Monthly Archives: November 2008

Just lately I’ve been really interested in finding out Google’s position on semantic technologies and their view on the Semantic Web.

I’d been asked before whether Google were making any efforts in developing semantic technology, but I couldn’t really say. Then I attended the Googleworld debate, at London’s ICA, but couldn’t really find the chance to pose any technical questions.

In an attempt to satisfy my curiosity – and anyway, to investigate something I believe to be of interest that, as far as I can find, hasn’t received any real attention to date – I wrote an open letter, of sorts, forwarded to Google and Semantic Web researchers I’ve found connected to Google, simply asking:

What’s the deal?

Dear Google,

My name is Marc Hibbins, I write a blog I’m sure you wouldn’t have read, I tend to cover new web technologies, online trends, my own development issues, but I’m also extremely interested in the Semantic Web.

Over the past couple of weeks I’ve become increasing interested in finding out Google’s position, or just their ideas even, on the Semantic Web and semantic technologies. I recently wrote about the increase in tech blogs covering the subject over the past couple of weeks, and I’ve been asked a few times – what’s Google up to?

I also recently attended a debate called ‘Googleworld’ – it covered, generally, the past ten years of Google and what’s to come. I wrote about it, and the chair of the meeting replied that he too, is unaware of Google’s position.

Could you shed any light on the topic? Having had a thorough look around online, I’ve found next to nothing. I’m extremely intrigued to find out if Google have any plans with semantic technology – or even if there’s any in place already that just might not be so visible?

If you’ve no plans, do you have any comments? Do you think it’ll even ever happen?

Kindest regards,

Marc Hibbins

I wasn’t sure what kind of response I’d get, if any at all. Or if anybody I did get in touch would be wary of offering any insight that might be misinterpreted as any ‘official’ position.

To my surprise, my first response came from executive ‘Google Fellow’ Jeff Dean. He works in the Systems Infrastructure Group (crawling, indexing and query systems – full bio here), but he couldn’t initially offer any real strong thoughts on the issue. He did say however, that he wasn’t sure if Google even had any real position on the subject at all. If nothing, at least this confirmed that my lack of findings wasn’t down to only an absence of research published externally from Google – or poor investigative work on my part.

My second reply was from Stefan Decker, professor at the National University of Ireland, Galway and director of the Digital Enterprise Research Institute, an internationally recognised institute in Semantic Web and web science research. He co-presented a very interesting Google Tech Talk last year, and worked in Stanford at the same group as Sergey Brin and Larry Page.

He said, very explicitly, that:

In short: The Google management does not believe in meta-data.

Craig Silverstein is on record several times negatively of talking about the topic, as well as Sergey Brin. It is very clear that they are not proactive – a serious mistake from my point of view.

Interesting. I got in touch with his co-speakers, Eyal Oren and Sebastian Kruk. Both said they have contacts at Google still, but neither are aware of any public developments.

Eyal pointed me toward Sindice, a semantic search engine and index as perhaps (though only speculatively – as likely any search engine), might one day receive interest from Google. Perhaps to incorporate their infrastructure for RDF and semantic data consumption. But as he said, there’s absolutely no evidence of it right now.

Sebastian on the other hand described the lack of address specifically as:

[Their] ‘anti-semantic’ approach.

An increasing trend he’s recognised. Suggesting an almost concious movement against any such development. He also expressed his disappointment at the very low turn out at the Tech Talk, that literally only one attendee showed any real interest.

My final response was initially the most exciting – from Ramanathan V. Guha, who leads development of Google Custom Search. He said he’d be happy to comment on what’s going on, although could only offer his own personal opinion and nothing official – but I’ve not received any correspondence from him since.

All in all, at least I know I’ve not overlooked anything major. Fingers crossed I get a response back from Guha, but otherwise I guess I’m left keeping a close eye out for any other developments.

Yesterday, I wrote a ‘how to’ on installing and running multiple versions and concurrent instances of Firefox on Windows XP.

But what about the other browser choices? After all, my original intention was a to develope a versatile testing environment, specifically for cross-browser, cross-platform intended web sites.

Surprisingly, running multiple versions of the other major browsers isn’t as complicated as the Firefox process.

Opera, for example, gives you the option whether to install the set-up as an upgrade or separately, straight out of the box. They offer alternate releases of the current version on their site (9.62 at the time of writing) and have a publicly available archive that goes back to version 3.21 for any old release candidates you need to test.

If you want to run multiple versions of Internet Explorer, you can alter various system and user profile settings in a similar way to my method with Firefox, but it’s far easier to take advantage of the many ‘standalone’ versions you can find online. These are generally third-party, non-Microsoft developments.

TredoSoft have collated standalone versions of Internet Explorer from 3 up to 6, ready to install from a single set-up – it’s called Multiple IE.

It’s brilliant to see IE3, I decided I’d use it as my default browser for a day – loved seeing the frantic alerts about some alien idea called a ‘cookie’ and whether I wanted to risk accepting it onto my computer.

NB: If you’re concerned about what’s being installed when you use Multiple IE, you can do it all yourself with the instructions on Manfred Staudinger’s Multiple IE page.

There’s standalone applications for other browsers too. I only use Windows nowadays, but I’ve recently found Michel Fortin’s standalone versions of Safari – he’s even numbered the icons for your dock (via). That page also links to instructions on running multiple versions of Firefox for Mac.

As for testing Linux system – and this goes beyond HTML and CSS debugging, I use VMware Player from VMware. Not only because when I’ve been developing server-side applications, I’ve not wanted to bother installing those on my home computer base – because it can be tricky, time-consuming, potentially damaging if things go wrong, etc etc and I tend to use Linux-based system for deployment anyway – but because appliances are so damn handy.

Virtual appliances run within a virtual machine like VMware Player as self-contained, packaged software. They can be created and restored as system images, so if something goes wrong – it’s so easy to turn back, with no risk to whatever personal data you might have on your computer as you would installing software as services on the base.

More than that, they’re readily available. VMware has an Appliance Marketplace, with over 900 ready-to-go appliances and a simple, central repository to develop or distribute your own.

There’s popular Linux distributions, various Red Hat, Ubuntu, Fedora – all pretty clean, the basic install, but also some interesting others.

I particularly like the Web Developer appliance, specifically designed to safely test and fine tune web apps. Based on Ubuntu, the creator has consciously included some trendy applications that are gathering more attention, like Ruby on Rails. On top of the expected with Apache, PHP and MySQL, you get a a handful of browsers, various database and debugging tools, code and graphics editors, all as standard, all configured and running – great way to get started.

I’ve recently completed building a new site for the BBC, this time a project pretty much entirely in HTML. As you’d probably expect, the BBC are pretty hot on maintaining a wide foundation of web standards and providing a high level of accessible content, two approaches I’d say I’m a keen practitioner of.

I truly believe that there is zero excuse for slapping ‘this page is best viewed with [browser name] in [screen dimension]‘ on any website, unless it is specifically designed otherwise – and especially if it’s text and image content only – when it’s really so straightforward to adhere to some basic standards which, with almost an exponential effect, can improve the way your content is delivered, cross-browser, cross-platform. That kind of disclaimer is just a cop out, plus Tim agrees - and he’s the man.

A part of that development process is to determine, amongst things, which browsers on what device or platform are the priority. In turn, getting hold of these browsers readily for testing and ideally, available frequently for an agile development. Not leaving testing until a stage where any damage might be irreparable because of time constraints and deadlines, or viewing testing time as an easily squeezable phase, first to suffer when scope is altered.

The BBC have outlined such requirements, they have very in depth guidelines publicly available on their site. The following table defines the ‘levels’ of browser support that all projects must comply with:

BBC Browser Support 

Abbreviated definitions:

Level 1:

  • All content and functionality must work, minimised variations to presentation, fully-styled, maximise user experience.

Level 2:

  • Core content must be readable, usable and navigable, any degradation to must be graceful, no content must be obscured.

Level 3:

  • No support or testing necessary. 

 

Read the full support documentation.

What can be really tricky sometimes though, is getting hold of the all those browsers and platforms to test with. There’s various ways of reconfiguring application and registry settings to install multiple versions of browsers, but the following method is how I installed and concurrently ran multiple instances and versions of Firefox on Windows XP.

Firstly, Firefox 3 (current version, 3.0.4) is readily available from Mozilla, as are previous releases of Firefox 2. Firefox 1.5 though is slightly harder to find. Along with a strong recommendation not use them, you can get other releases from the FTP archive going back to v0.10!

With each set-up, select Custom Installation, giving each version a different installation folder, so something like:

C:\Program Files\Mozilla Firefox 1.5
C:\Program Files\Mozilla Firefox 2.0
C:\Program Files\Mozilla Firefox 3.0

Be sure to uncheck the option to run Firefox when you click Finish, this bypasses writing some default system settings.

Then you’ll need to create a seperate Firefox profile for each version of the browser you’ll be running. Open the profile manager from Start > Run:

“C:\Program Files\Mozilla Firefox 2.0\firefox.exe” -ProfileManager

It doesn’t matter which you choose. Create three new profiles, I named mine ‘Firefox 1.5′, ‘Firefox 2.0′ and ‘Firefox 3.0′ to keep it obvious.

Then I created three .bat files in Notepad, these function as shortcuts to the different versions, as follows:

Firefox 1.5:

set MOZ_NO_REMOTE=1
start “” “C:\Program Files\Mozilla Firefox 1.5\firefox.exe” -P “Firefox 1.5″
set MOZ_NO_REMOTE=0

Firefox 2.0:

set MOZ_NO_REMOTE=1
start “” “C:\Program Files\Mozilla Firefox 2.0\firefox.exe” -P “Firefox 2.0″
set MOZ_NO_REMOTE=0

Firefox 3.0:

set MOZ_NO_REMOTE=1
start “” “C:\Program Files\Mozilla Firefox 1.5\firefox.exe” -P “Firefox 3.0″
set MOZ_NO_REMOTE=0

Obviously change the path and profile names if you don’t use the same as mine, open those up et voila!

Running multiple versions of Firefox

Click to enlarge [via].

Update (18.08.09): This also works with Firefox 3.5, just follow the same steps!

I was lucky enough to be of the 100 attendees for This Happened #6 at London’s BFI Southbank last night, as part of the Onedotzero Adventures in Motion.

Started last year, This Happened is a series of events focusing on the stories behind interaction design. Hosted by Chris O’Shea (Pixelsumo), Joel Gethin Lewis and Andreas Muller (Nanika), the evening showcased four recent works of interactive design, inviting speakers discus and present their creative process from conceptual brief through to installation.

First up was Markus Kison, Berlin based artist and creator of Touched Echo. He installed suspended speakers in to the structure of a railing at Brühl’s Terrace, Dresden, a popular tourist spot and a location most severely effected by the Dresden bombings of World War II.

Touched Echo

A discreet, minimalist installation, Touched Echo invites visitors to assume the position of the victims of the bombings on February 13th 1945, transforming them to performers as they cover their ears, creating a conductive connection from speaker, through railings and their bones to the inner ear, allowing them to hear the sounds of aeroplanes, the falling of bombs and sounds of explosions.

He noticed that although the location received many visitors, the visual landscape lacks a focus. So profound the difference but for the bombings, the gesture and immersive audio stimuli transports the visitor back in time but also acts as a modern, respectable memorial.

Next up was rAndom International, a London-based experimental design collective founded in 2005 by Hannes Koch, Stuart Wood and Flo Ortkrass. Their installation, Audience, recently at the Royal Opera House consisted of 64 head-size mirrors ‘objects’, motorised – and equally characterised – to interact with visitors and passers-by. Subverting the role of the opera goer from that of intending to watch, to be being (quite inescapably) watched themselves by an array of inquisitive and responsive objects.

It was only on for three days at the Opera House, I would have loved to have visited. As they said too, catching up via blog after-the-event can’t compare to the interactive experience:

Troika is a multi-disciplinary art and design practice, exhibiting their Cloud, a five meter long digital sculpture at the British Airways lounges at Heathrow’s Terminal 5.

An interesting organic form but mechanical mass, the Cloud is a 3D shape with a surface of over 4000 ‘flip-dots’, those traditionally used for non-digital signage in train stations and airports. Controlled by bespoke ‘animation’ software able to address each dot individually, the end result is almost as memorising as it is calming, looping on 24-hour evolving cycles accompanied by the rippling sounds of each flip-dot ticking over.

Finally we had UnitedVisualArtists, presenting ‘Constellation’, a light-based sculptural intervention designed for the indoors of Covent Garden Market Halls.

A very grand architectural installation, it is made up of 264 LED-strip lights hung from the cavernous space, cycling sequences of lights and patterns across the ceiling for the Christmas season. There’s also a touch screen interactive surface, allowing visitors to control individual lights or affect sweeping gestures across the structure.

It opens tomorrow night, hopefully I should be able to attend and give it a go.

Adam Neate gave way 1,000 more paintings to the public yesterday, claiming the streets of London as his public gallery. As afternoon dusk fell, hundreds of volunteers distributed the screen printings across the 32 boroughs of London.

Covered by the BBC, Independent and countless art sources online, try as I might scouring streets of Dalston, Hackney and Shoreditch last night, I returned empty handed.

Elms Letters has a great quote about what he’s doing. Printed on cardboard and shrink-wrapped in cellophane, there’s a deliberate attempt to blur the boundary between painting, print and product:

“I remember as a kid going into Woolworths and seeing laminated prints of that famous Tretchikoff painting ‘The Chinese Girl’ and thinking it was great that people could have that iconic image at home for next to nothing. I’m hoping that for some people who come across one of these new paintings, they’ll pick it up not because they recognise it as one of mine, but just because they connect with the image and would like to hang it on their wall.”

Tonight I head down to BFI Southbank for This Happened, a series of events focusing on interaction design, part of the Onedotzero Adventures in Motion.

Report back tomorrow.

When it comes to luck you make your own.