DougWeb – Doug's home on the Web

Just before leaving work on Thursday night, I sent an email to approx 11,000 people promoting some new publications the charity is selling.

Upon arriving at work Friday morning I discovered 2600 bounce-back emails in my inbox. So I’m not best friends with the records department at the moment – I spent all day Friday going through the database and deleting wrong email addresses. And I still have 2200 to go.

And then when I got home Friday evening, I found 1100 bounce back emails in my personal inbox, because some fuckwad spammer had got hold of my email address and decided to send out adverts for viagra and a variety of diet pills in my name.

If I believed in God, I would think I was being punished…since I don’t, I’ll have to be content with being extremely pissed off.

I’ve gained many a useful tip from blog posts, so it’s only fair that I share the results of an afternoon’s frustration.

If you’re having problems getting the awesome PHP/Java Bridge to work under Tomcat, make sure you’re using version 5.5 of Tomcat and not version 6. I had all of my problems go away instantly when I tried the older version (a final act of desperation). I assume it’s got something to do with the confusing table of servlet specifications outlined on the Tomcat homepage, but that’s just a guess since I try to keep as far away from Java as I can (I dislike strongly typed languages).

Anyway, that’s what worked for me. Hopefully it helps someone else out there.

As practiced by someone who obviously knows a lot more than I do…

Nice idea, but until it stops fucking up fonts with its ‘font smoothing’ technology (is that supposed to be a feature?) I will not be using it. At all. I don’t care if it makes the fonts look like they do on a Mac, they look shite. Part of running on Windows is the ability to take advantage of ClearType (Microsoft’s font-smoothing technology which actually works). Applications don’t even have to do anything – Windows takes care of it. So why the fuck have Apple ported the fucking shite Mac font smoothing over to Windows??!

Also, note to the Apple guys: when on Windows, act like a Windows application and ditch the fucking brushed metal look. It just looks sad and pathetic…

So I’ve just spent 2 days at @media (at @ ??!).

Highlights

• Thursday night’s free bar
• Joe Clark telling us that it’s not our job to compensate for the fact that screen readers are shockingly bad at their jobs
• The London 2012 logo being used whenever a speaker wanted to highlight a bad design
• Mark Boulton’s tips on improving typography
• Jason Santa Maria referring to London’s street layout as a ‘clusterfuck’
• Toupeepal – Dan Cederholm is a genius for many reasons, but that was a particular stroke of genius
• Håkon Wium Lie brought an OLPC ($100 laptop). And let people play with it. It’s so small!
• Jeremy Keith’s Ajax talk – didn’t teach me anything new, but it’s incredibly reassuring to know that thing’s I’ve worked out independently match his ideas.

Lowlights/WTF moments

• Missing out on lots of interesting sessions, because I was sitting in another interesting session taking place at the same time.
• Fucktards asking the world-class, top-of-their-field speakers absolutely stupid questions. Some people were blatantly at the wrong conference
• Apparently Gill Sans is a ‘quintessentially British font’. Eh?
• The Business and Design Centre is laid out really badly
• Getting somewhat lost on my way home Thursday night. My journey was supposed to be ‘come out of pub, turn right, keep walking until you hit King’s Cross’. However the pub was on the corner of a crossroads, and the entrance to the pub was on a different street than the map indicated. So when I exited and turned right, I apparently went south, not west. Instead of retracing my very drunken steps, I decided to ‘keep walking, you’ll hit a tube station / bus stop eventually’. Probably not the wisest move I ever made, but it all worked out in the end (I found King’s Cross!). The blue in the map below shows the direct route, red shows the route I think I took…

Overall it was immensely educational, whilst being extremely fun and I am very, very glad I went.

I’ve finally succumbed to the Facebook craze that seems to be sweeping that nation. In my defence as a geek, I did check it out back when it was all new and shiny, but registration was limited to US college students back then, so I didn’t get any further than the front page.

Now to track down some more people to add to my friends list…

Whilst ‘upgrading’ the water pipes outside my house (iron->plastic), the ever-helpful workmen have sliced through the phone cable outside the house, and it apparently won’t be repaired until Sunday. Not entirely sure how that’s happened, because we’re connected to the the phone network via an overhead cable to a nearby telegraph pole. Not going to put too much thought into working that out though.

In the meantime I’ve connected my mobile to my PC and am using it as an incredibly overpriced modem – at £1/Mb, I’m expecting to end up with a £100 bill next month. The water company will be getting a copy of the bill…

‘Developer Kit’ as he is known (a kit is a young fox), is part of a new marketing campaign by Mozilla – and so far he seems to be a success.

You can find more info about Kit at the Mozilla Developer Center.

I released version 0.4 beta 3 of the DVD Profiler Companion last night – the main changes over the last version being support for the new IMDb page design, and support for Windows Vista.

Users of DVDFolks, are strongly recommended to switch to the Companion, although I have detailed a temporary workaround for the IMDb changes on the DVDFolks page.

In the last 20 minutes I’ve had 4 people email me to say that DVDFolks doesn’t work any more because IMDb have had a redesign…

So if you came looking for a way to contact me about it – please don’t bother.

But thanks for the thought.

The situation

• Take a large voluntary organisation’s online shop
• Add a 20% discount for members of said organisation – provided that they supply valid membership details
• Add a box to the checkout screen that asks for a membership ID
• Add another box to the checkout screen that asks for the name of the entity possessing the membership if it’s not the person placing the order (because memberships are normally in the name of an employer, and not the employee who happens to placing the order).

The Result?

• Approximately 50% of our customers can cope with this
• 50% can’t.
  • Approx 25% manage it after they see a big red error message that explicitly tells them that if the membership isn’t in their own name, that they really ought to be filling in the box marked ‘membership name’ and not leaving it blank.
  • Approx 15% manage it on their 2nd/3rd/4th and in one case 6th attempt.
  • About 5% manage to fill in the name OK, but the membership ID they supply bears absolutely no resemblance to anything correct
  • About 2-3% make a typo that is quickly corrected
  • The rest just give up after a couple of attempts

Why does this worry me?

Several people in the office have looked at the layout/wording on the checkout page, and concluded that there’s nothing wrong with it. However 50% of people trying to purchase something fail miserably at basic reading comphrension. And these are the people that look after young kids whilst their parents are out at work…

Well, WebKit actually. But that doesn’t sound as exciting.

It’s from the guys at GetWebKit, and their browser is called Swift. And it’s actually a rather shitty-looking browser. But since the version they’ve released is 0.1pre, you can’t expect too much really in terms of interface. But it’s WebKit for Windows, so I don’t care. It does threaten to make my Mac Mini redundant though…

via Anne Van Kesteren

200 million people have downloaded the Firefox web browser (my browser of choice). That’s just astonishing. If you haven’t – try it out!

So yesterday Microsoft released an updated version of Internet Explorer 7 beta 2 preview. The key point about this version is that it is ‘layout complete’ – there will be no further changes to the way websites are rendered between now and IE7 final in 6 months time. So far, I’m actually quite impressed with it – especially compared to how beta 1 looked just a few months ago. It’ll certainly make my life easier once it’s finally released, as well as the lives of just about every standards-aware web developer out there. After 5 years of abandonment Microsoft have delivered!

Did you know that snail mail can be faster than ADSL? Indeed it can, although not the snail mail you’re probably thinking of.

In an experiment that can only be characterised as bizarre, Israeli researchers constructed a data transfer system powered by snails. Yes – snails.

From the introduction to SNAP (SNAil-based data transfer Protocol)

Snails are widely assumed to be slow animals. Yet the literature on sluggish speed is surprisingly limited, and only few have actually bothered to measure and record it formally. Further, reported gastropod speeds vary widely with species and circumstance, ranging from 0.000023 to 0.0028 meters per second. With that in mind, it is not surprising that the use of snails as data communications agents was not considered up until the research reported here. Indeed, one can hardly reconcile the phlegmatic disposition of slugs with the fantastic speeds at which information is expected to flow over the Internet. Yet as we show below, the negative attitude towards using snails in communications networks is an example of bounded rationality impeding bold and creative engineering.

The original research

Having downloaded it, installed it and spent an hour playing with it, I now have an opinion that can be summed up as:

The change from IE6 is amazing. Kudos. But since you’ve had 4.5 years since IE6, that shouldn’t be considered as a massive achievement.

The rendering engine seems to be a decent attempt at a large chunk of the CSS2 standard, instead of IE6’s which came across as a CSS1 implementation with a sprinkling of CSS2. Having said that, there are still plenty of bugs that (painfully) show the foundation that the new CSS bits are built on – some date back to IE5.

At the moment I’d say that IE7 is like Netscape 6 – there’s plenty to work with, but when you get bitten by a bug, you get bitten hard.

The best bit for me is that IE7 can actually render dougweb.org pretty much like Firefox/Opera/Safari (i.e. like it’s supposed to look). Sure I’ve had to add some of the IE6 hacks to IE7 as well – but I haven’t had to add anything specifically for IE7, unlike IE6. And IE6 still renders the page horribly.

It’s sad day that I have to praise a browser for simply not sucking at it’s job. But that’s the world we live in…

XML + skin = 780 HTML pages

Perhaps the most fundamental thing to HTML is the fact that an element is created by inserting a tag name, surrounded with < and > e.g. <elementname>. Because of this, any time you want an actual < or > to appear in the text (e.g. 2 < 4), it must be ‘escaped’ as &lt; or &gt;.

Any element content (which may include more HTML) goes after the opening tag, and is then followed by a closing tag </tag> e.g. <b>This text is bold</tag>

A tag may also have parameters passed to it via attributes e.g. <element parameter="value">

Code like that found in a typical DVD Profiler skin
<DP NAME="SOMETHING" ATTRIBUTE="<b>Some HTML embedded within the actual tag as an attribute value....</b>"> breaks all the rules on what constitutes HTML, and is thus a complete nightmare to handle.

How do you tell (algorithmically) where the <DP> element ends in this example? It’s a lot harder than just looking for the first > after <DP…

Even the source colouriser inside DVD Profiler gets confused with stuff like that, and that example came straight out of an IVS-created skin!

<td align="right" WIDTH="0"><DP NAME="LOCK_RELEASEDATE" IFSET="<IMG SRC='$DPIMAGES.Locked.gif'>"></td></TR></TABLE></TD>


I learnt HTML by reading the HTML specification.
I learnt CSS by reading the CSS specification.
I read the XSLT and XPath specifications, and got nothing more than a headache.

HTML is big on the web.
CSS is big on the web.
XSLT and XPath are not. Now I know why.

I’m currently in the early stages of a minor redesign for DougWeb. IE6 is holding me back with a collection of bugs I’ve never seen before, and hope never to see again once I’ve figured out the cause. Normally IE bugs take the form of let’s put this in a weird place. This can be dealt with by telling IE something ‘wrong’ that happens to create the intended effect. But not this time. A partial list of problems I’m encountering include:

1. Content not showing until you scroll past it, and back.
2. Backgrounds showing behind text, but not in the gaps between paragraphs
3. Any attempt to change the styling on a link when hovering over it, currently results in the entire page contents shifting down about 20,000px.

An example of the current IE crazyness is included below. Neither Firefox, Opera, Safari or Konquerer do anything remotely like this. But that’s not exactly a surprise…