Ben Griffiths’ weblog

As I sit here, my laptop can see 7 wireless networks other than my own - I suspect many of them have, like mine, been secured.

Here’s a question: what would happen if everyone with password-protected wifi in the UK reconfigured their routers so that up to 25% of their bandwidth was available unprotected to anyone who wanted to use it?

I don’t even know if that’s possible with current routers, but most of them surely have hardware that’s capable - and of making sure there was no route to machines on the the secure network from the public side.

Am I naive in thinking that this would create a sensible wifi commons across most major cities?

Update: James has pointed me to Fon who make a router that already does exactly this.

Parcel tracking is a pain. When I buy something online, I’d like the last page in the purchasing process to give me an RSS feed to which I could subscribe to track the delivery - I find reading feeds much less intrusive than email these days.

But I don’t want that feed hanging around in my reader after the parcel has been delivered. Somehow the feed would have to indicate to my reader that no more updates should be expected, that the subscription had expired.

My reader would then archive the feed for me and stop checking it.

One way to do this would be through the HTTP status codes. There is the 410 Gone return code that, according to the spec, implies that “the requested resource is no longer available at the server and no forwarding address is known.”

But this doesn’t seem quite right - it’s an error code for a start. I suggest we need another status in the 2xx range to imply discontinued or frozen.

Having this simple functionality expands the usefulness of RSS - I might want other things to be published in this way. A few examples - live train times that freeze once the train has arrived at its destination; project feeds that are frozen on completion; customer service cases; and so on.

I’m not sure it’s a killer feature, but I think it would help me in a small way - if a couple of the feed readers with large user bases added support for it, I’m sure that many more services would start using it.

Of course, it wouldn’t take long until the marketeers got their hands on it - they’d probably call it a ‘touch point’ - and started piping spam down the channel, destroying it like they destroyed email. But in the meantime, I can think of all kinds of interesting services that could be built around self-destructing feeds.

There’s been talk recently about the BBC’s problems with integrity – I heard one chap say that the current crop of BBC journalists are the best-trained they’ve ever been. But, their science and health coverage continues to be shamefully poor. Lets take today’s headline story – Alcohol link to bowel cancer risk.

Now, if you’d already read the Cancer Research UK you would be forgiven for feeling a sense of déjà vu.

The only piece of information the journalist has added was a quote from a chap at Alcohol Concern, the rest is copied almost directly from the press release. When I was at school, this was called ‘cheating’.

But, worse, the journalist seems to have made no effort to explain what the 10% increased risk actually means. Let me suggest two possible interpretations that any competent, numerate journalist might have considered and clarified on our behalf.

Let’s say my risk is 1 in 20. A 10% increase in risk could mean that that my chances of getting cancer moves to 1 in 18. In other words from 5% chance to 5.6% chance.

This is called the ratio of odds and it’s a common way for the media to misrepresent health statistics since it tends to give a bigger number to a smaller risk – 10% seems a lot scarier than .6% increase.

But, it could be that the 10% increase in risk mean that my chances go from 5% to 5.5%. This way of explaining things uses the ratio of rates.

Which is it? Well, thanks to the lazy BBC journalist, I can’t tell. To be fair, the press release doesn’t say anything about this either – but it’s the journalist’s job to investigate these things.

Also, note that I have no idea what the average and median alcohol consumption is – so can’t have any idea whether I’m in the 1 in 20 risk group, or given my alcohol consumption what my risk is.

Terrible journalism and what’s worse – deeply lazy.

Kevin Marks “In the end, this means that really great coders will keep getting paid less than they are worth and average ones will keep getting paid more, so the economic benefits of great skill will go primarily to the companies with the best employees and not to the employees themselves.”

I read this post a little while ago and was reminded of it today. I’ve found that sometimes good developers don’t get paid the amount they deserve because the management don’t know who they are.

A while ago I visited a company that was jammed to the rafters with below average programmers being paid wages that the management thought should justify a certain skill-level.

I sat with one of the programmers for the afternoon, and before long it became very clear that most of the development in the 20-person team was being done by two of the other developers, and from what I saw their stuff was pretty good. Which is more than can be said for the chap I was sitting with…

That afternoon, I think I prevented a dozen bugs being put into the codebase by this poor chap who was completely out of his depth, but also amazingly arrogant with it. It was something less than a pleasure.

I reported this back to the MD only to find that I’d been sat next to their highest-paid programmer. I was a little bit taken aback. The chap had the most certifications and qualifications and therefore commanded a higher salary. Hmm.

The two developers that wrote most of the good code? One was an older chap who was seen by the management as ‘not having a sense of urgency about him’ and the other had recently been given a written warning about not commenting his code.

Did you know that these guys are doing the same amount of development as the other eighteen together, I asked? The MD was completely shocked. So what should I do? he asked. Well, I said, I’d ask them.

I received an email from the MD today – four years on they’ve shrunk their team to six people, the two great programmers are still there, and as a company they’re much more productive.

I suspect, though, that as Kevin suggests they haven’t shared out the other 14 developers’ salaries…

More than 400 people will die this year because they need organ transplants.

Shame on you if you’re reading this blog and have not registered your intention to donate organs after your death on the NHS Organ Donor Register.

It’s the least you can do, and I can’t see how any ethical person could object to being on this register. Only 14 million people are on the register, even though opinion polls show that more than 80% of people support organ donation.

But there is more that you can do – if you write a blog then you should post about this today, and if you use facebook or twitter or other social network tools you should tell your friends about it there. Go on, do it now.

A couple of people have asked, after my last post about this, if I’m ill. I’m not – or at least I’m in no worse state than normal…

I propose my first theorem: “In any sufficiently close social group, the aggregate amount of facial hair reaches a certain level and thereafter remains constant.”

The colloquial term for the phenomenon is the ‘zz-top effect’.

The theory is easily tested, and no known evidence of its failure has yet been noted.

Observe a group of work colleagues: if a previously-bearded colleague turns up to work as cleanly-shaven as a child’s balloon, the theorem posits that enough of his colleagues would have neglected to shave that same morning as to maintain the constant level of facial hair in the group. Equally, as one colleague’s beard grows, often beyond his or her control, the rest of the team will become progressively more clean-shaven.

I propose no knowlege of the mechanism by which this happens, but I expect string-theory to validate the theory any day now.

In any case, for those struggling to achieve a convincing beard the best advice is to surround yourself with clean-shaven people and persuade a friendly bearded-one to suffer a quick trim.

It took me just 30 seconds to register on the NHS organ donor register. You should do it now. I’m pretty ashamed that I haven’t done this before.

Back in November the EU looked at the economic impact of open-source software. It’s a mammoth report, but it’s also an extremely important one – FLOSS software doesn’t have the lobbying power of the monopolists who peddle proprietary software.

It’s uncompromisingly positive about the benefits of FLOSS, going as far to suggest that they’re so great that FLOSS developers should be able to write off their code contributions as charitable donations.

I wonder how many people know that the UK government has a policy on open-source software. It’s a brief document with little or no commitment to FLOSS. According to the policy, open-source software is considered alongside commercial software with a final decision made on the basis of value-for-money.

I notice though that according to the EU report, some 72% of firms report that open-source software has provided them with “total cost of ownership” savings over commercial software. If you do decide to read through the report, the case study of Beaumont Hospital is enlightening – an enormous saving through the migration to FLOSS. Case closed?

Not at all, and certainly not while government departments are ludicrously suggesting that propriety software allows the Department to benefit from industry standards and trends. Thanks go to Conservative MP David Gauke for asking those questions; the extract is from the Department for Work and Pensions’ answer.

It’s about time that the government revised their 2004 report, to take into account the benefits of FLOSS pointed to in the EU report.

They must also commit to releasing as open source a much higher percentage of the software that they commission. But that’s the topic of a future post…

Chatting with my brother tonight - not a very frequent occurrence - about an idea he has to build a site around young professionals in the media industry.

Twenty minutes later, including finding and buying a domain name, he has a fully functional installation of Drupal running on one of my servers. Another ten minutes later, I’ve talked him through editing his /etc/hosts file so that he can see the site. Add in a couple of extra drupal modules and it’s pretty much ready-to-roll.

How did I cobble a whole site together in 30 minutes? Well, it’s thanks to the incredible automation that my hosting company have built into their control panel. Webfaction rock, no two ways about it.

It actually took me (according the Skype chat history) just 6 minutes to get it up and running - everything from the database to the virtual domain configured with a few clicks in their control panel. And of course, no editing of configuration files - it does it all for you.

Oh, and if you don’t want to use Drupal - and I sympathise - they support pretty much every popular web framework with the same astonishing automation. They don’t charge nearly enough for the service they provide.

The guardian reports today that average web developer’s salaries have risen by a whopping 23% in the last year.

Definitely a sign that we’re in the midst of a tech boom. Interestingly, the article claims that the retail sector is driving the demand.

I hope this signals the end of companies seeing development as something to be out-sourced or off-shored. It would be good to see more companies build core technical capability in-house and stop taking the advice of the snake-oil consultancies that seem to profit most from out-sourcing deals.

UK retail in particular has been very poorly served by the delivery of truly awful website technology from expensive and well-known consultancies.

All this while their board was wondering how Amazon and eBay were eating their online lunch.

As a side note, I wonder how those consultancies had any time to do anything at all. Surely they were busy taking government money to deliver mis-sold and poorly thought out public IT fiascos.

Anyway, good news that in-house development is starting to be recognised, financially at least, as a strategic differentiator after nearly a decade of being undervalued and underpaid.

Anyone looking to recruit right now should be warned though. With so many developers now opting again for contracting over full-time employment, I’m sure salaries haven’t yet reached anything like a peak.