The good news for Nokia is that things can't get much worse

[andrewducker]

I was hoping, that despite my dislike for Microsoft's app policies*, that they would make their phone system a success, and we'd thus have more competition. Much though I like Android, I'd like there to be three main players so that there's constant pressure on all of them to improve.

But unless Windows Phone 8 is a startling success, this isn't looking very likely:


I shall have to hope that Firefox OS takes off instead, which would at least provide an open alternative.

*You can only install from the Microsoft app store, and they have policies in place against adult content of any kind.

Comments

[bart-calendar.livejournal.com]

I don't understand the Samsung numbers. Nearly everyone I know in France who has a smartphone has a Galaxy.

[calum]

They're the US numbers for Q2 2012, and there was a legal injunction preventing Samsung selling phones in effect.

[andrewducker]

Samsung have the largest share of Android - which is what you're seeing. But the Galaxy is still a premium phone, it's about 600 euros for the newest one. Lots of people are going for cheaper phones, which other providers sell lots of. Those people just aren't your friends :->

[andrewducker]

That too!

[bart-calendar.livejournal.com]

Gotcha. Weird. I mean my friends aren't rich.

It must be the schemes they have here. Rome Girl is trying to convince me to get a smartphone and has pointed me to the offer offered by our Internet provider which is 19 euros a month for 23 months for a Galaxy SII.

I can't imagine any of my friends have 600 euros to spend on a phone so they must be going for similar offers.

http://mobile.free.fr/mobiles.html

[bart-calendar.livejournal.com]

Thanks!

[andrewducker]

Yeah. But that's for people who use a phone enough to have a contract (I know several people that are on Pay As You Go), and spend 19 euros a month. If you are only paying 5 euros a month you end up with a cheaper phone (like the ZTE Blade, or the Galaxy Y).

[bart-calendar.livejournal.com]

Yeah. My current non smart phone is pay as you go. In fact the people I know who don't have smart phones all have cheap pay as you go phones like mine. It's just the people with smartphones who have Galaxies.

Chris is convinced that it would be a good idea for me because I wouldn't have to stop by the apartment several times a night to check work emails. My worry is that it would leash me to work constantly and cost me money to boot!

[andrewducker]

Yeah - Smartphones are slowly eating into the cheap phone market. Not all the way there yet, but apparently over 50% of all phones are now smartphones.

I know what you mean about being leashed to work. I don't have a phone I can get work email on yet, but I'm sure it's only a matter of time...

[chess]

Is it not the case that Nokia's target market is actually the cheaper end of the market, and would therefore be under-represented in the US, though?

[andrewducker]

In that case Symbian is doing bearably well, but Windows Phone is still nowhere to be seen:
http://www.pcworld.com/article/256155/android_apple_own_80_of_global_smartphone_market_microsofts_share_22.html

[steer.livejournal.com]

I said this on twitter but what is up with that chart? The blackberry part "9%" is larger than the HTC 14%. The 51% for Android OS is only about 2 times the size of the 9% for blackberry. The 9% android other is well smaller than the 2.9 % for the HTC Windows Mobile...

[andrewducker]

I agree. Not sure what the hell Nielsen are up to, as the RIM Blackberry 9% is only slightly smaller than the Apple 34%!

Got 5 minutes to put together a more proportionate one?

[drplokta]

You should note that there's been serious criticism of that particular diagram, because the areas of the coloured boxes aren't remotely proportional to the percentages. The iOS one should be much larger, and the ones to the right of Blackberry (which should itself be smaller) should be little slivers.

[steer.livejournal.com]

Can't find the raw data anywhere alas (can't be fagged to type it all in).

[steer.livejournal.com]

As I said above -- for a start the "Android" box should take up most of the diagram (more than 50%). I wonder if they're doing something logarithmic on the x axis or if it's just made up?

[anton-p-nym.livejournal.com]

It's a pity that Windows Phone isn't doing better; I just got a Nokia Lumia 610 yesterday and I find WP works very well even on that limited hardware, and the integration with Twitter, Yahoo, and Xbox Live is brilliant.

-- Steve eyes European cell plans with deep envy, given that his 3G voice+data plan runs him CAD$50/month. (Canada definitely needs more players in the mobile market.)

[octopoid-horror.livejournal.com]

"use a phone enough to have a contract"

Given the status of hip up-to-the-minute phones as lifestyle accessories, surely you'd be better off saying "want a specific phone/use a phone enough to have a contract."

I first got a contract to spread the upfront cost of buying a fancy phone - I still don't use a phone enough to justify a contract, I just feel more comfortable spending more than the cost of a handset spread over a longer period than splurging a vast amount of money on a fancy smartphone!

[andrewducker]

True. I suppose I'm thinking of people who still think of phones as phones, and thus justify their monthly spend based on the minutes they get for it, rather than the phone they get.

[andrewducker]

I just spent ten minutes playing with Google's charting tool for Treemaps, which sadly doesn't display multiple levels at once, and also doesn't have a way of saying "Put all of the children of a parent in one column" and so I couldn't get anything terribly pretty out of it.

Have some raw data:
['Location', 'Parent', 'Percentage'],
['Total', null, 0 ],
['Android', 'Total', 0],
['Apple', 'Total', 34],
['Blackberry', 'Total', 9],
['Windows', 'Total', 0],
['Symbian', 'Total', 0.9],
['Palm/WebOS', 'Total', 0.6],
['Samsung (Android)', 'Android', 17],
['HTC (Android)', 'Android', 14],
['Motorola', 'Android', 11],
['Windows Mobile', 'Windows', 0],
['Windows 7', 'Windows', 0],
['HTC (Windows Mobile)', 'Windows Mobile', 2.9],
['Palm', 'Windows Mobile', 0.1],
['Other', 'Windows Mobile', 0.2],
['Samsung (Windows 7)', 'Windows 7', 0.5],
['Nokia', 'Windows 7', 0.3],
['HTC', 'Windows 7', 0.5]

[manintheboat.livejournal.com]

I have a Palm Veer because it's the *tiny* smart phone. The OS is awesome. I think I'm the only person who bought one.

[skx.livejournal.com]

I use my phone often enough that a contract makes sense - because it is both cheaper than pay as you go for the volume I use it and because I prefer *fixed* bills for budgeting purposes.

That said I did purchase my phone and arrange the contract separately. (HTC Desire, about 14 months old now.)

[daveon.livejournal.com]

Given the history of web based OS, I'm not holding my breath. People who've played with it have complained about performance too.

[andrewducker]

I'm not holding my breath either. I've not seen any sites that work as well as actual applications.

But I'd settle for them pushing things forward a bit more.

[momentsmusicaux.livejournal.com]

I'm never entirely sure about the 'competition to improve' principle.

Ultimately, it means duplication (or triplication) of effort. That's tons of hours of reinvented wheels.

[drdoug.livejournal.com]

The thing that strikes me about Windows Phone 7 and 8 is that almost everyone remotely techie I know wants it to be a success to bring in more competition ... but almost none of them are thinking of going as far as buying one. It's becoming the OS you wish other people would buy - which doesn't bode well for volume.

[andrewducker]

Yup. But different people like different wheels. If Apple hadn't done the smooth UI they did then I'm not convinced others would have moved that way so much. Android notifications were better than Apple ones, and they've pushed Apple to improve on that front.

But on top of that, there are advantages to different approaches that are mutually exclusive - I love some bits of software that other people find incomprenensible. Having both of them out there makes life better for both of us.

[andrewducker]

If I want an awesomely smooth phone that's locked into one way of doing things, I'm getting an iPhone. If I want a phone that I can do whatever I want with, I get an Android.

If the Microsoft phones were also open, so I could use different keyboards, home screens, etc. then I could see myself trying one as my next phone. But they aren't, so they won't get my cash.

And on the One True Awesomeness route they just aren't going to beat Apple. Even if they managed to produce something better than the iPhone they wouldn't get the majority of the Apple crowd to use an MS device.

[daveon.livejournal.com]

We're doing a lot with HTML5 and JavaScript with a native 'wrapper' and, on the whole, the experience isn't bad. It's how the Renovation app worked and, assuming we get the backend access working, the ChiCon app.

But on older devices you really notice the performance drop off.

[daveon.livejournal.com]

I'll be honest here, it is a nice OS and the Nokia Lumias are excellent.

BUT they made a massive mistake by picking a UI paradigm that's completely alien to Android and iPhone devs/users. It means that your development base is always going to lag.

[momentsmusicaux.livejournal.com]

I take your point.

I read an article (possibly one you posted...) about how Android was being developed to be very much like the Blackberry with a keyboard until the iPhone launched and they did a very hasty redesign.

[andrewducker]

Yeah, the original Android apparently looked like this:
http://tctechcrunch.files.wordpress.com/androidphone.jpg

Which really isn't very impressive!

[andrewducker]

Yeah - it's nicely unique, but it doesn't make getting devs on board easier.

Which, I guess, Metro on desktop is supposed to solve. But I'm not convinced that will be a success either!

[andrewducker]

Oh, on the other hand, there are dozens of sites that allow you to do a family tree, and that means that the data is fragmented all over the internet. That's a perfect example where having a single dataset would be incredibly useful.

[steer.livejournal.com]

PNG
http://www.richardclegg.org/graphics/phoneshare.png

or

PDF
http://www.richardclegg.org/graphics/phoneshare.pdf

Actually, it's bloody difficult data to display well so I'm not really happy with either of those.

Took me some time (a couple of hours) to get it even this good.

[steer.livejournal.com]

Oh, and obviously with something like this aspect ratio is important -- the rectangles will only be "right" if your screen aspect ratio is such that a "square" is "square". I *think* the PDF should print correctly like that.

[steer.livejournal.com]

R code

library("ggplot2")
phones<-read.table("rawdata.txt", quote="'",header=TRUE, sep=",")
phones$width<-by(phones$percentage,phones$os,sum)[phones$os]
phones$cumwidth<-cumsum(sort(tapply(phones$percentage,phones$os,sum),decreasing=TRUE))[as.vector(phones$os)]
phones$height<-phones$perc/phones$width
phones<-phones[order(phones$width,decreasing=TRUE),]
heightsum<-data.frame(unique(phones$os))
heightsum$height<-0
phones$cumheight<-0
phones$widenames<-""
phones$minpercentage<-""
for (i in 1:nrow(phones)) {
prow<-which(heightsum$unique.phones.os.==phones[i,]$os)
heightsum[prow,]$height<-heightsum[prow,]$height+phones[i,]$height
phones[i,]$cumheight<-heightsum[prow,]$height
if (phones[i,]$width > 5) {
phones[i,]$widenames<-as.vector(phones[i,]$device)
}
if (phones[i,]$width > 1) {
phones[i,]$minpercentage<-paste(as.character(as.vector(phones[i,]$percentage)),"%",sep="")
}
}
postscript("phoneshare.ps")
g<-ggplot(phones,aes(ymin=-0.16,xmax=110)) + xlab("Market share")+ ylab("") +coord_equal(ratio=100)
g<-g+geom_rect(aes(xmin=phones$cumwidth-phones$width, xmax=phones$cumwidth,ymin=phones$cumheight-phones$height, ymax=phones$cumheight,fill=device))
g<-g+geom_text(aes(x = phones$cumwidth-phones$width/2, y = phones$cumheight -phones$height/2+0.04, label = phones$widenames))
g<-g+ geom_text(aes(x = phones$cumwidth-phones$width/2, y = phones$cumheight -phones$height/2, label = phones$minpercentage))
g<-g+ geom_text(x=25,y=-0.02, label="Android (51%)")
g<-g+ geom_text(x=67,y=-0.02, label="Apple (34%)")
smallsize<-0.07
g<-g+ geom_text(x=92,y=-0.02,label="RIM (9%)",size=smallsize)
g<-g+ geom_text(x=95,y=-0.05, size=smallsize, label="Win Mob (3.2%)")
g<-g+ geom_text(x=97,y=-0.08, size=smallsize, label="Win 7 (1.3%)")
g<-g+ geom_text(x=98,y=-0.11, size=smallsize, label="Symbian (0.9%)")
g<-g+ geom_text(x=98,y=-0.14, size=smallsize, label="Palm (0.6%)")
g<-g+ geom_text(x=100,y=0.4, size=smallsize,label="0.9%")
g<-g+ geom_text(x=101,y=0.6, size=smallsize,label="0.6%")
g
dev.off()

[steer.livejournal.com]

Data

'device', 'os', 'percentage'
'Samsung (Android)', 'Android', 17
'HTC (Android)', 'Android', 14
'Motorola (Android)', 'Android', 11
'Other (Android)', 'Android', 9
'Apple', 'iOS', 34
'RIM Blackberry', 'RIM Blackberry', 9
'HTC (Windows Mobile)', 'Windows Mobile', 2.9
'Palm (Windows Mobile)', 'Windows Mobile', 0.1
'Other (Windows Mobile)', 'Windows Mobile', 0.2
'Samsung (Windows 7)', 'Windows 7', 0.5
'Nokia (Windows 7)', 'Windows 7', 0.3
'HTC (Windows 7)', 'Windows 7', 0.5
'Nokia', 'Symbian', 0.9
'Palm WebOS', 'Palm', 0.6

[steer.livejournal.com]

Corrected plot
http://www.richardclegg.org/graphics/phoneshare.png
http://www.richardclegg.org/graphics/phoneshare.pdf

See above -- only "correct" if aspect ratio of screen is such that a square is square. :-)

(Corrections welcome -- plotting code in R is above).

[andrewducker]

I'm fairly sure that unless I tell my PC to stretch things to fit the screen squares are always shown as squares.

Also, thank you!

[steer.livejournal.com]

No problem -- I needed an excuse to spend some time playing with ggplot in R anyway as I'm supervising a student using it.

The "correct" one looks startlingly different. You can see why the consultants picked the distorted one as it is much more eyecatching but when compared with the "honest" one it is remarkable how wrong the original is.