Saturday, February 23, 2008

Oscar predictions 2008

Trying to second-guess who and what the Academy will deem gong-worthy at tomorrow's 80th Annual Academy Awards ceremony is almost the dictionary definition of a fool's game, but hey, I've never shied away from fooldom before (or, for that matter, from making up new words). Below are my predictions for who will be waking up with an Oscar in their pocket on Monday morning.

If you think I'm way off beam and you could do better, please post your predictions before midnight (GMT) on Sunday (either by commenting on this entry or by posting on your own blog and linking back). The highest score gets a bag of Jelly Babies. It's out of fifteen (I've omitted some of the more specialist categories) and the full list of nominations can be found here.

Interestingly, the category I spent longest agonising over was Best Cinematography which boasts five genuinely outstanding nominations, including two from the magisterial Roger Deakins ('No Country For Old Men' and 'The Assassination of Jesse James by the Coward Robert Ford').

Best Picture: No Country For Old Men a
Best Actor: Daniel Day-Lewis (There Will Be Blood) a
Best Actress: Julie Christie (Away From Her) r
Best Supporting Actor: Javier Bardem (No Country For Old Men) a
Best Supporting Actress: Cate Blanchett (I'm Not There) r
Best Director: Joel & Ethan Coen (No Country For Old Men) a
Best Adapted Screenplay: No Country For Old Men a
Best Original Screenplay: Juno a
Best Original Score: Atonement a
Best Song: Falling Slowly (Once) a
Best Foreign Language Film: The Counterfeiters a
Best Animated Feature Film: Ratatouille a
Best Art Direction: There Will Be Blood r
Best Cinematography: No Country For Old Men r
Best Documentary Feature: Sicko r

Update (25th February 7:00): I got 10/15 - the Jelly Babies are mine, all mine... (full winners here)

Related fabric of folly posts:
My Top 25 Films of 2007
40 best songs from film soundtracks
Mild peril: The inadvertent humour of film advisory warnings

Photo: Caleb Sconosciuto. Used under licence

Thursday, February 21, 2008

Video streaming and ISP traffic shaping

The above chart shows the data transfer usage for my broadband connection over the past month. The noteworthy change from the same usage chart for November? Over 2GB of peak-time streaming. The culprit? BBC iPlayer. Why significant? Because my ISP (PlusNet) uses traffic shaping to discourage/penalise peak-time usage, which I've been doing a whole lot of ever since my colleagues in BBC Future Media & Technology added a streaming component to the iPlayer in December. As a result, my "up to 8Mb" connection has been throttled to a painfully slow 125.87 Kbps (according to thinkbroadband's Speed Test) during peak hours, rendering web browsing tortuous and streamed video unwatchable (which is how it will remain until the end of this month's billing period).

PlusNet has a helpful page explaining traffic prioritisation (presumably so-called because it sounds slightly less sinister than traffic shaping), which contains the obligatory layman's metaphor:

"Think of it this way, the broadband network is like a motorway. When the traffic is light, all vehicles can move at the national speed-limit. Some lanes of the motorway have been reserved for important traffic, such as buses or emergency vehicles. During rush hour, most vehicles are forced to slow down. However, the traffic on the reserved lanes can continue to travel at their full speed."

The interesting word here is 'important' - an inherently subjective term (surely streaming video is important to me if that's what I happen to be doing? I don't want to be stuck in a bandwidth traffic jam if I'm trying to watch BBC THREE live or catch-up on the Six Nations). What ISPs are really interested in, unsurprisingly, is limiting bandwidth-intensive activities such as video streaming and P2P downloading which eat into their profit margins.

Until relatively recently, ISPs had a handy justification for traffic shaping: that the vast majority of video streaming and P2P downloading was illegal. This is becoming less true as more and more legitimate streaming and download offerings emerge (the BBC may have taken most of the heat on the bandwidth implications of iPlayer, but ITV, Channel 4, Five and Sky all offer similar services). Add new entrants Joost, Zattoo, Vuze, Babelgum, Jalipo, Veoh, Brightcove and Democracy (all reviewed here) into the mix and you're looking at a burgeoning market for legal downloads and streams.

So, what's an online telly addict to do? One option would be to change ISPs, although as David Meyer points out in a comment on ZDNet, "Any ISP which says it doesn't use traffic shaping at all is lying, unless it simply doesn't have enough subscribers to fill up its pipes". Part of the problem is that in the race to offer cheaper and cheaper (and in some cases free) broadband, profit margins have been squeezed to the point where a high-bandwidth user is no longer an economically viable customer. Unfortunately, that category of high-bandwidth users looks sets to grow exponentially as streaming and P2P downloading become increasingly mainstream.

One possible scenario, suggested in a typically polemical piece on The Register, is a return to metered pricing. Whilst this may feel slightly counter-intuitive, it is consistent with the idea of broadband as utility. I'm happy to pay for my water, gas and electricity on the basis of how much I use - why not my broadband? Personally, I think this is pretty unlikely. Most people were so delighted to see the back of metered dial-up access that it feels implausible that they'd accept a return to a pay-as-you-go model. A more likely scenario is that slightly more expensive, higher-bandwidth packages will increase in popularity for heavy users who (like me) would happily pay a bit more not to have their streams endlessly buffer.

My short-term solution is to return to off-peak downloading using Azureus, with its handy Speed Scheduler plug-in ensuring that it only downloads between the hours of midnight and 4pm (hence no purple in the Peer-to-peer bar). Not my preferred solution, not least because it requires me to decide in advance what I want to watch rather than sampling on a whim (which I've been doing a lot more of since iPlayer introduced streaming). I'm now back to thinking 'do I want to watch this programme enough to download a 600MB file?' to which the answer's often no.

Longer term I think I could well be shopping around for a package with a more generous bandwidth allocation and/or less severe traffic shaping. Any recommendations welcome.

Disclaimer: I work for the BBC. The opinions expressed on this blog are my own and do not represent the views of my employer.

Related fabric of folly posts:
Broadband as utility
Interesting times for the BBC online
Round-up of Internet TV services

Tuesday, February 19, 2008

On turning three, leaving blogspot and becoming a dot-com

fabric of folly is three years old today so, to celebrate, I've decided to take the plunge and move this blog off the much maligned blogspot domain (with it's 77% splogs) and on to my previously dormant .com address. Whilst I'm guessing I may take a hit in terms of Googlejuice, all old links should be automatically redirected and, thanks to FeedBurner (yet another Google property), there shouldn't be any disruption to my RSS feed (let me know if you spot any weirdness).

If you'd like to join in the birthday festivities please feel free to make use of the newly added tip jar in the column to the right (replicated below for the terminally lazy / those using feedreaders). The default donation is 10 pence so don't be shy... ;-)

Leave a tip!

Related fabric of folly posts:
On being misidentified as a splog
Blogger beta & fabric of folly 2.0
Losing my blogging virginity

Photo: Damgaard. Used under licence

Tuesday, February 12, 2008

BBC THREE reborn

Congratulations to assorted colleagues on the relaunch of BBC THREE, not only on-air (where the blobs have made way for user-created junctions - not to everyone's delight), but also online, where the site has been transformed from a rather static, sombre affair (above left) to something more dynamic and befitting the channel's target demographic (above right).

Five particularly cool things about the new site:

- live simulcast from 7pm every day (UK only)
- full integration with the BBC Programmes BETA (= a permanent page for every episode)
- it uses's new visual language (hence the extra width)
- it's got an innovative new Flash schedule, which expands whichever day/programme you click on to give you more information
- it's not an island, with sensible presences on Bebo, Facebook, Flickr and YouTube

Lots of people contributed to the redevelopment of the site but a special shout out to Jo Twist, Al Boley, Siobhan Mulholland, Yuri Kang, Simon Clarke, Oliver Bartlett, Beth Meade, Paul Condon and Venus Speedwell who all worked their socks off to turn it round.

You can read what my boss, Simon Nelson, has to say about the BBC THREE relaunch on the BBC Internet Blog.

Related fabric of folly posts:
Lily Allen take-away widget
Interesting times for the BBC online

Torchwood ARG

Sunday, February 10, 2008

Channel 4's Big Art Mob

More good web stuff coming out of Channel 4 in the form of Big Art Mob - "a collective effort to create the UK’s first comprehensive survey of Public Art" and part of the wider Channel 4 Big Art Project. Users are encouraged to upload photos (or videos), captured using their cameraphones and tag their location so they can be added to a national map. All the usual Web 2.0 staples are there including comments, Google Maps integration and tagging (though no ratings - presumably they decided art is above ratings). With the exception of some rogue HTML tags on the About page, it's a pretty slick implementation - the functionality of the take-away widget (embedded below) is particularly nice.

Related fabric of folly posts:
JPG Magazine & Picture This
Photo Friday
The joy of Creative Commons

Saturday, February 09, 2008

Is it ever ok for websites to start playing audio automatically?

A few years ago there seemed to be general consensus within the usability community that auto-starting audio on webpages was 'a bad thing'. A 2004 survey, quoted on Jakob Nielsen's Alertbox in an article entitled 'Most Hated Advertising Techniques', found that 79% of respondents answered "negatively" or "very negatively" when asked about online ads which automatically play sound. Of course, people get annoyed by most aspects of online advertising, however it wasn't just audio from ads which got people's goat, as this wonderful 2005 thread from an Apple mailing list demonstrates. The thread starts with a query from a web developer about how to add background music to their client's webpage but the discussion soon turns to why he would want to commit such a crime, climaxing with the immortal quote "Background music on a web page makes baby Jesus cry".

Fast-forward a few years and the debate has moved on. Gone are the conversations about the niceties of embedding a MIDI file as background music (thank the Lord). The main game-changer has been the widespread adoption of Flash streaming, precipitated by the growth of broadband and the unprecedented success of a certain video sharing website. Suddenly, every man and his dog is streaming in Flash and the issue of audio auto-start is very much back in play.

Leaving the technology to one side for a moment, the most significant impact of YouTube on the debate is that more and more webpages have A/V as their primary purpose. Auto-starting audio on a page where the media is the main event is potentially very different from a page where it is secondary or, worse, incidental.

That said, many of the same issues remain. A huge amount of internet use is office-based where not everyone has headphones and not all colleagues are likely to appreciate a sudden burst of Chocolate Rain. The growth in broadband has also meant more users will already be listening to audio whilst browsing the web (75% of broadband users have listened to radio whilst browsing according to the Radio Advertising Bureau). Another, slightly less obvious issue, stems from the growth in tabbed browsing (pioneered by Opera and Firefox and then thrust into the mainstream by IE7). Anyone who's opted to restore their tabs from a previous browsing session will most likely also have spent some time trying to track down which tab is responsible for the audio blaring out of their speakers.

I guess it ultimately comes down to user expectations. The web isn't yet at a point where users expect audio to play without their specific say-so. YouTube is a partial exception in that the ubiquity of it's brand promise means that most users clicking on a YouTube URL in an email or elsewhere on the web will know to expect video (with accompanying audio) to start playing automatically. Few, if any, other websites enjoy that expectation. I still feel surprised and annoyed when I land on a MySpace page and audio starts playing automatically. Even the websites of inherently aural brands such as radio and television broadcasters are not yet expected to auto-start audio (yes, ABC, I'm talking about you).

Whether or not audio auto-start will ever become wholly acceptable on the web is a moot point. Web-enabled devices tend to be so inherently multi-function that it seems unlikely that the expectation of self-starting audio will ever take root in a way it has with traditionally single-function devices like radio or television sets. Interestingly, the majority of online advertisers have cottoned on to user expectations in this area, favouring visually arresting but silent videos, with an invitation to users to switch on the sound.

Perhaps user behaviours will shift (I'm already in the habit of hitting the mute button on my MacBook whenever unsolicited audio would be disruptive) or maybe the technology will evolve to automatically detect the appropriateness of audio in any given situation (e.g. computer thinks: it's after 10pm which means the kids are probably asleep so I won't play audio).

I'm personally of the view that audio auto-start remains a no-no on the web. Any potential benefit (e.g. attracting more attention, saving the user a click) is offset by the annoyance it will cause to others, many of whom will be scrabbling for the Back button / volume control and vowing not to return to your impertinent little website.

Interested to hear what other think - please leave a comment or vote in the below poll. Having road-tested PollDaddy a couple of months ago, I thought I'd give Vizu a spin this time...