Sunday, February 29, 2004

Spam: the results are in.....
Of the 84 valid responses to the spam quiz, we get some interesting information:

  • 95% of people receive more than 50 emails a week, with the average being over a hundred.
  • More than 40% of these messages are spam
  • They average 11 messages from new contacts a week (suprisingly high, I feel, but there was a little confusion about the question and so some people may have counted spam messages as from 'new' people) - this is just under 10% of all messages received.
  • It's not clear that those with spam filters get less spam: lots of it gets through. It may be that they are more email-active, have had accounts for longer or their address is more widely in the ether, and so get lots of spam attacks anyway, but nevertheless the amounts that they still read are quite high.
  • Most send more than 60 emails a week: on average, more are sent than valid ones received, so this sample is a net source of email, not a net sink.
  • We send to nearly 10 new people a week (also, a number I'm suspicious of as it's very high. Or maybe I don't make email friends as easily as the people in the survey - or I've been on email so long I've met all the people I can already.

This means that most people correspond with people they know already, most are subject to appallingly large amounts of junk mail, and yet still need their addresses to be public as there is a not insubstantial amount of 'new' correspondence created each week.

All this has interesting implications for spam filtering systems - I'm still working on the ideas. Incidentally, does anyone know if a Denial of Service (DOS) attack is illegal, or not?

p.s. at least we do get a blog entry for this quite unusual date, too :-)

Saturday, February 28, 2004

Wireless presentation gateway
I like the look of this - a 'b' device that allows a projector to be shared by wireless-equiped notebooks, allowing different people to grab the screen and share their stuff without all the usual plugging and unplugging nonsense. Now this is good practical HCI. As to whether it will work or not, my only experience of wireless screens is the Toshiba one, which acts as a lightweight, portable touchscreen but communicates with the host PC wirelessley (basically, think tablet PC without a processor or memory but with a wireless link back to a base station that has the necessary stuff - well, maybe don't think of it like that, but you get what I mean). This worked fine - slightly slow to respond to touchscreen inputs, but usable - and would be fine for presentations. I have to get one of these. And a projector, I guess, rather than keep borrowing one.....

And I like their wireless media adapter, and wireless web cam too: watch this space for reviews.....

Problems at UK e-University
The UK's eUniversity appears to be losing out the the UK's individual Universities, and its future is being looked at. But the fascinating bit is, using the figures in the article, the UKEU has over £75,000 per student (for remote students, remember) - which puts the possible £9,000 from fees for conventional universities in the shade, really. Ah, the efficiencies of the internet.

Friday, February 27, 2004

Form design issues and the lottery effect
EPSRC, the government funding agency for science research, has a form that you complete in order to apply for funding. Strangely, whilst it is laid out quite well, it is a pain to enter information into, edit that information or move it around - and it doesn't do any automated summations or summaries either. I simply can't understand why not. Perhaps it's to discourage flippant applications?

The bigger issue for EPSRC is that it gets so many decent applications that it can't even fund a significant proportion of them. The best solution is to increase science funding, but, assuming that isn't on the cards, what else could we do? My view is that we should tackle the lottery effect. This is where people decide that the best approach is to submit many, many proposals, in the hope that a few get through. This puts both an excessive burden on the EPSRC reviewing and panel selection process, and means that people's research becomes dictated by the chance of which proposal gets funded. Better to restrict people to, say, two proposals a year: they would then have to focus on making these as good as possible, in the direction they wished to go, and a far greater proportion of them could also be funded, making it more worthwhile applying in the first place.

Thursday, February 26, 2004

BBC NEWS | Technology | Hackers exploit Windows patches

"It's a myth that hackers find the holes" is the quote that stood out for me. Apparently some people are surprised to discover that hackers do not spend their every waking hour looking for holes in Windows - they take the far easier option of waiting for Microsoft to admit to a problem and then they reverse-engineer the patch to work out how to exploit it (of course, even once a patch is released, there will be thousands of machines that remain unpatched and hence vulnerable). Now, this doesn't surprise me at all - why waste time looking for holes when Microsoft periodically owns up to them. I guess what this story makes me worry about is that increased awareness of this type of hacking could lead to a secretive approach to dealing with vulnerabilities, one that in the end will be counter-productive and lead us back to the bad old days of companies like Microsoft (it's not just them) not even admitting there's a problem.

So what's the answer? Damned if you tell 'em, damned if you don't, it seems. One intriguing solution might be to re-design the software and even hardware to help prevent major hacks in the first place. For example, AMD & Intel now have plans for chips that keeps data and program instructions in two separate parts of memory, thus preventing the "buffer overflow" problem (see link for a better explanation than I could give here). Of course, buffer overflows aren't the only way to gain access to a machine, but it's a start. I wonder what else we could design chips to fend off... spam maybe? now that would be nice.

Wednesday, February 25, 2004

BBC NEWS | Science/Nature | Real pain dulled in virtual worlds

Virtual anaesthesia: a genuinely new use for virtual reality and gaming! Patients with chronic pain are invited to dip into a VR environment to distract them from their discomfort.

There are pointers to some other virtuous VR apps as well, including PTSD and phobia therapy.

Tuesday, February 24, 2004

T-Mobile: postcode hell
Update on the T-mobile saga: the reason Tina failed the credit check is that T-Mobile's postcode database is incorrect. And they have no procedures to override it, so if your details don't match their postcode information, you are assumed to be a charlaton and hence a poor potential customer. Technology seems to stop you every step of the way - but I though it was supposed to make our lives simpler and easier?

She's now passed it. How - because the postcode databases are different, we found letters with different postcodes on, found out which one they wanted, and sent in copies of that one..... So it's really very secure, this approach. But, for now, Tina is a T-Mobile customer. Now, if O2 had had an open customer service department, things may have been different - but I expect they were too concerned with being taken over. - Fax Or Email Your MP For Free

A great public service, given that many of our elected representatives refuse to use one of the most common forms of communication in the country. And I have to say that it does work - from personal experience. I faxed my MP, who did reply, did pass my comments on to anther relevant person, who did also reply. From a usability perspective, it was not quite like email - though I know for a fact that the fax had more weight than any email message would have done.

Russell's guide to surviving conferences or research visits
Some golden rules have come out of our recent decision to send Will (the newest recruit to the Mobilearn project) to a project meeting in Italy. He's not been in the job 10 days yet, but is the only one available to attend it, so has to go on his own. Poor soul - so here is some distilled wisdom.

Key points to remember:

  • It's the social events that are important, not the organised bits in between
  • Talk to whoever you can - they may be important to you in the future
  • Be honest but not too derogatory, as it's a fairly small world in your research area, and someone is bound to know someone who you've just slagged off
  • Take a piece of paper with your hotel address printed on it, so that when you're completely lost or too pissed to think, you can hail a cab and hence get home
  • Use a credit card for everything
  • Collect receipts. Blank ones are especially useful (cos sometimes you can't get a receipt for a different legitimate purpose)
  • You can speak the language if you can order 'two beers' and be understood. Learn the phrase, wherever you happen to be. It tends to get you into and then out of all sorts of situations
  • Take a few notes - you'll need them because the things you remember won't be what others were interested in
  • Take email addresses - you'll ignore the point above and this becomes the only way of getting the information you need
  • Feel free to skip sessions if you can be talking to interesting people instead - at a conference at least, the material will be in the proceedings anyway

Monday, February 23, 2004

The devil is in the detail: 18 pence debt puts me on a BT blacklist (well, sort of)

Ahh, another example of ridiculous interactions with telecoms companies I'm afraid, but maybe with a small lesson about the stubborn determination of complex systems to follow their programming to the letter (bit?)

just called BT to make some changes to my payment details (which I actually requested last month but nobody did anything about it). I was told that I am unable to make any changes at the moment because there is an amount outstanding on my account. how much? 18 pence. yes that's right 18 pennies. why is this outstanding? because a few months ago, with some credits applied to my account (another story there about awful usb adsl modems), my bill came to 18 pence. they never bothered to collect it. I reasoned that it would cost them more than 18p to collect, so they hadn't bothered, and had written it off. I've had this happen before on tiny bills. but no, this amount was still showing on the account, and despite the best efforts of the representative I spoke to, nothing would shift it, and I was told to call back in APRIL once another bill had been produced. in the end I had to offer the solution: a Switch payment of 18 pence.

so 2 problems here: a billing system that doesn't collect small amounts but leaves them on the account, and a management system that doesn't allow such amounts to be written off. she said it was the first time she'd seen it, but maybe she hasn't worked there that long.

Interesting set of HCI links
Check out this for a slightly quirky but interesting set of internet links. Also, some interesting and effective design choices in the layout of the page - it's got a lot of dynamic and database content.

Saturday, February 21, 2004

BBC NEWS | Technology | Video game to help flood planners
Laudable - play games and stop people's homes flooding sometime in the future. The convergence of education and gaming is here. But forgive my cynicism, as I have previously worked with the government's own organisation who developed the models for flooding. We were working on the next generation user interface to the system, and kept getting some strange results. On checing back though the code, we couldn't find that we'd done anythnig wrong, so checked the engine that they'd written, and which was in widespread planning use. Ah - problems with imperial to metric conversion, and a few fudge factors - yes, in places the predicted values were three or more times less than they should have been. We corrected this in the newer versions, but I'm not sure they ever informed their other users. So when I see a quite recent housing estate flooded, and people standing around saying it shouldn't have happened, I am not as suprised as some.

Computers may be magical devices, but if they are programmed incorrectly then they still spew out rubbish...

Friday, February 20, 2004

BBC NEWS | Magazine | Has Text-porn finally made computers 'human'?

I don't want to lower the tone too much here, so I'll let you read the details yourself. In summary, a popular text chat 'bot' that is used by companies to satisfy people's need for smutty texts is apparently so good at its job that some think it might even pass the Turing Test. Or maybe it's just that people keep on failing it...

T-Mobile - an exercise in customer engagement to produce record churn?

Buying a new mobile phone should be a fun experience, especially if you're one of the more valuable contract customers with a decent monthly spend and international usage. So how can phone companies, im theory so keen to get our business, get it so wrong? I relate the sad but true tale of Tina, my partner, as a prime example.

She went into Carphone Warehouse yesterday to upgrade her phone. She had previously been told that she could change her current phone for free, hence the browsing of catalogues and eventual choice. An O2 customer, they went through the discussion and she was then told that it would cost about £30 to upgrade. "But I had been told it was free" said Tina. "Ah", said the assistant, "we've only been told to give it to you free if it looks like you'll be leaving O2". When pointed out that Carphone Warehouse promises to get you the best deal, and yet was trying to sell her the phone when it knew that it could get it for free, the assistant promised to talk to O2.

"Bit of a problem", she reported. "Their Customer Service division that does upgrades is closed to incoming calls, and so is the disconnections department. They're busy". What? A communications company with a CLOSED customer service division that is supposed to keep customers with them? As an O2 shareholder, I'll be writing to them about that.

"Fine" says Tina, "I'll leave O2 then." After a discussion on what networks have what coverage where we live, she settled on T-Mobile. In retrospect, this was an even bigger mistake than trying to leave 02. She gave all the details, and after a bit of hassle when the system couldn't find our house (they have changed the postcode and some systems struggle initially) she was passed through the credit check and duly signed the contract. When she finally got in, late at night after another long day, there was an answerphone message. Excited at having a new phone, she put it on to charge, and then listened to the message.

"T - Mobile here. You've been connected to our network by mistake. An administration error. Please take the phone back to Carphone Warehouse. Sorry. Thank you." No name, no contact number, no explanation. Just an assumption that because they seemed to make a mistake, she should run around after them. And, thankfully, we don't live right next to a Carphone Warehouse, so it will mean a specific trip.

But why should she return the phone? Legally, the contract has odd clauses in it that mean they can cancel it if she gives false information (she didn't) and in a box which she didn't sign it suggests that there may be another credit check - but as that part of the contract is not signed it seems not relevant. But morally, surely they should come and collect it, at least?

According to the number for T-Mobile on the back of the Carphone Warehouse documents, there was a number to call - 7am to 10pm though, so too late for that night.

Today dawned bright and early - Tina brought the phone into the bedroom so I could hear the conversation with T-Mobile.

"Press 1 to ask about your bill, Press 2 to upgrade". No-one to talk to about administration errors. So she tried 1. "But I do billing" said the assistant. "But I want to talk to someone about my connection and immediate disconnection" said Tina. "Oh, well, you can talk to me then" the assistant replied.

Tina explained the problem. "Yes, administrative error, sorry", she was told. "Can't tell you why you have been refused, you'll have to write in." Well, that was useful. Can't talk, can talk, can't get any information.

I decided I'd write about it in my blog, and thought I'd ask T-Mobile for a comment. Their press office doesn't have email, so I used the Customer Service contact form on the website. It crashed. So that doesn't work either. On the failure page, there was an email address for the webmaster, which I send a copy of my message to and a note that their page was failing in it's customer support role.

"Dear Dr Blake" started the response. Strange, as there's no Dr Blake here, and I can't work out how to type Beale and get Blake. The rest quoted the Data Protection Act and suggested Tina phone them. It was an unsigned email.

I replied that she had, and asked for a general response to the overall issued raised by dire customer service. The next response (admittedly, they at least replied quickly) said "Neither would a general statement be satisfactory, since each case is dealt with on its individual merits."

So you can make your mind up about them yourselves, I guess. Customer service seems to insist that you can't get anything useful over the phone (you need to write in) and if you email them they ask you to phone. And all because they messed up in the first place and want us to run around after them. Strange way to treat users, I think.

Sex, drugs and cans of spam
Offers to help you use the porn flooding into your inbox beats even the porn itself, new work shows. Spam accounts for more than half the email traffic. We don't like it but too many people respond to it and so it becomes a viable business model. Even your mobile is not safe. And though fines have been handed out recently, most notably for a phone dialing scam, it's still a very cost-effevtive way of selling your product.

I'll publish the results of my brief spam questionnaire shortly, but rest assured that a few new ideas are being considered to fight it.

SAP Design Guild -- Usability, User Interface Design & Visual Design
Not really known for their user friendly software, this site has actually got some interesting stuff on it.....

Ericsson T100 - some people don't like it...
Usability Watch India :: by Suman kumar - his blog has a usability rant against the phone. Some of the points (phone rings the numbers in its phonebook by itself) seem quite serious to me - could get you into all sorts of trouble.....

Thursday, February 19, 2004

Searching the BlogSphere

The most interesting part in this page is the bit that says
Sometimes I want to know what the world thinks (google)
Sometimes I want to know what I think (my weblog)
Sometimes I want to know what those I respect think (blogs I read)

Very good - I like this a lot. A community approach to better search without the effort.

'The clueless users who refuse to upgrade'
Funny, well-written, but possibly missing the point. If the computer is marketed and sold as an end-user, consumer device, then it needs to be able to be used as such. Fridges and games consoles and telephones and t.v.'s do not crash, whilst home cinema systems are complex and sometimes hard to drive but still rarely fail altogether. Users can be, and often are, not the most clued up on their technology - but they want to use it not understand it.

The suggestion is that Microsoft stop supporting older software so that all becomes new and secure. But if it's unsupported then it'll stay unpatched and so be more vunerable. The better approach is to have a better costing model so that it becomes advantageous to update and upgrade your system, rather than it costing money and time to do so. But updates and patches are not small things - on a dialup connection at home, I can't really afford the time to download 20MB patches. If MS put them on cover disks like they used to then I'd be happier.

Wednesday, February 18, 2004

Spam quiz
I'm doing a survey on spam, as I've a few ideas on how to address it. It's becoming the cholesterol of the internet, clogging up the www and making us feel ashamed of the mess that's our inbox but somewhat helpless to do anything about it! But by filling in this brief (30 second) questionnaire, you'll be helping me address the problem. Thanks.

Sunday, February 15, 2004

Source code leak coverage reveals widespread lack of understanding of computers [Bill Thompson writing for BBC News]
As I hope my posts to this blog will indicate, I am an advocate of making things easier to use and letting the user get on with things without worrying about how the technology underneath it all actually works. But Bill Thompson argues (convincingly) that we also need to keep up with the other side of this equation: educating people about at least the rudiments of how new technology works so that they can a) use it better and b) protect themselves when things go wrong.

The backdrop for these remarks is firstly the recent leak of Windows source code, with the press struggling to explain to the general public exactly what source code is, and secondly the big holes that keep appearing in Windows security, again with the press struggling to explain exactly what the problem is. No complex computer system will ever be 100% bug free, and making it easy to use is just part of the solution - proper education about what computers are, how they do what they do, and what steps are needed to cope with potential problems, has become a forgotten priority.

Friday, February 13, 2004

BBC NEWS | Technology | Microsoft probes secret code leak
Storm in a teacup? Conspiracy theorists, your time has come? Neither: somewhere in between. Microsoft's rivals can benefit? Hardly. Most people who are knowledgeable are aware of how Windows does its stuff - in order to program it effectively you get to see a lot of the inner workings. The source code is littered with profanities. So what? I get frustrated when programming, and whilst I don't put comments in my code that I wouldn't want others to read, I can understand a culture where it doesn't actually matter - and anyway, people were not expected to read them anyway.

Where it came from is a bigger issue, not least because Microsoft releases source code to its operating systems to select Universities and individuals for research purposes, and they tend not have such secure networks as MS itself. So my guess is that it was hacked, or leaked, from a MS partner, not from MS itself - and this mat result in an understandable if unfortunatel backlash from MS against it's University and other partners. They would be wise to consider the longer term effects of this, as these partnerships are mutually beneficial and good at developing MS's longer term relationships with other organisations, and a few hundred meg of old source code is hardly likely to be a significant issue. If it were Longhorn code that had escaped, then they might be worrying more.

People lie more on the phone than in email [New Scientist]

Hmm, this flies in the face of some established literature that suggests that people will lie *less* when you can actually see the other person. the further 'away' you get, goes the theory, the more you will lie - so face-to-face is usually assumed to lead to most truth telling, followed by video conferencing, audio phone, text chat and email.

what's really interesting is the author's assertion that the lying comes about because over a synchronous link (ie phone rather than email) we can be confronted with unexpected questions and feel compelled to answer appropriately. the other issue is one of accountability - emails leave a trace and could be retrieved as later evidence of dissembling, whereas a comment made on the phone could be dismissed as a misunderstanding. so the big question is, as it always has been, do different media lead to different rates of honesty? this is crucial when we think about the use of new technologies like videoconferencing in court cases, psychological assessments, and job interviews. how accountable do people feel when they're in a video call?

Tuesday, February 10, 2004

RSS my Blog! Add to my Wiki! IM my bb!
RSS syndication now makes it easier to collect other people's blogs, or news stories, or whatever, in one place - even this blog now has an XML feed (actually, an Atom feed, but it's essentially the same thing). Feel free to syndicate this, but please drop us an email to let us know you are doing so, as the more readers we have the more we'll write for you!

It's interesting how the internet is expanding the number of tools available for people to communicate, whether it's to comment on things, aid their memory, or chat to friends. Given the popularity (after an initial pause) of SMS in the mobile space, and the growing use of Instant Messaging, I wonder how long it'll be before all the other systems come into wider public use? Bulletin boards are undergoing somewhat of a resurrection owing to neater interfaces, better designs and a 'cooler' look, whilst Wiki's are becoming more common especially for group-owned concepts or corporate information, and as for blogging, it needs no more introduction.

Probably the most underused is the Wiki - a simple tool for generating web pages without knowing HTML, allowing content to be interlinked, outside sites to be referenced, and so on. Sounds familiar? It's like a blog that isn't really talking to anyone..... But that makes a difference to what appears in it - notes from meetings, research materials, web bookmarks, and so on, all contribute to a web of knowledge and information - a personal-wide web, not a world-wide one.

Monday, February 09, 2004

Google Slaps Booble
Booble - you can't say it without smiling. Google don't find them funny and have slapped a cease and desist order on them.

Booble is an adult search engine. Google is a general purpose search engine. Booble says it is a parody, Google says it is a trademark infringement and copying. Amusing to most, annoying to Google, money-making for the lawyers, but it raises some bigger questions about what is copying, what is trademarkable, what is parody and so on. Big ethics raises its head once again - where next? Consider the Mike Rowe case, where Microsoft took objection to a domain name. I quite liked that, and I think the settlement where the kid gets an XBox and other goodies is just great. Mind you, asking for a Playstation would have been cooler.

Thursday, February 05, 2004

Lost: one cat.....
Jake, our apricot Persian cat, went missing. A homely soul, he set off into the snow, and was gone for days. Phone calls, meeting the neighbours, putting the word out in the pub and via the schoolkids in the school, and the whole village was looking for him. 5:30 am trips round the garden looking for him, late nights spent tromping the local fields, and not a single sign. Then, three dyas later, at 6:30 am, there is his trademark strangled 'mow' sound, and he's at the foot of the bed asking for breakfast - and has been around the house ever since.

Why is this in this blog, apart from the fact that we all love a good story with a happy ending? Well, I got to thinking that surely in this hi-tech world of ours, it should be easy to make a small device that combines a GPS receiver, mobile phone chip, and small battery, that can be rung to be activated, and then sends out a location. Small enough to attach to a cat's collar, and quite cheap. If anyone knows of one, please let me know, as I suspect Jake has the wandering urge still residing in there somewhere.....

Update RF ID tags may provide the answer, am checking out both passive and semi-active solutions with a directional antenna reader. Could be a lifesaver app (not a killer one).

This page is powered by Blogger. Isn't yours? (c) 2003-2005 Russell Beale