Tuesday, January 31, 2006

This is not a Mobile Me...

Both Engadget and the Unofficial Apple Weblog are running leaked pictures of what purports to be a a pre-release test unit for the much rumoured Apple MVNO. Apparently their anonymous tipster claims that Apple are almost ready to rollout their own cell service under their recently recently trademarked "Mobile Me" name...

The new Mobile Me?

Personally I'm not convinced, and judging by the comments on the two blogs not many other people are either. This looks a lot like an off the shelf Samsung handset, not something I'd expect from Apple. Surely if they really are going the MVNO route they'll ship their own hardware, if only to differentiate themselves from the other networks? I mean, if they don't ship their own hardware what's the point, for Apple, of doing it in the first place?

People have also mentioned that the SIM looks a lot like a cut up iTunes gift card, but to be honest I don't think that's not such a strong argument, after all it just means that its carrying the same branding as another of Apple's products and one which is closely related to the phone. Surely any Apple phone would ship with iTunes? However the argument that the SIM should have an identity number on it is bogus, those numbers are usually printed on the same side of the SIM as the chip contacts, the number wouldn't be visible in this picture.

That said, this is pretty much has to be yet another in a long line of distinguished Apple fakes. I've talked about this before, the hardware margins in the mobile phone industry are vanishingly thin and Apple is used to much larger profits per unit. Move along, nothing to see here...

Update: Apparently nobody was ever claiming that this was Apple hardware, and it was only the SIM card that was supposedly an Apple product. In which case, what's it doing inside that Samsung phone?

Monday, January 30, 2006

802.11n busted?

The Register is reporting that recent trials by Tom's Networking of the latest 802.11n gear were a total disaster.
We found that the product's maximum speed was indeed impressive, topping 100Mbps under the right conditions. But we also saw behavior that does not bode well for use with mixed WLANs... depending on how close the two WLANs are, legacy gear can be essentially knocked off the air.
Depressingly it appears that the interference issues aren't just a hardware or a software issue, although the gear appears to have some problems there as well, but is designed in as part of the original specification. Err, it's not as if there are that many existing wireless networks or anything though, so I'm sure it'll all be fine...

Sunday, January 29, 2006

Get 'em while there is some left..

When Apple announced the new Intel iMac at Macworld 2006 earlier in the month I'd only just bought a 20 inch iMac G5. Well it looks like my still almost shiny new iMac, along with its 17 inch little brother, is about to officially move into end of life status...

If you still want a G5 you can still pick up the 17 inch [USA, UK] and 20 inch [USA, UK] models at Amazon, with a mail-in rebate of $150 if you're buying at Amazon.com. You can't complain, although if you're looking for a bargain you should probably wait for the load of barely scratched second hand models that'll no doubt be flooding the resale market very soon.

Update: It appears that Apple is trying to clear out the last stocks of iMac G5, as TAUW is reporting that there has been a big price drop at the Apple Store. Looks like the UK store has had a similar drop. So as I said, get 'em while they're going if you want one...

Update: The Register is reporting that Apple has officially discontinued the 17 inch iMac G5, although for now at least the 20 inch version is still on sale on the UK Apple Store.


A ubiquitous Earth?

Speculation is rife that Microsoft might be gearing up to try and compete with Google Earth. Indeed Alan Glennon argues that recent press releases and job announcements hint that it's not if, but when...

Alan also talks about the <LookAt> tag in KML, which allows you to tag not only a specific position, but also the direction and view point at the position. Brian Flood takes this up and discusses automating the process, so that any picture he takes with his camera is automatically geotagged with a position an <LookAt> tag.

In the very near future you should be able to go out with your camera, which has an integrated GPS or is connected to your GPS module over a PAN, and take a bunch of pictures or video without really worrying about having to tag them manually. When you got home you'd hit the "upload" button on the camera. The camera would then make a Bluetooth or WiFi connection to your desktop system, upload all your geotagged photos and kick off Google (or Microsoft?) Earth which would display them as placemarks scattered across the globe. You could click on the placemark and zoom down and actually have the view point of where you took the picture an a thumbnail of the picture itself, all automatically. Neat? Well not really, it's almost achievable today, so not really very interesting...

Of course what would be interesting is if your camera tried to do some data mining for you while our were busy trying to take the picture. If you were living in a world of ubiquitous computing your camera could use it's integrated Bluetooth or WiFi to grab data from the local area. Tourist attractions could have information about their site advertised over some sort of zeroconf networking. Your camera could seek the data out and attach this information as meta-data to your photograph. After visiting some ancient temple all the facts and history about the various sights could be automatically attached and then redisplayed when you clicked on your geotagged photograph inside Google Earth. Although the camera would have to be fairly smart, some sort of agent based based system would be necessary. After all, what about all the advertisements and plain old spam that would get pumped out along with the interesting historical facts around a major landmark or tourist attraction?

Of course that still isn't as interesting as it could be, what if your live position and your photographs are uploaded to the net as you take them via a Bluetooth connection to your 3G mobile phone and could be displayed live in Google Earth. Wouldn't that be interesting?

What do you need to make it happen, well some sort of zeroconf networking for Bluetooth, and a ubiquitous deployment of Bluetooth so that creating PANs is done automatically. We also need manufacturers to stop crippling the Bluetooth software they are deploying. Which is mostly sociological, the technology is already here to allow this sort of thing...

The future of computing is in ubiquity and geolocated data, and I'm beginning to think that it could all come down to personal area networks of small devices connected over Bluetooth rather than convergence devices such as my recent purchase the HTC Universal. Your camera, your phone and your PDA will all share your GPS unit over your PAN. My own experiences seem to indicate that so called convergence devices aren't really the way to go, people want decent ergonomics from their widgets. Phones should of necessity be small, and don't need to have embedded WiFi, GPS units, MP3 players or even a camera. They just need to be able to make a connection to a device that does, and of course offer their own 3G data capabilities to the other devices on your PAN.

Despite what the networks seem to think, video telephony isn't the killer application for 3G, the killer application is data. But only if they make cheap enough, and simple enough, so it gets used automatically by the growing collection of devices your average person carries around with them.

Whatever happens I doubt I'm going to get someone to fund my way to Where 2.0 in San Jose in June this year, which I think could become one of the more important conferences on the calendar, in many way more important than OSCON where people are still talking about the desktop as if it were important. Although that said I'd probably I'd probably kill and maim to get to go to OSCON, and unlike the last three years, nobody is going to fund my way there this year either.

Friday, January 27, 2006

Another flight for Global Flyer?

An attempt to make the longest non-stop un-refueled flight in history will see a second around the world flight for the Virgin Atlantic Global Flyer piloted by Steve Fossett. Take-off on this record breaking attempt, from NASA's Kennedy Space Center in Florida, is currently scheduled for between dawn twilight at 6:46 AM and sunrise at 7:11 AM (local time) on the 1st of February. The final decision to go will be taken around 1:00 PM UTC this Sunday...

Take off for the Global Flyer on it's previous flight

New Scientist has more background information on the new record attempt by Steve Fossett and the Global Flyer.

Update: The weather and Chinese New Year intervened and the earliest date for take-off is now the 6th of February...

Update: Take off is currently scheduled for the 7th of February, BBC News has more...

Update: The flight has been delayed again due to a fuel leak, this time with less than an hour to go before the scheduled take off.

Update: After repeated delays, Steve Fossett has now set off on his new around the world record bid...

Wednesday, January 25, 2006

First impression of the HTC Universal

I mentioned at the tail end of last year that I was getting my hands on the new Orange SPV M5000 (aka the HTC Universal) . Due to stock levels, it seems everyone and their dog wanted one, this took a lot longer than I'd thought. I finally had to phone Orange again, and then threaten to leave and go to another network, before they finally shipped me one just this week.

The first thing I noticed on on opening the box, no mini-VGA port. Pre-release rumours had suggested that while the prototypes had one, it had been dropped on later production models. I couldn't find anyone at Orange to confirm or deny this, and since at one point they'd apparently forgot that it had WiFi, I wasn't sure I was going to take their word for it anyway. So while I was sort of expecting it, I was still disappointed by its absence. The lack of VGA display out makes this a device a whole lot less interesting, the thought of being able to do power point presentations directly off a palmtop of this size was very appealing.

Another disappointment is that the mini-USB connection doesn't seem to support host mode, and you can't just plug the handset into a your computer and mount it as mass storage. That's just unforgivable considering that pretty much everything from digital cameras, iPods, the PSP, flash drives costing less than what I paid for lunch, and yes even a lot of mobile phones can do that. It's almost certainly going to turn out to be a software limitation as well, why on Earth haven't they implemented this?

The biggest disappointment though has to be the handset's Bluetooth connectivity, it took me a good half hour to figure out how to connect the handset to my Mac for file transfers, why isn't this ready to go out of the box? Worse yet the device doesn't advertise itself as a mobile phone, instead it declares itself a "device". This would be fine except that I'm still running Mac OS X 10.3.9 on my laptop, which doesn't let you use any Bluetooth devices that aren't classed as phones to connect to the Internet. Even if they do advertise dial-up networking amongst their available profiles....

This problem at least seems to go away Mac OS 10.4. Looks like I've finally got a reason to upgrade...

There are also a few irritating software issues, a good many of them generated by the "customisation" done by the network. However a lot of this can be turned off but going to Start > Settings > Today > Items and deselecting Orange from the list of things to appear on the default home screen. If you don't disable this there isn't any point in selecting anything else from this list as any additional information won't be displayed. Another totally unrelated but equally irritating issue is that the software keyboard popup keeps appearing, even when your typing on the hardware one, at certain types of prompt. You might want to install a null keyboard CAB file. Worked for me...

What doesn't work is synchronisation, both Missing Sync and Pocket Mac don't support Windows Mobile 5.0 yet and neither are saying much about when they will. This isn't great, although again not to unexpected. I am living at the bleeding edge here after all...

However I had presumed I could just transfer my address book and calendars as vcard and ics files onto the palmtop and import them into the local applications. This sort of works for vcards, although there are disappointingly some problems with field entries migrating into the wrong places, for instance all my fax numbers ended up filed under the wrong headings. But for calendars the it doesn't seem to work at all as only the first appointment off any calendar I try and import into the handheld's software is registered, the other several thousand in the calendar file are discarded. This is pretty shoddy programming on Microsoft's part, a couple of for( ) loops wouldn't have hurt too much, surely?

There also isn't much information about developing for these handsets, unlike Nokia with their open source strategy and Python and Perl ports it's a lot harder to find out how to do things under Windows Mobile 5.0. One good resource I have come across is the XDA Developer Forums and Wiki which has a lot of information about the handset.

So do I like it? From reading the above you probably think I don't, but I have to admit having a palmtop with a proper VGA screen and WiFi is going to prove really useful. To be honest however, I might pull my shiny new Orange 3G USIM out of M5000 and put it in my reliable old Nokia 6600, replacing it with a cheap pay as you go USIM and keep my main number in a handset that I can actually carry around easily. Thankfully, unlike some other networks rumour is that Orange's USIMs will work just fine in a 2.5G handset. I use the phone a lot and the M5000, while a nice palmtop, is pretty impractical as a phone...

Update: I've now solved the issues I was having connecting to the Universal as a Bluetooh Modem from my Powerbook.

Monday, January 23, 2006

Another five minutes?

It looks like Michael Arrington's criticism of Ning has pushed them into addressing at least some of his points. In what appears to be a response to the criticisms levelled at the start-up they've posted a summary of what's been happening and news of some new features coming soon on their blog.

Diego Doval has now made a more measured response than his initial reply, and people are jumping to the defence of Ning. On the other hand some people are pointing out how "lost in space" Ning has become. Back when I originally grabbed a beta account I added a couple of Ning-related blogs to my feeds list. But despite that I don't remember seeing anything other than puff pieces about the site since then, at least until Michael's post. If a Web 2.0 company can't be bothered to write about itself on it's own blog, it's evidence of something, although I'm not exactly sure what...

What Ning is trying to accomplish is laudable, but bubbles are about mindshare, and they've made a poor first impression on the great and the good. It's going to be hard for them to recover from that, perhaps their competition will learn from their mistakes? You don't just have to have the right service at the right time, it has to be pushed at the right people as well, and that also has to happen at the right time.

Your start-up has five minutes to sell itself to me, don't waste it, you probably won't get another five minutes...

Sunday, January 22, 2006

Mobile as server...

I've had a web server running on my phone for a couple of years now, and I'm not the only one, so I was fairly confused to see people making a fuss about having Apache available on your phone. Or rather I wasn't confused, after all it is a big deal, because now we have a full scale stable enterprise level server available on the Series 60 platform. But the publicity surrounding this seems to be focusing on the fact that it is possible at all...

This frankly surprises me a lot, I've been waiting for someone to come out with something vaguely like Rendezvous... sorry, I mean Bonjour, err, I mean Zero-configuration Networking... over your Bluetooth personal area network for Series 60, but it hasn't happened...

Maybe this is why? I guess people just haven't caught on that they have a computer in their jacket pocket. I mean most phones these days have heavier duty processors and more memory than the machine I did most of my PhD work on, there really isn't anything stopping people doing serious work on them now. Except the fact people think about phones as client only devices, and especially with the availability of both Python and Perl ports for Series 60, this hasn't been the case for a while...

After all, if you can have a web server for your PSP, what's the big deal about your phone?

Friday, January 20, 2006

Back in transit

I'm sitting in the departure lounge of Edinburgh Airport logged onto the BT OpenZone, rather than the T-Mobile, wireless hotspot. I actually have a T-Mobile account and would have used their service, but the hotspot seems to be totally shot at the moment. It's even refusing to give me an IP address, let alone allow me to talk to the proxy so that I can get out onto the Internet, hence the fallback to BT.

Of course it's nice to have a fallback, it's not often you see competition between wireless providers. Normally providers lock venues into contracts where they insist on being the sole service provider. Although thinking about it, I'm not actually entirely sure I'm supposed to be on the BT OpenZone network, as I've a funny feeling that I'm in the spill over from the BA Executive Club lounge.

I've got a late flight, but I should be home before the plane is turned back into a pumpkin, so I guess that's alright...

The AstroGrid Workshop, Final Day

The fourth and final day of the workshop is being run as a question and answer session. Now we know more about the software we get to pin the developers against the wall and ask them why it works the way it does; why that button goes there, why do you have to double click on that thing, why can't we just click that, and what's with all the XML? Towards the close of the meeting we were also standing up and talking about how we're thinking about using all this stuff, so I plugged the laptop into the projector and trotted out my half hour's worth of tinkering with Perl and the ACR, no doubt to the general amusement of rapidly thinning audience.

It's been interesting to see how the AstroGrid code base has evolved since last time I looked at it seriously, and it was especially interesting to see the state of the ACR. If the VO is to succeed I think something like ACR is very necessary as it lets the community access these new resources in the way they've always done things, in their favourite scripting language.

Related: Day 1, day 2 and day 3 of the workshop.


I never did tell you how I got on with Ning. I guess that's because I didn't find it particularly interesting I just couldn't figure out the advantages of using Ning's proprietary stuff over just sitting down and writing your mashup by hand.

It looked fine if you just wanted to clone one of the existing applications, and make some minor changes, but not so fine if you wanted to write something unique. You also lost control of the hosting, user registration and lots of other stuff. In the end the disadvantages seemed to be much larger than the advantages for those people who wanted to do something novel. But I figured it was just me, that I didn't "get it", so I kept quiet...

Well it looks like it wasn't just me, Michael Arrington has just posted a damning inditement of the service on TechCrunch. Michael starts off by saying that he considers Ning to have been the perfect service at the perfect time, but from his post it looks like he's a disappointed man. He talks about the four main problems he sees with Ning, but to me one of these is overwhelmingly more important than the rest,
You have to know PHP, or at least HTML, to build anything unique on Ning. They promise to create tools to allow non-programmers to build stuff in the future, but for now, 99.9% of the Internet population is effectively locked out from creating new stuff. - Michael Arrington
If you have to know PHP or HTML anyway why would you use Ning unless they made your life significantly easier? Why share the fame and the glory of having the latest and greatest mashup with the Ning crowd instead of just writing it yourself?

The prime market for the Ning service sort of had to be the people who didn't know PHP or HTML, but had a good idea they wanted to rollout anyway. In the end the service didn't cater for those guys out of the box and that's probably why they're going to be one of the first casualties in the Web 2.0 bubble.

Update: Diego Doval replies to Michael's criticisms, and there are some some general murmurings that maybe 3 months is too short a time scale to judge the success of something like Ning over...

Update: Is any publicity, good publicity? Will Michael's criticisms breath new life into Ning? Are they worth another five minutes?

Thursday, January 19, 2006

The AstroGrid Workshop, Day 3

Today we've been looking at how to publish data into the VO and how to wrap command line applications so they can be used by workflow and other VO aware applications.

Update: However I've also been continuing to fiddle with the ACR to try and accomplish something vaguely useful. I've now got some perl code which will grab a VOEvent RSS2.0 feed look through it for new events, and upload these to the user's VOSpace. Not exactly Earth shattering, but it's a start...

Update: It also turns out that just about everyone is currently generating VOEvent messages with bogus id's. Looks like we're going to have to fix that one...

Update: Day 1, day 2 and day 4 of the workshop.

Wednesday, January 18, 2006

The AstroGrid Workshop, Day 2

Today we're looking at AstroGrid's user interface into the VO, the Workbench, along with their distributed storage system VOStore...

The Workbench GUI

Update: I've uploaded an quickly hacked together example of how to access the ACR RPC services via Perl. Not having played with any of the three main XMLRPC implementations for Perl before, being mostly a SOAP person, I was wondering whether any of them implement syntactic sugar for the RPC calls? The following,
   my $rpc = new XMLRPC::Lite();
   $result = $rpc->call('astrogrid.community.isLoggedIn');
while perfectly serviceable isn't really very pretty. I'd obviously prefer to do this,
   my $rpc = new XMLRPC::Lite();
   $result = $rpc->astrogrid.community.isLoggedIn( );
and at least with SOAP::Lite I can. But it doesn't seem to work with XMLRPC::Lite. I'm guessing Perl really doesn't like the dots in the method name. A brief, but not exhaustive, look at the XMLRPC::Lite source seems to suggest that this is the case?

Update: Day 1, day 3 and day 4 of the workshop.

Tuesday, January 17, 2006

Microformat wars?

It's pretty easy to to create a microformat, in fact it's so easy it's hard to stop users doing it. Some will catch on and some probably won't. Both geotagging, and more recently celltagging, have become standards from the ground up. People are following them because they think it's a good idea, not necessarily because some standards agency has given it's blessing.
I think there's a lot to be said for following standards or microformats. However there's nothing stopping people from tagging however they like, and if enough people do it becomes its own useful standard. - Dan Catt
The key word here is "useful", people want software and formats to be useful, no useful added semantic content, no microformat.

People are starting to get worked up about the supposed wisdom of crowds and the possible problems created by Folksonomy and I don't really understand why, surely we should let the market decide?

In the end people use software and standards because they're useful, they solve problems. If standards are neither useful, nor solve problems people care about, then they won't be used. Conversely, no matter how unofficial, standards that solve problems get used. Standards that almost solve problems get twisted until they do, even if that "breaks" the standard. That's what's happened with triple tagging and the solution isn't to stop users doing it, it's to fix the infrastructure so that when they do we can use the extra semantic content in a useful manner.

More Web 3.0

We're no longer on the way to Web 3.0, some people are so sick of the hype surrounding the second bubble that they've decided to skip directly to version 3.0. Amongst other interesting things, Jeffery Zeldman has this to say to those of you out there building social software,
To you who are toiling over an AJAX- and Ruby-powered social software product, good luck, God bless, and have fun. Remember that 20 other people are working on the same idea. So keep it simple, and ship it before they do, and maintain your sense of humor whether you get rich or go broke. - Jeffery Zeldman, A List Apart
I have to admit I agree with him, you only have to look at the sites offering Flickr like functionality for video to know that there are only so many ideas out there, and that everybody is trying to get rich off those ideas. Not everybody is going to succeed, and it's not necessarily going to be the best technically implemented site full of AJAX-powered goodness that's going to win either. If Flickr offered video feeds tomorrow, the ten or so startups trying to make money off this would immediately go under. Bubbles are all about mindshare, and while I'm firmly behind the Web 2.0 ideals, lets admit, at least to ourselves if not the people offering us vast sums of money, that we're stuck with another bubble and just get on with cranking out some good code...

The AstroGrid Workshop

The first day of the AstroGrid Technical Workshop kicked off with Andy Lawrence giving a broad overview of the Virtual Observatory.

Andy Lawrence talking about the VO

He then handed over to Nick Walton who talked about the aim of the workshop, which is to show us how to deploy the AstoGrid infrastructure and for us to gain hands on experience with the tools they've built for the community.

Nick Walton talking about AstroGrid

Apparently we're going to be split into groups which will have to decide on a topic or a problem which they'd like to work on during the meeting, and then we're actually going to cut some code. Should prove to be interesting...

Update: Nick then handed over to Guy Rixon who gave us an orientation to the various AstroGrid components and dumped the buzz word dictionary we'll need for the rest of the workshop. We've now been split into groups and parcelled out with various tutors to install the infrastructure onto our own laptops...

Installing the infrastructure...

Update: Day 2, day 3 and day 4 of the workshop.

Where is Perl 6?

Chromatic has written up a good summary of the current state of Perl 6 development for the O'Reilly Perl.com site. He talks at length about the advantages of Perl 6 and addresses the question everyone is of course asking. When's it going to be ready? The answers of course is, as always, it'll be ready when it's ready...

I actually didn't spot this article till it hit Slashdot, and looking through the comments it generated I was actually fairly surprised by the amount of anti-Perl feeling out there. After all, everyone I know loves the language, and for the record a list-of-lists isn't hard...
my @list = ( [1,2], [3,4], [5,6], [7,8] );
I don't really think the syntax is so hard, so what's the problem here guys? In fact if you understand the underlying data model you're dealing with, the syntax is obvious. You wouldn't try to code in a language where you don't understand how the data is stored, right?


Monday, January 16, 2006

In Transit...

I'm back in transit yet again, on my way to the AstroGrid Deployment Workshop being held at the NeSC this week. I'm booked onto the commuter flight to Edinburgh out of Exeter this evening, no doubt another in a long line of Dash-8's, and with luck I should be holed up in my hotel some time around eight o'clock. At least I don't have a talk still to write...

Update: Another day, another Dash-8...

BE215 from EXT to EDI

Sunday, January 15, 2006

Disassembling the iMac

Some people just can't seem to help themselves when they get their hands on new hardware, they just have to take it apart (via TUAW)...


But I must admit to being somewhat confused, compared to the previous generation of G5 iMacs the new machines seem to be disappointingly messy. Apple normally put a lot of thought into both the outside and inside of their machines, and this looks like it's just been thrown together. You have to hope that this is a pre-production sample rather than a production model, because to be honest, it just doesn't look like Apple kit.

Safe landing for Stardust

At 05:10 EST (10:10 GMT) this morning the Stardust sample return capsule successfully made a soft landing in the Utah desert. NASA TV has podcast coverage of the landing...

At 5:10 a.m. Eastern time, Stardust's return capsule landed in the Utah Test and Training Range. Top: An infrared camera captures the Stardust sample return capsule as it descends to the Utah Test and Training Range. Bottom: The NASA TV image above shows an infrared view of a helicopter on the ground at the capsule landing site. The capsule contains cometary and interstellar samples gathered by the mission.

Update: The sample return capsule has now been recovered and moved to a temporary clean room at the Michael Army Air Field,

Top: The Stardust sample return capsule in the Utah desert. Middle: The capsule being lifted from the landing site. Bottom: Stardust sample return capsule being wheeled into a temporary cleanroom at the Michael Army Air Field in Utah.

Update: More from the New Scientist...

Saturday, January 14, 2006

Stardust on final approach

In just over 18 hours from now the Stardust probe will hopefully make a soft landing in the Utah desert at 03:12 local time (10:12 GMT) after a 2.88 billion mile round trip. The velocity of the sample return capsule, as it enters the Earth's atmosphere at 28,860 mph, will be the fastest of any human-made object on record.

The reentry of the sample return capsule should be visible over most of the western part of the United States, and should provide a spectacular show, with a predicted peak visible brightness of around -7 magnitudes seen from a distance of 100 km (60 miles), which is two and a half times brighter than a Iridium flare.

Update: A safe landing for Stardust...

MS Windows on Intel Macs

While, at least until Windows Vista arrives, you won't be able to dual boot Windows on the new Intel Mac hardware, it now appears that you will still be able to run Windows. Yesterday Microsoft committed itself to a release of Virtual PC for the new Apple Intel platform.

Considering an entire layer of emulation can be removed from the product, the new Virtual PC should run Windows software at almost native speed. The downside is that Microsoft did not announce when it would ship the new version, and while you're waiting for it, you can't run the current release under Rosetta.

Although thinking about it, that's probably going to be for the best. If you could run Virtual PC in Rosetta it would be a huge resource hog. Native x86 code would be running in a virtual machine emulating a PowerPC, on-top of another layer translating the emulated PowerPC instructions back to native x86 code. Anyone want to take bets on how slowly that would all run?

Update: More from Apple Insider...

Friday, January 13, 2006

A day of Google searches...

An paper by Rob Pike, Sean Dorward, Robert Griesemer and Sean Quinlan of Google is starting to make the rounds on the net (via Inside Google and Information Aesthetics). The paper is all about how to analyse large distributed datasets, which might not interesting enough in itself to get people talking about it, however it comes with some interesting eye candy...

...an animated image of the distribution of Google search queries over time, during the course of a day. Cool, huh?

CREDIT: Google
An animated image, generated by Google Labs, that shows how search queries to Google were distributed around to world and over time on the 14th of August 2003

Not the iPhone, but the iCell?

In yet another round of post-Macworld rumours, this time surrounding the recent trademark filing by Apple, the possibility that Apple might be contemplating entering the mobile market has yet again become fodder for the Apple rumour mill.

Speculation about possible plans to release an Apple branded mobile phone have been doing the rounds for several years, with rumours of the Apple iPhone stretching back to the start of 2004. Although you might of expected the rumours would be put to bed after the release of the Motorola ROKR, instead the release of such an obviously poorly designed device just whetted people's appetite for the real thing.

This time we can all blame Ars Technica for starting things, and Slashdot for spreading it far and wide. Of course the reason that rumours of the iPhone continue to appear is that it is such a reasonable and likely thing for Apple to do, despite the fact that the mobile phone business traditionally has such a low margins. After all the PC business has low margins and Apple is seemingly able to get away with charging higher prices there...

What's the difference? It's only Apple's history in the computer market that lets charge those higher prices, the mobile industry is a very different, with consumers used to low cost or even free handsets. Where would Apple make it's money?

I'd rather like an iPhone, but I don't think it's going to happen...

Update: It looks like I'm not alone in thinking the latest iPhone rumours are a based on a collection of rather thin suppositions. On balance I think that the other rumours that are spreading about the new range of Intel Macs are much more likely to be true...

Thursday, January 12, 2006

One more thing?

Is there more to come? Only two days after the Macworld keynote rumours are already spreading that there may still be "One more thing..." to come and key products were missing from the Steve's keynote speech due to supply issues with the new Intel Core Duo chips...
..."They can't get enough Core Duo (chips)," said my source. He also said that if he were me, he probably wouldn't order one of the new MacBook Pros. I asked if there would be MacBook replacements for the 17-inch and 12-inch PowerBooks, but he said, "Oh, it's much cooler than that. Much cooler." - Leander Kahney
Update: I'm starting to wonder whether it was supply issues with Core Duo chips that was the problem, and whether the predictions of instant-on Intel Macs are the "much cooler" thing. The recent announcement that Samsung will start producing 16 gigabit (that's 2GB) Nand Flash chips this year, coupled with Apple's supply agreement which everyone thought at the time was to handle the growing demand for the iPod, might mean that at least the replacement for the 12" Powerbook might be shipped without a hard drive. It'd certainly fit with my speculations about the future of the Powerbook line...

Wednesday, January 11, 2006

The MacBook Pro

The first up close shots of the new Apple hardware are starting to trickle out of Macworld in San Francisco. The images below are curtsey of UNEASYsilence who are doing a really great job of reporting the odd, funny and cool corners of the show that other bloggers aren't reaching.


Thanks guys...

Update: Tristan Louis has posted a comparison chart (via Apple Matters) comparing the MacBook Pro with the PowerBook and comparable Windows laptops. It makes interesting reading...

Update: It looks like all might not be lost for those of us who desperately wanted a updated 12" Powerbook, as the rumour is that a 12" and 17" version of the new MacBook may be coming real soon. Straight from someone, who overheard someone, who once talked to Steve Jobs. So it's just got to be true...

Update: Looks like Apple's use of EFI means that despite the move to Intel current releases of Windows won't run on Apple hardware. Presumably these problems will go away when Windows Vista finally ships.

Update: It also looks like there are some concerns about the specification of the new MacBook. You do have to wonder about the IR port, why doesn't the Apple Remote use Bluetooth? and since this is a laptop, can you use the remote to drive a PowerPoint or Keynote presentations? and if not, why the heck not?

Update: But there are also people who seem to be raving about the new MacBook. I guess the only way to be sure is to go down to your nearest Apple Store and try one...

Update: Or alternatively you could watch the video that Kevin Lim has helpfully made from the show floor at Macworld of a MacBook Pro in action.
Kevin Lim (via YouTube)
Fifteen minutes with MacBook Pro and iLife '06

Now for the good news...

With a lot of non-Intel Macs kicking around, and with people already talking about buying another cluster of the new Intel iMacs, computing at Exeter is going to be in transition for the next couple of years.

I've gone through several of these over the years, including the shift from VAX/VMS to a mix of DEC Alpha and Sun workstations, and then from those to clusters of Linux beige boxes, and most recently from Linux to Macs clusters. However Apple's move to Intel architectures won't really effect us as much as it might have, as we still have Linux boxes around and were therefore already operating in a mixed environment.

However since I'm at least partially responsible for the system administration of our current Mac cluster, and will probably be responsbile for any new Macs, I'm going to be following the PPC to Intel transition very closely.

The good news is that it looks like the essential admin tools are going to be able to cope with mixed architectures environments out of the box. More good news is that target-disk mode should still work with the new Intel based Macs. The best news is that at least some of this goodness should be available already in the 10.4.4 update. You can't complain really...

Update: ...well, not much. Rui Carmo has taken a look at 10.4.4 and it seems that the Bluetooth Updater may have some problems.

Update: It also looks like we have some incompatibilities between the partition tables on Intel and non-Intel Macs. Nothing is ever easy, is it?

Update: ...and Rui is still having problems with 10.4.4.

An iMac?

I can sympathise with Jason Kottke and Scott McNulty, I've also just bought one of the previous generation of iMacs...
I about fell out of my chair yesterday when Steve announced the new iMacs, since I just bought one of the last PowerPC models. 'But the iMac was recently refreshed', I moaned, 'you were supposed to introduce a new Mac mini!' - Scott McNulty
..because like everyone else the last thing I was expecting was a new iMac, they'd only just been updated in October. I turned slightly green when I heard, and I still feel a little queasy. I'm still really happy with my new machine, just a little bit less happy than before.

Tuesday, January 10, 2006

Google Earth for Mac OS X

Closely following the leaked copy of Google Earth into the wild, Google has just officially released a version of Google Earth for Macintosh today at Macworld. While this is great news, however on the down side it appears the official release is still Mac OS X 10.4 only so those of us at the trailing edge still have to wait.

Update: First impressions and usage notes from Ogle Earth.

Update: Flickr users out there probably also want to look at Flickrmap and their tutorial on how to use Google Earth to geotag your Flickr images.

Update: It looks like JPEG images aren't a supported format on the Mac version of Google Earth for image overlays, and while I haven't played with the Windows version at all the KML tutorial seems to suggest they should be...

JPEG images aren't supported?

Update: Okay, oddly enough it looks like it sometimes will load JPEG images, and sometimes not. I'm no longer sure whether this is a transport or format issue, although there doesn't seem to be any difficulty with other formats like GIF.

Sometimes yes, sometimes no...

Update: Well it looks like I'm not the only person having troubles with image overlays, that's somewhat reassuring...

Update: Rumours about a Microsoft Earth, and me speculating about ubiquitous computing...

Macworld 2006

I've got five different browser windows open spread over the four different machines on my desk, all of them are full of tabs in an attempt to keep current on what's going on in San Francisco at Macworld.

CREDIT: Engadget
Macworld 2006 in San Francisco

Update: The keynote kicked off with the traditional retail update from Steve, but quickly moved onto the product announcements...

CREDIT: Gizmodo
The keynote

The first product to be announced, an FM Tuner and remote for iPod nano at $49, scores some points for Kevin Rose who predicted it a few hours ago, the tuner and remote will go on sale immediately. No mention of the rumoured iPod shuffle upgrades, the iLounge is suggesting that these might be unveiled in a separate event real soon now.

CREDIT: Gizmodo
...and here's Steve

Update: It looks like Kevin Rose's predictions might be on the nose, as both the rumoured iLife '06 and the promised Mac OS X 10.4.4 updates are shipping today. Does that mean Kevin's prediction of a dual-core Intel iBook is going to be "One More Thing"..?

Update: Kevin's predicted iPhoto updates along with the new Photocasting plugins, which sounds a lot like an Apple branded version of Flickr if you ask me, also seem to have come up. Looks like he's on a roll...

Update: Some iMove, iDVD and Garageband updates. Small stuff, are we all just waiting for "One More Thing" now?

Update: ...and here comes the rumoured iWeb which is going to allow you to share iLife content via the web. It apparently allows you to share photos, movies, video podcasts, music and blogs with single-click publishing to .Mac.

Update: iLife '06 will be $79 and is available today, however it will also be bundled free with any new Mac.

Update: We're on to hardware, the biggest news last year was the the move to Intel. Now Apple is ready, and Intel is ready. We're getting an Intel Mac today and shockingly the first of the new generation of machines is going to be an iMac rather than the long rumoured dual-core Intel iBooks. The new iMac will have the same design, the same features at the same price points. The difference? The new iMac is ×3 faster than the iMac G5 and uses an Intel dual core chip, each core faster than the G5. It's going to be called the iMac Duo and it'll ship today.

Update: Microsoft is committing to delivering a Universal binary of MS Office on time in March, and guaranteeing that they'll continue to ship the product for at least the next five year.

Update: The entire Mac range will be going Intel by the end of 2006, well ahead of the schedule announced last year, almost nobody predicted this...

Update: Here comes "One More Thing". Apple have been trying everything possible to put a G5 inside a Powerbook but they've failed. But there is an updated Powerbook, it's called a MacBook Pro and, as Kevin Rose predicted, it's a 15.4" widescreen dual-core Intel laptop, only 1" thick, running ×4 to ×5 faster than its predecessors with a brighter display and a built in iSight.

Update: Winding down, the new MacBook supports both Front Row and the Apple IR remote. It also has a new feature called MagSafe, a magnetic power chord. The MacBooks will start shipping in February, but you can pre-order today for $1,999 for a 1.67GHz Duo or $2,4999 for a 1.83GHz Duo.

Update: ...and that's the lot.

Update: Now the hysteria of the keynote live reports has started to die down, more details of the new iMac and MacBook along with iLife '06 and iWork '06 are starting to filter out...

Update: You can now watch the keynote on Apple's Quicktime site, and order the new hardware on the Apple Store.

Update: The first pictures of the new hardware are now starting to trickle out of Macworld in San Francisco, along with a fifteen minute video of the new MacBook being put through its paces on the show floor.

Last minute Macworld rumours

The last minute rumours now spreading across the blogosphere are a welcome return of the Apple plasma display and DVR mutterings that were knocking around earlier this month. I have to agree with Nick Carr here, it's a wonderful pre-Macworld rumour, if only because it's so unlikely, but yet so possible...

Update: Kevin Rose, who has a good history of predicting the Macworld announcements, has made just made a midnight post with his predictions for this year.

He's expecting a new thinner, dual core, 15" Intel iBook which will ship in February. Along with this he's also predicting the release of the long awaited iPod FM receiver, the much rumoured iLife '06, "photocasting" software in iPhoto, the Mac OS X 10.4.4 update and some sort of remote control, perhaps to go with a new release of Front Row?

Update: Kevin Rose's predictions just appeared on Engadget...

Update: Both the Apple Store and .Mac have gone down for the keynote, and the press are being let in from the holding area. Here we go...

Update: Blogging the keynote live...