Hacker News new | past | comments | ask | show | jobs | submit login
iOS 6 Adoption Stats One Month In (stuartkhall.com)
46 points by appbot on Oct 19, 2012 | hide | past | web | favorite | 49 comments



I haven't updated to iOS6. I was worried that google maps would be deleted from my iPad and replaced with apple maps. I do not want this.

I heard somethings about Safari being buggy.. I didn't really want this either as I only really use my iPad to surf the web.

I was concerned about the memory footprint. As I understand it app's now download retina resources whether you have a retina iPad or not. I have limited space on my iPad. It runs fine at the moment. I didn't see any reason to update.


> I haven't updated to iOS6. I was worried that google maps would be deleted from my iPad and replaced with apple maps. I do not want this.

I'm also waiting for 6.1 at least to see how Apple Maps will evolve, I have no sense of direction whatsoever so I need maps to show me where I am, and Apple Maps still seems a bit too risky for now. It's possible to use the Google Maps site directly, but in my tests it's less responsive and the geolocation doesn't work as well as in the application itself.

> I heard somethings about Safari being buggy.. I didn't really want this either as I only really use my iPad to surf the web.

I've seen developer reports, but no reports of user issues. Quite the opposite, webkit's been updated in iOS6 so Safari is apparently significantly faster.

> I was concerned about the memory footprint. As I understand it app's now download retina resources whether you have a retina iPad or not.

It's always done that (ever since the release of the original iPad and the creation of "hybrid" iPhone/iPad applications), the version of iOS you're running is completely irrelevant to that issue: if you're running an application compatible with all iOS devices (3GS to iPad 3 at native resolutions) it may very well bundle 4 different copies of bitmapped assets (or even 5 for e.g. startup screens, since the iPhone 5 has a different resolution than the 4/4S). I don't believe iOS AppStore bundles have every been stripped. I'm not even sure they could be since they're signed (unless you jailbreak).


Just as a data point on this whole Maps thing, Apple Maps in iOS 6 has already saved me on numerous occassions in the US and Europe. In Europe, I had a dedicated Garmin Nuvi, but it was Maps that got me to the the tiny village in Alsace. In the US I had a dedicated TomTom, but it was Maps that was able to correctly locate sites of interest for Civil War fans (Eastern Theater). Both dedicated GPSes had had their maps updated in the last month or so.

Make of that what you will, but my personal experience has been that Maps is actually pretty reasonable. In major cities it maybe isn't quite as detailed as Google maps, but then, you can always use Google Maps as a web app for that...


In general, a map will always be good enough until the moment they aren't. It's extremely localized, and your experience can't be compared to another persons' 1-1.

There are now entire countries with completely useless maps. Is that fair to iPhone users in those countries? Even in Canada and the USA, thousands of small cities that formerly had detailed maps were "wiped off the map" with this update. So while millions of people think things have gotten better, millions of people can also legitimately claim Maps is not just "worse" but completely broken.

For the people who live in cities with turn-by-turn navigation and 3D flyovers, hey the new Maps is great. Apple has sacrificed a product that was good enough for most people with one that is great for some and terrible for others. It's a tough decision for them to make, and they should have still used Google Maps app for those who are affected negatively since theres still so much time left on the contract with Google.


> It's a tough decision for them to make, and they should have still used Google Maps app for those who are affected negatively since theres still so much time left on the contract with Google.

If they had given themselves a year where they made Maps available incrementally "by region" then they could have given themselves more time to test and refine the data and enticed users waiting in expectation instead of disappointing a whole bunch of them en-masse. Everyone outside the SF Bay Area could have looked on in expectation while us poor saps would look forward to the day.


"...where they made Maps available incrementally "by region"

I can see where you're coming from but I don't know if isolating map availability by region would work. Hyper-local travel is used in the majority of cases, but there are cases of regional travel too. What happens if someone wants turn-by-turn navigation from California to, say, Oregon, to attend a conference, weekend getaway, job interview, or just simply meet at a coffee shop in the new region? I live in Seattle, and often drive to Vancouver BC or Portland for weekends, just as an example.

IMHO, the real mistake here was setting expectations so that it would be lower and not affixing the "Beta" label to it (as was done to Siri.) Apple should also have, side from collecting data logs, ensured there's a more prominent feedback loop. As to your suggestion, I think a broader approach would work better, and perhaps work in layers of information, drilling down more deeply with richer details in regional pockets, one broad layer at a time.


> I can see where you're coming from but I don't know if isolating map availability by region would work. Hyper-local travel is used in the majority of cases, but there are cases of regional travel too.

No duh. It wouldn't "work" in the sense of excluding absolutely 100% of people. So what? From Apple's perspective, it would be fine if it worked just 85 to 90% of the time. The point is to avoid a massive amount of customer dissatisfaction.

> IMHO, the real mistake here was setting expectations so that it would be lower and not affixing the "Beta" label to it (as was done to Siri.)

No disagreement there. Doing both things would have been even better.


> I had a dedicated Garmin Nuvi, but it was Maps that got me to the tiny village in Alsace.

Huh? What kind of village in Alsace is not referenced on your Garmin Nuvi (or any kind of maps)? For example, all villages have a wikipedia page!


Thank you for the data point. I think I'll still (admittedly a bit irrationally) wait for 6.1 to be released, and I'll probably update at that point.

I've little reason so switch yet anyway.


You realize that a dot-release likely won't update Maps a whole ton right? The application itself that's installed on the phone is fine. It's the data that some people find to be inaccurate or not detailed. Apple has been updating the data independently of OS releases, AFAIK.


Of course, but I think it'll iron other potential issues and will likely be released at a point where Apple feels they have improved a map enough to say "here look new version, more shiny, maps no sucky anymore". The iOS 6.1 point I don't see as something which will improve map, but as something which will mark maps as improved.


> I don't believe iOS AppStore bundles have every been stripped. I'm not even sure they could be since they're signed (unless you jailbreak).

My understanding is that when you submit your app to the App Store, Apple actually re-sign it with their identity instead, so I guess that theoretically they could remove resources from your bundle and then sign it again, before sending it to your device. But they don't do this.


> My understanding is that when you submit your app to the App Store, Apple actually re-sign it with their identity instead

That is correct.

> so I guess that theoretically they could remove resources from your bundle and then sign it again, before sending it to your device. But they don't do this.

But then they'd have to split the application into a dozen different bundles, and if the user uses iTunes to manage his device this dozen different bundles would then need to be downloaded as there's no way for Apple to know for certain which devices will be synchronized using that iTunes installation in the future.


>if the user uses iTunes to manage his device this dozen different bundles would then need to be downloaded as there's no way for Apple to know for certain which devices will be synchronized using that iTunes installation in the future.

Good point - I'd forgotten that some people still manage their applications using iTunes!


I've updated my 4s and put at least 1000 miles on turn by turn maps and everything is okay (turn by turn is beautiful, if that matters at all). I did add the maps.google.com shortcut to one of my folders, but I haven't needed to use it.

I havent noticed any bugs in safari, I did have to disable the iCloud tab sync since my son is the main iPad user and all of his mine craft tabs were showing up on my phone and computer.

You should try the maps for yourself though, all seems fine to me.


> I was worried that google maps would be deleted from my iPad and replaced with apple maps. I do not want this.

Apple Maps does just fine for me with driving directions. If you don't like them, you can also use Mapquest. On a 4G connection, Google Maps also works very well as a web app. HopStop takes care of my commuting needs just fine.

> I heard somethings about Safari being buggy.. I didn't really want this either as I only really use my iPad to surf the web.

Safari on my New iPad on iOS 6 works significantly better than Safari on my original iPad on iOS 5.1.1. For one thing, it never crashes.

> I was concerned about the memory footprint. As I understand it app's now download retina resources whether you have a retina iPad or not. I have limited space on my iPad. It runs fine at the moment. I didn't see any reason to update.

As an iOS developer, let me assure you that in most instances, Retina resources amount to a rounding error in most cases. (~35kB in the case of a full screen PNG image for an iPhone.) Also, they only take up a part of the permanent flash store, not RAM.


> Safari on my New iPad on iOS 6 works significantly better than Safari on my original iPad on iOS 5.1.1. For one thing, it never crashes

That's because the new iPad has 4 times the RAM of the original iPad. Lack of RAM was the cause of almost all Safari crashes on the original iPad. Safari under iOS 5.1.1 on the new iPad was as stable, or even a little bit more stable, than under iOS 6.


I updated last weekend. I use Safari a lot and haven't had any issues. The weird thing I experienced during the upgrade is my iPad (3rd gen) got 1 Gb of memory freed - in that section iTunes calls Application Data. Still, 1 Gb is 1 Gb and I'm glad to get it back. And no, I haven't noticed anything amiss. The memory footprint on my iPhone remained the same.

As far as maps goes, I don't use maps much (either Google or Apple). When I really need a map for long distance trips I use my Garmin. My experience with the mobile map apps so far is they're junk compared to the Garmin. I do need to do some traveling this Monday so I'll have the opportunity to do a side-by-side comparison test.


Had the opportunity to seriously use Apple maps yesterday for the first time. I liked it better than my Garmin (which is a 2008 model). I like that it gives you street names for turning and that it tells you of upcoming turns and exits well in advance. The fact that it cuts in over the music being played by my iPhone is a practical plus - the Garmin simply can't do that. Finally, a plus for pointing out hotels and gas stations. My Garmin definitely didn't do that either.


Safari has been fine on my iPhone 4 - probably better than in the previous version...


1. Graph is a bit suspicious--why does the adoption rate sometimes go down? (Some quite large drops as well.) 2. Any equivalent graphs for other versions? 3. Adoption rate pretty flat since about two weeks after the iOS 6 release. 4. What percentage of devices can be upgraded to iOS 6? (Probably 95%+, since the 3GS is supported--?)


Since this is "the % of our app's users using iOS6", the rate going down it probably means that in a given day the rate was 70% and the other less than that (meaning more total users, or just more active users that day with less than iOS6, etc).


Nothing to be suspicious about, that's just we are seeing, can't tell you if it's representative of what others are seeing (although the %'s seem to be in the right ballpark to what others have reported). Spikes seem to be related to when an update for one of our apps come out, e.g. around the 2nd (the big spike) is when our update to Discovr Music came out (https://itunes.apple.com/au/app/discovr-music-find-new-music...) I guess people who updated to that are more likely to of updated to iOS 6. In that actual release (2.6) we are seeing more like 75% iOS 6 usage.


Yeah, that would be interesting to also show what percentage of people are on the latest version of your software. A lot of less techie people just never update anything, unless it breaks.


If this were an adoption rate, then a rate can certainly drop. Imagine 5% of current users upgrade on Monday and then only 2% upgrade on Tuesday. A rate is not "raw user numbers."

In the case of this article, additional new users on an older version of iOS would cause a drop in the percentage of users on iOS 6. For example, I have not yet upgraded my iPhone 4 to iOS 6. If I start using this app today, I'd increase the number of users, but cause a decrease in the precent of users on iOS 6.


It's not a rate, though, it's "adoption". How can it drop? Unless users tried iOS 6 and went back to 5 en masse.


I submitted before I was finished editing. Apologies for that. I noticed your reply after I was finished. I address this above.


I can see your point, I guess an accurate description of the graph would be the percentage of people using our app on iOS 6 on each particular day


Could someone please describe this graph in detail? Unfortunately there are no numbers to go along with it which is challenging for those like me who are totally blind.


This is a bar chart of iOS6 adoption over time.

It starts at September 1st 2012 on left and ends at October 16th on the right, with a bar for each day that indicates what percentage of their users were running iOS6 on that day. The days are labeled terribly so none of the dates I'm about to give are exact.

From September 1st to around September 18th very few of their users ran iOS6. It's hard to tell because the graph is incredibly coarse but at most, 2% of their users were using iOS6. The graph points out that these were the beta testers.

The official launch is labeled, and on that day iOS6 adoption soars. On around September 21st, just a few days after the iOS6 launch, already 35% of their users were running iOS6. By September 25th, a little less than 50% were.

On October 2 the adoption rate jumps to the highest shown on the graph, a little under 70%. The next day it sinks down to 60%, and the next day it sinks back down to 52.5%.

For the last few weeks of data starting on October 6th, the adoption rate stays steady at around 55 or 57%.


The graph shows percentage of users having installed iOS6, one bar per day. The bars are very close to 1-2% in early september (beta testers), then around the release 19th of September, the bars quickly rise each day: 30%, 35%, 45%, 47%, 50%, and from there on a quite steady (but jumpy) rise towards about 60% today. There's one weird peak at about 1st of October at about 65%, but that's probably just happenstance.

Edit: The difference in the numbers me and the fellow who posted just at the same time shows how coarse the percentage-axis was ;)


I think I like yours better. Much more succinct, with pretty much the same data.


Thanks to both of you for your descriptions.


Sorry there are %'s on left axis, it goes from 0% at the bottom to 70% at the top.


Have they fixed the ridiculous "Only one connection allowed at a time to an http server" change yet? :/ As far as I'm aware, the browser is fairly unusable because of this.


This limitation doesn't exist. Safari happily opens up to 4 connections per server.

There is however a bug with long-polling connections: If you fire off the XHR for the long-polling endpoint (or Server Site Events endpoint for that matter) before window.onload has fired, then Safari will not fire onload (nor load any other external resource) until the long-polling connection closes or returns something.

What you are seeing is people not updating their scripts yet to take this bug into account.


> What you are seeing is people not updating their scripts yet to take this bug into account.

Why? Shouldn't the expectation be that the bug gets fixed?


Because meanwhile, your users/customers experience the bug, which for you means lost page views, mindshare and probably revenue...


I found the browser a lot better/faster on my 3GS after the update, probably due to Javascript speed improvements, so it's very far from "unusable".


I haven't noticed anything and I use Safari/Chrome pretty heavily on my Touch. Under what conditions would this noticeably affect my browsing?


Mobile safari should use SPDY, then this wouldn't be as much of a problem :-)


Another source shows the same plateau at 65% or so: http://david-smith.org/iosversionstats/

I am a bit surprised from those numbers that there weren't more people left on 5.0, since 5.1 was an OTA update like 6. But maybe this is more of an "I don't want the update" thing than a "I don't know about the update" thing.


My ex wife's iPhone 4 is still using ios4 since I'm not around to do it for her. I chuckle a bit every time she sends me a text message.


I’d be really interested to see iOS 6 adoption stats compared against previous releases of iOS. Does anyone know if there’s anything published on this?


Marco gives one datapoint for iOS 5.0 here for Instapaper: http://www.marco.org/2011/11/30/more-ios-device-and-os-versi...

49 days = 45%


This is a very good question. For example, I haven't upgraded because my phone is jailbroken. As of 2009, 10% of idevices were jailbroken with Cydia installed. Couldn't find any more recent stats.

http://en.wikipedia.org/wiki/Cydia


Yes, from what I have read there is still no "untethered" jailbreak for iPhone 4 iOS6. That's what's keeping me from upgrading.


And all the iPad 1s can't upgrade.


I wonder how many of non-adopters use jailbroken devices.




Applications are open for YC Summer 2019

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: