sallonoroff /about /blog /blog/archive /blog/stats /search

Why is wifi onboard a train so slow?

If you’ve ever tried to use wifi onboard a train in Britain you’ll know it to be an exercise in futility and frustration. I’m sure i’m not the only one that has heard or read someone complain about the experience, or even thought to themselves “I might as well just use 3G”.

You don’t have to massage the little grey cells for very long though to realise that 3G is all the train’s wifi service can be using for internet connectivity too. The tin box speeding through the British countryside isn’t plugged into an ethernet port back in Kings Cross, some spindle of cat5 cable viciously unwinding as it goes.

Icomera, a company you’ve almost certainly never heard of before, provides the wifi kit used by East Coast Trains. You can read all about their solution here, but this is the bit we’re interested in:

East Coast trains carry up to 500 passengers at speeds up to 125mph (200kmph) through 400 miles of urban, suburban and rural areas. The Icomera Moovbox M800 was identified as the only solution that could deliver the level of service required in the challenging environments. By combining up to eight 3G/HSPA modems with a broadband satellite link and implementing Icomera’s patented seamless switching technology, East Coast passengers can enjoy high speed Internet access throughout their journey.

Right, so the trains use a mix of 3G and satellite broadband – how fast is that “high speed internet access”?

Wikipedia has lots to tell us about 3G. The most pertinent information I took from that article before falling asleep is that to meet technical standards, 3G only need offer a peak speed of 0.2Mbps. Of course, many telecoms providers now offer speeds that far exceed that minimum standard but we’re not talking more than 22Mbps in an absolute best case scenario (HSPA+). Personal experience certainly tells me that it’s nowhere near that fast – I’m fairly sure the best speed test result i’ve ever had a 3G link was around 3Mbps. As for satellite broadband, Wikipedia suggests that the average download speed is just 1Mbps. Meanwhile, this article about the launch of “upto 50Mbps” broadband satellites suggests that the typical speed existing systems in the US give is around 12-15Mbps.

Now it’s time for some beer-mat mathematics. Let’s be hugely generous and assume that Icomera’s eight 3G modems each bring 22Mbps to the party and the satellite system a further 15Mbps. Further, let’s ignore how all these modems actually work together (and all associated technical overheads involved) and just assume their connection speeds easily sum together to provide us a nice round maximum speed that is made available to the train’s wifi users. I make that 191Mbps. Not bad, except it’s potentially being shared by hundreds of passengers. If only 100 passengers share it, that’s 1.9Mbps each. But you’ll only see 0.38Mbps if the train is full to capacity and everyone is trying to use wifi. And remember this is taking place in A RAINBOW-LADEN PERFECT DREAM WORLD OF IDEAL NETWORK CONNECTIVITY. 0.38Mbps in paradise. Hmm.

Come back to damp and dirty reality where those 3G modems are getting a shitty signal as they whip along at 100mph, all the various bits of networking kit involved have overheads, some part of the system is faulty, and there’s contention over those paltry 3G signals with all the mobile-phone-toting passengers who aren’t trying to use the train wifi (not to mention all those not on the train but connected to the same mast), and you’ll hopefully realise why you’re sat there complaining about the speed of the wireless connection.

Hopefully 4G will improve matters somewhat… but then again, there are more people, with more devices, looking to use such services. So maybe in future i’ll try to remember to pack a book.

PS. Most of the above also applies to the wifi we’re increasingly being offered aboard buses too. But i’m sure you’d already worked that out.

An update on the Google diet.

I was surprised to discover that it’s almost two years since I started my Google diet, so I thought I’d write a quick follow-up to reflect on how it’s going.

The answer is, in short, very well. I’ve cut a lot of Google from my life and, as I predicted two years ago, I’ve barely noticed. I don’t need Google Analytics, I’m happy with WordPress’ own simple statistics for this blog. I don’t miss Calendar, Apple’s alternatives are more than adequate for my purposes. Google Chrome is still installed on some of my devices but I can’t remember the last time I had cause to launch it (Firefox is my current preference for browsing). I’ve replaced Google Maps with Apple’s offering, which I find is a perfectly good alternative (though admittedly I do occasionally have to hit to use the incomparable StreetView). Google+ is history. Or at least I hope so; it doesn’t seem to take much to accidentally re-activate it. I’ve even dropped Google for search – the thing that brought me into Google’s arms all those years ago and that which I admitted I’d struggle to give up – and have been using DuckDuckGo full-time instead. I’d tried moving to DuckDuckGo in the past and not got on with it but in the last year or so it has dramatically improved. DuckDuckGo is now set as default in all my browsers (desktop and mobile) and I have been thoroughly impressed with it. I haven’t had cause to resort to googling once. Seriously, it’s great.

The bad news is that I’ve not broken completely free of Google yet. The reason? Gmail. I thought I’d grab all my archived mail from Google (for posterity) and be able to ditch their service but it hasn’t worked out that way. Partly it’s because the account is still tied to some other services I’ve not got around to updating yet. Partly it’s because friends and family still send me messages to Gmail, despite me only sending them stuff from my iCloud address now. The main reason though is that I just can’t trust iCloud mail. I don’t know if it’s generally flaky or an overly aggressive content/keyword filter but sometimes my mail just doesn’t seem to get through. I don’t recall (perhaps “know of” would be a better term to use) many problems receiving mail but I really can’t trust iCloud to deliver my messages. They appear in the sent folder but never arrive. I would blame the recipients except that it has happened with things I’ve sent to myself. Bothersome.

So if you’re no longer interested in being Google’s Product, I’d say it is reasonably easy to loosen their grip… but it might not be so simple to extricate yourself fully. I suppose in the end it comes down to willpower and determination. As ever, a diet is probably a good start but a change of lifestyle is better.

Removing duplicate photos in iPhoto.

Apple’s iPhoto software very cleverly checks whether you really want to import that photo that you already have in your library. Nevertheless I always manage to happen across duplicated photos from time to time. And when I do, I always think “Oh, i’ll just check if there are any more” then promptly waste 3-4 minutes checking and re-checking the various menus and buttons for the “Find Duplicates” function. It doesn’t exist. There’s one in iTunes but not in iPhoto.

If you’re similarly afflicted, I can confirm that the tiny little AppleScript available at is excellent and, despite being published in 2008, still works with iPhoto 9.6 under OS X Yosemite. Just remember to select a bunch of photos in iPhoto before running the script. I definitely didn’t forget to do that and sit wondering why it hadn’t worked. Anyway, thank you Karl!

Now, whether the script will work, could be re-written to work or, indeed, is even required in the upcoming Photos app for OS X remains to be seen.

Not dead.

Hello website visitor. Hello person whose RSS reader I’ve just turned up in.

I noticed that it has been nearly a year since my last post here and I felt I should write something to avoid it actually reaching a whole 12 months of neglect. I do keep thinking of things to write about – and there are plenty of hurriedly jotted words, sentences and links sat in the WordPress Drafts folder that are testament to this – but never get around to turning them into something suitable for public consumption. I won’t make any rash promises about updating this thing more regularly but I am going to try to go through some of the half-baked drafts and see if anything is worth posting.

Thanks for reading.


iPod. It’s a name that I often find myself idly pondering.

I always thought that “iPod” was a fantastic, forward-looking bit of branding from Apple. It was a name which would allow the humble MP3 player to evolve and advance way, way beyond just audio. Add photos. Add videos. Add games. Add PDA functions, telephony and internet connectivity. Add anything. And all with an ever-increasing storage capacity. It was clear that this vessel, this capsule, this pod, might one day carry your entire digital life. So I was very surprised and a little disappointed when the name was changed with the introduction of the iPhone and the clever iPod marque apparently demoted. I assumed, being left to gradually and graciously fade away.

Indeed, Horace Dediu, the king of analysing all figures related to Apple, recently predicted that the iPod as we know it will die in 2015.


Bob Cringely still foresees an Apple device we’ll carry around our life on:

It should be no surprise, then, that Apple — a company known for its market timing — has just started shipping a new Mac Pro. That amazing computer is overkill for 95 percent of the desktop market. It represents the new desktop PC archetype, which is a very expensive hugely powerful machine tightly aimed at the small population of professional users who still need a desktop. Unless you are editing HD video all day every day, you don’t need a new desktop PC.

What the rest of us will get are new phones and whole new classes of peripherals. The iPhone in your pocket will become your desktop whenever you are within range of your desktop display, keyboard and mouse. These standalone devices will be Apple’s big sellers in 2014 and big sellers for HP and Dell in 2015 and beyond. The next iPod/iPhone/iPad will be a family of beautiful AirPlay displays that will serve us just fine for at least five years linked to an ever-changing population of iPhones.

It’s an idea that i’ve had myself before now, so unsurprisingly I need little convincing. And there’s no reason that a future iPhone device could not return to using its ancestral moniker. Apple have surprised me before on this front and could do it again.

Meanwhile, there are lots and lots and lots of rumours of an “iWatch” or other wearable device(s) coming soon from Apple. As Gruber rightly notes, you can easily see this sort of product adopting the iPod name.

So perhaps, in one way or another, the iPod brand does have a future and could yet refer to the device we carry our whole digital life around on. I do hope so.