Driving from San Diego to Silicon Valley up the 101, we passed an airfield where helicopters were loading up with water and fire retardants. They were helping to control a wildfire in the Santa Cruz mountains, where dry bush had been burning for nearly six days. I knew what they were doing, and why, as I’d just had a crash course in California’s fire problems.

Back in San Diego, at FiRe, I spent some time listening in to a group of CTOs and other tech luminaries trying to come up with an improved IT architecture for fire fighters dealing with wildfires. Inspired by the response to 2007’s disastrous fire season, science fiction author and TV presenter David Brin presented a panel from all sides of the tech industry with a challenge from the local supervisor.

The panel included Larry Smarr, the head of UCSD’s super computuing visualisation lab. He had experience of helping coordinate volunteer imaging specialists during the fires, and of using the university’s IT resources to help disseminate information. The panel was joined by two local subject matter experts - one of whom was a fire chief who’d had to put his own men in the path of the fire to help track the source of the flames.

It turns out that San Diego has a lot of the basic infrastructure needed to build an effective fire detection and warning system - including sensors on mountain tops in risky areas. What’s really needed are a way of increasing sensor coverage at times of maximum risk - and of pinpointing fires directly. Information also needs to be routed to help support decisions that need to be made quickly - and presented in a manner that makes sense. Visualisation tools are important here, as they can bring information from multiple sensors and display it in an easy to understand manner alongside appropriate geographic information.

Two days weren’t enough to solve the problem, but plenty of good ideas made their way into an overall system diagram. FiRe’s brains trust may not have prevented the next round of fires, but some of its ideas will go back to the team at UCSD - as well as to the local fire departments. San Diego may not yet use airships to spot fires, but better image processing and improved sensors could go a long way to saving lives and property.

San Diego’s fires also made it to this week’s Google IO (our San Francisco destination). In a presentation on Google Earth, it tutned out that a local radio station used Google’s tools to create an impromptu early warning system on its web sites. Fire reports were plotted on a map, and used to help predict the likely trajectory of the wildfire.

Imaging and visualisation are critical technologies. We’re visual animals, and a well designed image can compress huge amounts of information into very few lines. Appropriate imaging (if it’s on UCSD’s super computers, or on Google Earth) is a powerful decision support tool - and one that in the face of wildfires most certainly saves lives.

One of the highlights of the Future in Review conference is the chance to go to the supercomputing visualization lab at the University of California in San Diego, CalIT2. It’s run by Larry Smarr, who used to run the National Computing Supercomputing Applications where told one of his graduate students, Marc Andreessen,  to write a visual browser for the World Wide Web Tim Berners-Lee was working on over at CERN. When they showed NCSA Mosaic off, “everybody told us nobody needed it”, he says.

Given how wrong, that turned out to be, it’s worth keeping an eye on what Smarr thinks is important – bandwidth and pixels. Not content with the bandwidth of Internet2, he’s been putting together a multigigabit network connecting universities around the world for sharing data and collaborating over video conferencing. And making video real enough to suspend your disbelief means a lot of pixels; the 60-foot screen in the CalIT2 lecture theatre has four times the resolution of HD, the standard digital cinema will use when the movie theatres work out how to make money from it. To kick off the evening, Smarr invites Microsoft’s Curtis Wong to show off the 12 terabytes of images in the new World Wide Telescope, a map of the sky that zooms from star fields to galaxies to the solar systems coalescing inside them out of dust, fading into infra-red and wavelengths that show more structures.

The 30″ screens on most desks around the lab are dwarfed by the 200 megapixel video wall - eleven rows of five 30″ Dell screens crammed side by side to make one giant display with 100 times the resolution of HD. There are displays that wrap around the edges of a small room, stretching over your head and powered by eight HD projectors, that show us the surface of Mars in 50 million pixels rather than the 2 million pixels from the Word Wide Telescope.
 
It’s not size to prove screens can keep getting bigger; Larry Smarr thinks we need the bigger view. “We’ve artificially limited our brain by this stupid million pixels on a screen and we’ve unblocked that.” So how much more can we see; is there a limit? “Reality! You don’t see everything you think you see - it’s not as simple as pixels. There’s a limit to what you can resolve spatially, above 24 frames per second you don’t really see more. But the brain is capable of absorbing about 1gigabit per second, 24 bits deep 16 million colours. ”

From medical images to satellite maps, there are plenty of images to enjoy at that size. You can see the intricate details inside cancer cells or watch winter spread over the world. You can stand inside a building that exists only as a CAD diagram and walk through lifesize doors to see if the layout works. You can step forward to see the hidden sketch under a Leonardo painting, revealed by infrastructure-red photography and displayed so you can see every line. Or you can watch life-size opera live from the Concertgebouw in Amsterdam, or the opening ceremony of the Nobel prize from Japan and fill like you’re almost there. Every candle flame, every reflection, the brocade patterns on every kimono, the expression on every face.

These are the technologies that are coming to office video conferencing if you have the network bandwidth. Smarr advised HP on developing the Halo system and he’s putting in a Cisco TelePresence room at CalIT2 for academics to use for collaborations. The commodity hardware and open source software that powers the high-resolution screens isn’t as expensive as those. Each screen of what Larry Smarr calls the optiputer - systems connected by optical fibre that make up a worldwide computer system - costs about $2,000. But of course the bandwidth is what really raises the price tag.  Cisco TelePresence needs about 10Gbps; the big screen system is over ten times more.

No battery ever lasts long enough. The extended battery on the HP 2710 tablets we carry give us a full day of work, nine to ten hours or less if we turn on Wi-Fi. I’ve been typing since 8am this morning and online a few times and it’s now 1pm and I have four hours left. That’s just about acceptable, but it’s never enough - I’m wondering where the nearest power socket is. Two technologies we saw at the Future in Review conference this week could produce much longer battery life - if they ever make it to market.

Lithium ion batteries work by packing as much lithium as possible into the positive and negative electrodes inside the battery and them moving ions from them, through the electrolyte fluid and out to your device. The more lithium you can get into the electrode, the more ions you can get out of it. That’s how Yi Cui of Stanford is hoping to get a battery that lasts ten times longer. He’s replacing the usual copper electrodes with silicon, which can store ten times as many lithium ions .

That’s not news; we’ve known for 30 years that silicon stores more lithium, but it also swells up more than copper because of that - and when it swells up, the electrode breaks. Yi Cui’s breakthrough was using silicon nanowires that are much more supple; each wire is only 100 nanometres wide, but they’re very long. Silicon is also more stable than copper, so increasing the energy density doesn’t make it more likely for batteries to explode the way it does with current batteries. It doesn’t make it hotter either, because it’s the internal resistance of the battery that causes the heat, not the capacity.

Ten times as many lithium ions doesn’t mean ten times the battery life; by the time you add in the rest of the battery system, including the electrolytes and the packaging around it all, and some further developments that are still under wraps, you could get double the battery life of lithium ion today.

Startup Seeo is starting with the other half of the battery, replacing the electrolyte fluid with a plastic film that’s very like the polymers used to make motorcycle helmets. For one thing that means it’s much safer - no matter how hot the battery gets it won’t catch fire. But it also works with other battery chemistries than lithium; according to Seeo, some of the lithium replacements they’re looking at could give you 50 to 70 times the energy density of lithium, so you get a choice between smaller devices or longer battery life in the same size we lug around today.

We’ve seen a lot of new battery technologies over the years and few of them have made it to market. One promising zinc battery might finally show up in notebooks PCs this year, maybe, possibly - four years after I first saw it running a laptop. It’s not just that the chemistry might turn out not to work as well as it did in the lab. At the moment you can only charge a silicon lithium battery 100 times before it won’t charge enough to be worth using; that has to go up to 500 times before you’d think about putting it in a mobile phone you’d keep for two years and more like 1,000 for a notebook. Both Seeo and Yi Cui are aiming to charge as quickly as lithium ion, but they’re not there yet - silicon lithium batteries could take an hour to charge.

And hardware manufacturers have to see enough of a demand to change the power supply and charging system in a laptop or phone. Seeo’s lithium battery might fit into an existing device but that’s more about safety than longer battery life; a different chemistry will need a different charger. Silicon lithium batteries run at a slightly different wattage and the value that tells the system the battery is fully charged and doesn’t need more power is also different.

So are these new technologies going to languish the way others have? Maybe not. For one thing, people will pay more for longer battery life, so manufacturers have an incentive to switch. And for another, with the price of oil and petrol still rising, electric cars are looking more likely and both these technologies promise to scale up enough to power cars. When you can do that, a smaller battery for a phone or a PC almost comes for free.

RIM’s WES 2008 event wasn’t just about the shiny devices (though the Bold is very impressive). It’s also about the nuts and bolts behind RIM’s platform, and about how you actually run your own Blackberry service. With plenty of RIM staff on hand, and giving presentations on everything from web design for Blackberry browsers to configuring RIM’s tools to work with the latest versions of Lotus Domino, there’s a lot of material to help you through your working day.

Blackberry Enterprise Server is a hefty piece of software, and RIM’s own network administrators came down to Orlando to show how they run their service. Not surprisingly they have one of the most complex BES infrastructures around, supporting all three mail server platforms (Exchange, Domino, Groupwise) and with three distinct user classes: bleeding edge Alpha users, advanced Beta users, and general everyday Production users.

With a complex environment like that, spread across the world, what’s the secret sauce? It turns out to be their domain database strategy, which is regionalised and segregated (by mail platform and by user class). Even so, RIM aim to have as few databases as possible, with the intention of keeping management simple.

The team gave out some numbers. They currently support 9300 Blackberrys, of several different generations. There are 42 BES servers in the company, supporting all the different combinations of geography, user type, and mail server (of which there are more than 55). All that’s handled with 9 domain databases - running on centralised high availability SQL Server systems with remote connections from RIM’s global BES network.

If you’re worried about your BES performance you can take a tip from RIM, which uses a mix of virtual and physical BES systems. They use 2GB of RAM for Exchange nodes and 4GB of RAM for Lotus (as BES runs on the same server as Domino). Best results come from placing BES local to the mail server it’s using as a message source. The servers are also connected to the local PABX systems, as part of RIM’s new voice service.

So how’s it all managed? Users are supported with a service desk and with self-service. Self-service is increasingly important, and using RIM’s web-desktop manager rather than the standalone desktop tools they can set their own activation passwords and upgrade devices without having to call on the help desk and server administrators.

Server administration is handled by the corporate database team and a team of BES administrators, with the aim of tracking the code people are using on any given day. Systems are monitored using familiar enterprise management tools, including HP Open View and Microsoft System Center. A reporting database handles configuration queries, while custom scripts and the Blackberry Enterprise Resource Kit handle log analysis (though there are always ongoing evaluations of alternate tools).

It’s important for RIM to have good management tools and practices, as it’s using several different versions of the BES code - the current release, the next service pack, and the next generation release. At any time 1000 users are early adopters, already on the next generation Blackberry Enterprise Server.

So what are the five key tips from RIM’s own administrators for a successful BES implementation?

1. Keep your messaging environment healthy
2. Protect the domain databases
3. Ensure adequate server resources have been provisioned
4. Remove orphaned/unused accounts from your mail servers
5. Always document custom configurations (and also save logs)

– Simon

The new BlackBerry Bold sounds like a hot new smartphone: half VGA resolution, 600MHz processor, 1GB of internal memory, 2 megapixel camera, H264 video in hardware, GPS and Wi-Fi, 3G and EDGE, HTML email and a better Web browser ready for Web 2.0. Think YouTube, hi-res movie trailers, photo blogging, Facebook updates, ringtones, online shopping…

Oh, and think SAP CRM, encrypted email, secure Bluetooth for authenticating with a smartcard and all the other enterprise tools you expect from a BlackBerry. The Bold has a QWERTY keyboard and a screen you might point at but can’t touch (wait about six months for that). This isn’t a sleek little Pearl designed for consumers and the pockets in tight jeans; this a full-size BlackBerry aimed at the business. But it has all the consumer features – Bluetooth, 2 mpixel camera, multimedia player and a microSD slot – the absence of which was once the reason business picked BlackBerry in the first place.

RIM isn’t only relying on the security options in BlackBerry Enterprise Server to lock those features down to keep the IT department happy to ensure that BlackBerry stays popular in business. It’s deliberately putting consumer features into business devices to make sure that business users like them enough to keep using them. And businesses might want consumer features in more devices - to encourage employees to buy their own BlackBerrys and bring them to work. If your users are going to bring in a smartphone, RIM could argue, wouldn’t you rather it was one you can control and secure?

If you’re bemoaning the consumerisation of IT because you have to deal with users bringing in social networking applications – the way they brought in IM, Wi-Fi and PDAs – try thinking the RIM way. The consumerisation of IT isn’t going to stop any time soon. Your users have a faster network connection and more advanced communication options at home than they do at work; they probably find it easier to share photos with their extended family than to share planning documents with partners. You can complain about them not thinking about the security implications or not being able to judge which systems are good enough to work in a business setting, but you can’t stop them trying new tools and bringing them to work in an attempt to make their jobs easier.

And if you’re one of the IT professionals who Computer Associates says feels ‘disenfranchised’ because business decision makers don’t think about the IT impact of their business decisions, you definitely need to try thinking the RIM way. IT adds value to the business – but don’t expect the business people to know that (especially if you just told them not to use their favourite social network to talk to their peers in other companies). And expecting them to worry about impact on IT? That’s kind of forgetting who pays the bills and why a business puts in IT in the first place. To get the business to see IT as a value rather than a cost, a source of innovation rather than administration, try figuring out how you can leverage the technologies business users are trying out and give them what they want. And if you just don’t have the budget or the visibility to do that easily, remember that giving them that support will buy you a lot more consideration than burying your head in the sand.  

Those technology-aware business users aren’t going to make your job easier on purpose and they’re not going to use the tools they bring in purely for business purposes, because there aren’t neat dividing lines between work and play any more. Instead of trying to lock down everything, look for how you can partition personal and business use – something RM makes easier than a lot of other vendors.

I find it easy to spot most of the phishing messages that hit my inbox, because there’s nearly always an egregious grammatical mistake in there somewhere. Real messages from banks may be full of logical errors (like a regular savings account with a headline rate of 7% that never tells you that actually it averages out nearer 4% because not all of the money gets to earn the high rate for the whole year), but the spelling is spot on.

And spammers are in such a hurry to put up the Web pages they want to earn ad money on, or use for drive-by downloads to increase the size of the botnet they use to spend most of the spam from zombie machines, that they often make stupid mistakes. If you’re checking 100 messages a day in your junk mail filter for anything real that got in there by mistake, I’m not sure if it’s any comfort to remember that spammers are only human. But Google finds it useful.

According to Matt Cutts of Google at Web 2.0, Web spammers often use templates and tools to build their pages. And fairly often they follow the commented-out instruction to ‘type your hidden text in here’ - but never delete that instruction. The tools they use to fill in forms are simplistic too; the captcha you have to complete to leave a comment here is enough to defeat most of them - but so is a box labelled email address with the instruction not to fill it in. When the bot adds whatever email address it’s abusing, you know you can just delete the comment. Simple maths or the instruction to type in a specific word are beyond bots - at least until Jeff Hawkins perfects Hierarchical Temporal Memory.

If you have a site, you need to think of things that raise the blood pressure of the spammers without doing the same to your users. It’s like being chased by any dumb but dangerous pack animal, says Cutts; you only have to run faster than the slowest person you’re willing to sacrifice. If your system is a little different from the default installation of whatever you use, the default attacks are less likely to work and the spammers may move on to slower prey.

Apart from the obvious advice to patch, patch and patch again, Cutts didn’t say much more - because every time you tell spammers how you’re spotting them, they get a chance to stop doing that. A lot of what Google knows about spam comes from the analysis it does of real Web pages, which lets it work out what things go together. If you know that timepiece and chronometer are synonyms for watch, those strangely-worded Rolex spams are easier to stop. You can see this classification in Google Sets and it’s used in Google Spreadsheets. The equivalent of Excel AutoFill does more than days of the week and months of the year, without you having to add the lists by hand; start with red, yellow and blue and Google Sets will add other colours. Start with lion, tiger, bear and you get other animals.

But you might also get wood, tin and cotton. That’s because Google Sets can’t always tell the difference between the list of animal names and the list of animal toys on the Web sites it looks at. It will learn; like spammers it will learn more quickly if someone tells it what it’s got wrong. But at this point, we get into a race between whether the anti-spam tools can learn faster than the spammers…

So Microsoft walked away from its bid for Yahoo! after raising its offer to $33 a share.

I’m not really surprised at this result. As much as he’s like to think so Yahoo! isn’t worth the $37 a share that Jerry Yang was holding out for, and I really don’t think Microsoft wanted to go hostile considering the damage it would have done to the Yahoo! engineering teams it wanted. Yahoo!’s stated strategy would do much to destroy the company that Microsoft wanted, and that wouldn’t have been what anyone wanted. If the price had risen to more than $33 a share, and Steve Ballmer would have been risking an awful lot of additional gearing that would have ended up diluting Microsoft’s control of its own destiny.

So what’s next? One option is to for Microsoft to take the same approach it did with Borland in the 90s, when it failed to purchase the developer tools company. Instead Microsoft just started to hire the talent it wanted, bringing on board the skills it needed to give it a more mature development platform. If that’s the case here, then recruiters in the Bay Area can probably look for a bumper year as Microsoft starts to cherry pick the talent it wants from Yahoo!’s engineering teams. That’ll be considerably cheaper for Microsoft, though any results will take time to filter through its product pipeline. It took nearly 10 years for .NET to get to where it is today. There’s one problem though, in that Google’s checque book may be a little bigger than Microsoft’s - and unless Microsoft substantially increases the size of its Silicon Valley campus it’s going to be hard to entice developers to move from the balmy South Bay to the damp of the Pacific NorthWet.

The other option is, I think, going to depend on how the Microsoft and Yahoo! stock prices behave over the next quarter or two. The first few days of trading should see a steep drop in Yahoo!’s price, and an equivalent (but not so dramatic) rise in Microsoft. The spectre of a hefty gearing has depressed Microsoft’s stock, and the prospect of a payday has pushed Yahoo!’s up. If Yahoo! continues to trend down, its board is going to come under considerable pressure from institutional shareholders as to why it didn’t take the $33 offer. Yahoo! will end having to approach various suitors, but there won’t be a white knight until Microsoft comes in with a bid at around $28 (or possibly even lower) a share, which the Yahoo! board will be forced to accept.

That certainly seems to be what the market is expecting. Looking at the stock prices a day or so after Microsoft’s Microsoft’s share price rose, but not hugely, and Yahoo!’s hasn’t fallen all the way back to its pre-bid lows. Key shareholders are making a lot of noise about Yahoo!’s boards performance, and it looks as though Jerry Yang is going to be in for a rough ride. Yahoo! is going to have to pull out all the stops to get its new Yahoo! Open strategy announced at the Web 2.0 Expo up and running, and board-room turmoil will be an unnecessary distraction.

It’s still too soon to call it. The story’s not over by a long way - and the dealer’s just started to put down the cards for the next round. There’s a lot of money on the table somewhere for someone, and whatever happens over the next few months Microsoft gets the people and skills it wants for less than it was originally planning to pay, though the second option adds a few additional properties and the trauma of a merger…

–Simon 

Do you speak fluent geek? Or would you rather your computer learned to speak your language? To those of us who’ve done a little programming, a regular expression is pretty clear. But when I’m reminding myself to call an airline or make a payment on my credit card, I think in comfortable phrases with fuzzy edges; Monday morning, any time before the banks close on Thursday. When it comes to doing the accounts, I’m more likely to think ‘last January’ than January 2007. When Windows XP said ‘My Documents’ it sounded like a toddler in a tantrum; when Outlook says Last Week, Last Month or A Long Time Ago, it sounds halfway human. A halfway human teenager, to whom everything more than four weeks old is ancient, but that’s more comfortable than inhuman precision.

Fuzzy human thinking is hard for computers, because sometimes the rules are hard to learn (when do banks close? When does my particular bank close? Is next Thursday tomorrow or in a week’s time if it’s 11.55 on Wednesday?). For other things they’re impossible to learn because we don’t know what they are. How do you tell the difference between a photo of a cat and a cartoon of a dog? You just know which is which, and you know instantly - but you can’t describe how you know well enough to teach the rules to a computer.

Jeff Hawkins, who once founded Palm, has been working on the neuroscience of what humans can do to turn it into something you could teach a computer. The neocortex of the brain, where this recognition happens, unfolds to about the size of a dinner napkin; ‘my dinner napkin is talking, your napkin is listening,’ as he puts it. And in the neo-cortex, recognition patterns are distributed hierarchically and in sequences. We learn spatial patterns and the sequences things happen in, which Hawkins calls hierarchal temporal memory. The brain is predicting what’s likely to happen next and confirming what you’re seeing, hearing and feeling by passing signals up and down the hierarchy. Hawkins’s company Numenta has software for working with hierarchal temporal memory; car manufacturers are using it to try and understand traffic, governments are more interested in identifying who is speaking on a particular phone call.

Even though Hawkins thinks HTMs in silicon can be millions of times faster than the rather slow neurons in your head, it’s going to be a while before computers really understand what we say. In the mean time, there are a few systems that can fake it quite well. Tripit is a travel service that knows the format of the confirmation messages you get from airlines, hotels and car hire companies; forward all the confirmations for your next trip and it will extract the information and combine it into an itinerary, along with weather reports and suggestions for local restaurants and activities. Instead of having to print out a sheaf of papers to carry with you, you can get the details you need in a single email on your phone.

Tripit doesn’t understand everything; the company needs to work out the format for every hotel chain individually and they haven’t started on conference registrations yet. ReQall aims to understand free speech, as long as you use the right keywords – like ‘remind me’ or ‘ask Simon’. You can email, text, IM or phone ReQall and use standard English – “remind me to call Virgin Atlantic at noon on Monday”- and come Monday you get a reminder by email, IM, phone or SMS as you prefer. It’s natural, it works the way people work and it understands at least half of the things you want it to understand.

It’s like asking a friend to remind you of something, but always having them remember to do it. In fact, sign a friend up, and as long as they agree,you can have them sent reminders too. Think of it as outsourced nagging…

-Mary

There’s a proliferation of online document editing services. At Web 2.0 Expo every man (and his dog) seemed to be demonstrating another cloud-based document service. Now Google’s applications have been bundled with Salesforce.com’s online business services. Are desktop applications doomed?

From where I’m standing the answer has to be no.

Microsoft’s Office file formats were often described as how it locked customers into a never-ending cycle of upgrades. New version, new file format. That changed with the arrival of tools that decoded Microsoft’s files and provided compatibility. Open Office and the like had effective file converters that meant you weren’t locked in, and Microsoft’s new XML formats mean you can get at your text no matter what…

Online the story’s different. Create a document in Google Docs and it goes straight into their database. Yes you can save it locally, but that’s not the point of cloud services - you want to be able to get at that file wherever you are. So it stays on Google’s hard drives with no SLA, and no guarantee that you’ll be able to get your files if the service is ever withdrawn - and no idea if you’ll even get notice of a service’s imminent demise. It’s the same for Zoho or for Buzzword or any of a myriad other services. The cloud may be big, but there isn’t enough market out there (especially at $5 a month a user) for all the services to survive. When one goes down, and one will, sooner than later, how will you get your documents?

It’s the ultimate lock in, far worse than anything Office ever did. Unless, of course, you explicitly remember to save documents locally (and it’s not surprising that you have to click an export button to do just that). The convenience of online access to files isn’t enough to give up the freedom to store files where and how you want.

That’s enough to make me stay away. Things get even worse when you actually try to use them for what they’re apparently best at: collaboration.

If two people are trying to work together on a document, it’s good to be able to share edits and to quickly change focus. Try to use Google Docs, and things get a little tricky. Is the current section of the document locked or not? It’s impossible to tell - and if it is, it’s also hard to find out when the lock times out. The locks vary from application to application, and vendor to vendor. It’s not difficult to lock on a per word or per cell basis, and Excel has been offering per cell locking for network collaboration since the mid 1990s. Flickr developer Kellan Elliot-Mcrea put it last week, talking about online tools, “Good model for lightweight collaboration. It’s great up to a point - for most people that point is when you leave your office.”

Then there’s the time warp effect, when you’re taken back to the heady days of Word 6.1 and suddenly all the functions you’ve come to rely on suddenly vanish.

It’s time to make a stand. If you want me to use your online applications then let me have easy access to local copies, give me an effective collaboration platform, and throw in a decade or so’s worth of UI and functional improvements. Until then I’ll stick with Office and I’ll just move my files around with Live Mesh or .Mac or whatever synchronisation tool works best for me.

–Simon