I came across Phorm. In summary this allows people to deliver targeted content by using deep packet inspection, which is a little like your broadband router doing SPI, except this takes notes of what your applications are actually saying! Ok, bit simplistic, but the issue is that someone is taking your conversations and using them against you. As someone pointed out, what is there to stop people modifying the packets ?!?

Numbered headings in CSS

“New” thing for today, I’ve been looking at doing outlined numbering for headings, like you do in various formal documents. For example:

1 Heading 1

1.1 Heading 2

1.1.1 Heading 3 Heading 4

Now doing this automatically is difficult, but in CSS, well CSS3, it can be done. Unfortunately not in IE7 :/

Anyway, here is the basis of the code, the full CSS that does the above is under the “Cut” below.

The idea is very simple, for each heading, we maintain a “counter” within CSS to store the number. Before first use of the counter, you must “reset” it, so each enclosing level should “reset” the counter. Also each level needs it’s own counter.

So to kick the CSS off, let us assume we have a numbered page with a body:

body.numbered {
    counter-reset: mynum-h1;

This defines BODY element with a “numbered” class, that does a reset of the H1 counter. Next:

h1.numbered {
    counter-reset: mynum-h2;

This defines a H1 element that does a reset of the H2 counter, and now for something different:

h1.numbered:before {
    content: counter(mynum-h1) ” “;
    counter-increment: mynum-h1;

Now, this defines a “before” style that is used before rendering the rest of the H1 “numbered” element. It has a content: statement that returns the current value of the counter mynum-h1 followed by a space (” “), which is the numbering that was reset by the BODY. The counters start from 1, just in case you were wondering. For the next level H2, we have:

h2.numbered {
    counter-reset: mynum-h3;

h2.numbered:before {
    content: counter(mynum-h1) “.” counter(mynum-h2) ” “;
    counter-increment: mynum-h2;

This pattern repeats for each level of the H elements. The above will do the numbering of the H2 elements, starting from 1 which is reset after every H1. The output is the outer counter mynum-h1 then a period “.” then the counter for H2: mynum-h2. The H2 numbered element will also reset an H3 counter.


Continue reading

The BBC News Redesign ramblings

I guess people might call it a divisive issue or a bit of a philosophical debate about the pros and cons of the new re-design. It appears that some provide a more granular breakdown and analysis, but hang on:

A handful of iPhone users commented that it didn’t work very well on their browser. I’m sure it won’t look great on my Nokia N95 browsing with the images off, but I don’t really expect the BBC News design to worry too much about edge-cases like that. I guess though, if you’ve invested in the most expensive exclusive UK contract phone that money can buy, you’ll lash out and blame anyone for a poor internet experience rather than Apple 😉

Ok, that was a red flag. Edge cases ? It doesn’t work properly in Safari! Yes the iPhone is expensive, and yes, if your favourite website (or application) doesn’t work you’re going to be a bit annoyed, and yes you need a hefty dose of reality if you think that killing off parts of your audience and telling them what they can or cannot buy or how to use their equipment is a good idea.

BBC News stories on the web… unless you have a normal device.

As I’ve written before in this blog, the new BBC Homepage is somewhat annoying when viewed on my iPod. Now with the new News pages, they have somehow managed to cripple the Landscape mode on Safari on the iPod. This what you get when you load the website, then rotate the iPod. The navigation panel on the left is now hidden for some reason, which stays hidden even when you then rotate back to portrait, and notice the space on the right… for some reason the page “shifts” over to the left rather then filling the screen.


(apologies for the terrible focus, the protective cover doesn’t play nice with cameras).

This on top of a number of other rendering errors that mean the page doesn’t resize properly and the text is a bit harder to read. This is due partly to the fact that wider pages need more scaling to be used on a mobile platform. Pity that we couldn’t just use the flow layout for the text…

Brilliant. So not only do I need a new widescreen laptop just to read a story, I can’t use a mobile variation either.