5 ms·

Symbian Imploding As Foundation Head Resigns

20 points by noarchy 16y ago

codedivine 16y ago
Nokia's earnings call is coming up in a few days. Lets see what type of numbers they put out before declaring the death of Symbian.
imaginator 16y ago
Good. I won't miss their stupid branding (http://developer.symbian.org/ http://developer.symbian.org/). I won't miss their slow app signing (http://twitter.com/buddycloud/status/13062324669 http://twitter.com/buddycloud/status/13062324669 (ended up taking 6 weeks)). I won't miss all the crap we went through to get buddycloud working on this godforsaken platform (good work @incrediblepants). It could have been so good. But Nokia dithered around opening the platform. Even with an open platform, how many hackers really want to look at some long-closed source code and start fixing strangers bugs? Meh! Had the platform had gained adoption amongst hackers with an ecosystem comparable with Android, writing and debugging code on it was still at least 5x slower than iOS or Android and about as enjoyable as pulling fingernails with pliars. Then you were fighting back up against a terrible widget-set and UI (http://n8fail.tumblr.com/ http://n8fail.tumblr.com/). Yes, I know the "but Nokia sells more phones blah". A big ship will also keep drifting for 10km after its propellor has fallen off. On the other hand, working on Symbian has been a good example of where concentrating on niches can reap benefits; Symbian never received the hype that some other platforms have enjoyed and did allow buddycloud to gain about 500k users that were underserved by the dearth of decent social and location apps on the platform. That's been a win for us. This article is about Mr Williams leaving. The writing has been on the wall for ages - all the good people left two years ago: hired away by other mobile startups in London like Truphone or to start their own ventures. I hope Nokia can get their shit together because I like their hardware and have some friends working for them. But they don't deserve to.
- codedivine 16y ago
  Did you develop your app in Symbian C++ or Qt? What are your thoughts on Qt? I have never done any Symbian C++ stuff so cannot comment on that, but I am very much liking Qt overall.
  - imaginator 16y ago
    We started buddycloud development before Nokia did anything with Trolltech. Having switched to Qt earlier would have helped and mean that we could easily port to the N900 and what I hope nokia see as the future of their platform. Qt does seem nice and has helped Skype build a client that looks decent on most platforms. But for programming mobile apps, I would be more inclined to try and answer the framework question from a "What can I use on Android or iOS" POV. I see there are some java bindings for Android-land but the project is now abandoned. For iOS-land, I guess it depends on Steve's mood that month.
  - defen 16y ago
    Ahh but Symbian C++ is not really C++...more like its bastard stepchild which manages to remove most of the nice stuff from C++, and is just different enough to make you rewrite your whole app </bitter>
    - noarchy 16y ago
      This is precisely the sort of thing that no doubt scared away potential developers, including me. Having owned a smartphone that ran on Symbian, it was only natural (for me) that I'd want to look at how to develop for the platform. When I did, I was not encouraged.
  - daliusd 16y ago
    My impression about Qt on Symbian: it does not look native (yet) and I believe it is slightly slower than native apps. So I have not even started looking into that deeper. It is really easier to develop for Qt and I would use it for my own personal app, but for that PythonS60 is better IMHO. As for Symbian C++ API looks reasonably good. I am investigating possibility to write one commercial application and I have all required libraries/services in my phone. Documentation is good. Community support is reasonably good as well. As for language itself - it is not real C++ as mentioned in other comment. Development on Linux is possible but be ready for unpleasant surprises: no native emulator, SVG files that can't be drawn using Inkscape and etc. There is QEMU initiative but it is not ready for general usage. I have tried to build syborg but I have failed to do that.
pavlov 16y ago
From the article: "Symbian may be the largest phone operating system in the world by marketshare, but its free fall is approaching terminal velocity." I know it looks that way from the American perspective, but globally Symbian is still very much holding its own. Nokia, Symbian's corporate daddy and only remaining serious user, is selling more smartphones each quarter and increasing marketshare. That does not justify such hyberbole as "free fall at terminal velocity". (iSuppli reported 11.6% growth for Nokia/Symbian in Q2 2010: http://arstechnica.com/gadgets/news/2010/10/isuppli-smartphone-numbers-show-its-good-to-be-the-droid.ars http://arstechnica.com/gadgets/news/2010/10/isuppli-smartpho... ) I know the next argument: you're not talking about last quarter's numbers, but an impending implosion of Symbian under the onslaught of Android... Well, so far that's just speculation. Nokia has managed to hold its global marketshare for over a decade against both big Western competitors and nimble Chinese copycats. They seem to know something about this market. For what it's worth, Nokia is predicting that 50 million smartphones with the latest Symbian^3 platform will be sold next year. That's on top of cheaper units running Symbian^1 or even older variants. These devices are not iPhone killers or any other such nonsense, but they're well built and usable. They will keep selling by the millions in countries where Nokia has a strong brand and people know how much they're paying for their phone, unlike the U.S. where the phone's real price is hidden in an exorbitantly overpriced wireless contract. (I can't believe people regularly pay $80+ / month for their mobile over there...)
- imaginator 16y ago
  It doesn't matter how many low-end phones are sold. It's near impossible to put apps on them and even if the user can find a way to install an app, the users that buy low end phones have very little correlation with those that have a data contract. And even if they can a) get an app installed, b) have data, you have to deal with the crappy S40 UI. Totally not mom-friendly. My point is, it's pointless even considering these Symbian^1(who came up with such a fucking stupid branding as "^1"?) as being anything as a write off. They don't play a role in the app economy and there is such price competition at the low end that few profits could filter up to subsidise the smartphone division. So let's focus on Nokia's smartphone numbers rather than total shipments. Again... a big ship will keep drifting for a while. But eventually it will come to an end. I've seen many defections from the Symbian camp to the Android camp and have yet to find one developer yearning to go back.

20 points by noarchy 16y ago

Nokia's earnings call is coming up in a few days. Lets see what type of numbers they put out before declaring the death of Symbian.

imaginator 16y ago

Good. I won't miss their stupid branding (http://developer.symbian.org/ http://developer.symbian.org/). I won't miss their slow app signing (http://twitter.com/buddycloud/status/13062324669 http://twitter.com/buddycloud/status/13062324669 (ended up taking 6 weeks)). I won't miss all the crap we went through to get buddycloud working on this godforsaken platform (good work @incrediblepants). It could have been so good. But Nokia dithered around opening the platform. Even with an open platform, how many hackers really want to look at some long-closed source code and start fixing strangers bugs? Meh! Had the platform had gained adoption amongst hackers with an ecosystem comparable with Android, writing and debugging code on it was still at least 5x slower than iOS or Android and about as enjoyable as pulling fingernails with pliars. Then you were fighting back up against a terrible widget-set and UI (http://n8fail.tumblr.com/ http://n8fail.tumblr.com/). Yes, I know the "but Nokia sells more phones blah". A big ship will also keep drifting for 10km after its propellor has fallen off. On the other hand, working on Symbian has been a good example of where concentrating on niches can reap benefits; Symbian never received the hype that some other platforms have enjoyed and did allow buddycloud to gain about 500k users that were underserved by the dearth of decent social and location apps on the platform. That's been a win for us. This article is about Mr Williams leaving. The writing has been on the wall for ages - all the good people left two years ago: hired away by other mobile startups in London like Truphone or to start their own ventures. I hope Nokia can get their shit together because I like their hardware and have some friends working for them. But they don't deserve to.

codedivine 16y ago
Did you develop your app in Symbian C++ or Qt? What are your thoughts on Qt? I have never done any Symbian C++ stuff so cannot comment on that, but I am very much liking Qt overall.
- imaginator 16y ago
  We started buddycloud development before Nokia did anything with Trolltech. Having switched to Qt earlier would have helped and mean that we could easily port to the N900 and what I hope nokia see as the future of their platform. Qt does seem nice and has helped Skype build a client that looks decent on most platforms. But for programming mobile apps, I would be more inclined to try and answer the framework question from a "What can I use on Android or iOS" POV. I see there are some java bindings for Android-land but the project is now abandoned. For iOS-land, I guess it depends on Steve's mood that month.
- defen 16y ago
  Ahh but Symbian C++ is not really C++...more like its bastard stepchild which manages to remove most of the nice stuff from C++, and is just different enough to make you rewrite your whole app </bitter>
  - noarchy 16y ago
    This is precisely the sort of thing that no doubt scared away potential developers, including me. Having owned a smartphone that ran on Symbian, it was only natural (for me) that I'd want to look at how to develop for the platform. When I did, I was not encouraged.
- daliusd 16y ago
  My impression about Qt on Symbian: it does not look native (yet) and I believe it is slightly slower than native apps. So I have not even started looking into that deeper. It is really easier to develop for Qt and I would use it for my own personal app, but for that PythonS60 is better IMHO. As for Symbian C++ API looks reasonably good. I am investigating possibility to write one commercial application and I have all required libraries/services in my phone. Documentation is good. Community support is reasonably good as well. As for language itself - it is not real C++ as mentioned in other comment. Development on Linux is possible but be ready for unpleasant surprises: no native emulator, SVG files that can't be drawn using Inkscape and etc. There is QEMU initiative but it is not ready for general usage. I have tried to build syborg but I have failed to do that.

pavlov 16y ago

From the article: "Symbian may be the largest phone operating system in the world by marketshare, but its free fall is approaching terminal velocity." I know it looks that way from the American perspective, but globally Symbian is still very much holding its own. Nokia, Symbian's corporate daddy and only remaining serious user, is selling more smartphones each quarter and increasing marketshare. That does not justify such hyberbole as "free fall at terminal velocity". (iSuppli reported 11.6% growth for Nokia/Symbian in Q2 2010: http://arstechnica.com/gadgets/news/2010/10/isuppli-smartphone-numbers-show-its-good-to-be-the-droid.ars http://arstechnica.com/gadgets/news/2010/10/isuppli-smartpho... ) I know the next argument: you're not talking about last quarter's numbers, but an impending implosion of Symbian under the onslaught of Android... Well, so far that's just speculation. Nokia has managed to hold its global marketshare for over a decade against both big Western competitors and nimble Chinese copycats. They seem to know something about this market. For what it's worth, Nokia is predicting that 50 million smartphones with the latest Symbian^3 platform will be sold next year. That's on top of cheaper units running Symbian^1 or even older variants. These devices are not iPhone killers or any other such nonsense, but they're well built and usable. They will keep selling by the millions in countries where Nokia has a strong brand and people know how much they're paying for their phone, unlike the U.S. where the phone's real price is hidden in an exorbitantly overpriced wireless contract. (I can't believe people regularly pay $80+ / month for their mobile over there...)

imaginator 16y ago
It doesn't matter how many low-end phones are sold. It's near impossible to put apps on them and even if the user can find a way to install an app, the users that buy low end phones have very little correlation with those that have a data contract. And even if they can a) get an app installed, b) have data, you have to deal with the crappy S40 UI. Totally not mom-friendly. My point is, it's pointless even considering these Symbian^1(who came up with such a fucking stupid branding as "^1"?) as being anything as a write off. They don't play a role in the app economy and there is such price competition at the low end that few profits could filter up to subsidise the smartphone division. So let's focus on Nokia's smartphone numbers rather than total shipments. Again... a big ship will keep drifting for a while. But eventually it will come to an end. I've seen many defections from the Symbian camp to the Android camp and have yet to find one developer yearning to go back.