Symbian OS - one of the most successful failures in tech history | Hacker News Reader

Symbian OS - one of the most successful failures in tech history

eu.techcrunch.com

47 pointspavlov23 comments

Waning, maybe, but it powered boatloads of phones, employed armies of people, and made David Potter insanely rich, so I'd say Symbian hasn't been a failure at all.

I remember PSION from really back in the old days.

http://www.worldofspectrum.org/infoseekid.cgi?id=0008953

I often wondered if David Potter himself wrote that?

Edit: And I spent far far too long playing PSION Flight Simulator.

Seems like saying, MS DOS had been a failure.

Symbian will be around longer than one would expect, due to the fact that it will receive life support in the shape of Euro22M (US$33M) from the EU. Or as burst out when I saw that "WTF? Symbian gets 22 M Euro from EU... What a colossal waste of money."

http://www.electronista.com/articles/10/11/02/symbian.gets.3...

That's peanuts compared to amount of money Nokia put into the platform.

Indeed. 22M € is not a lot of money compared to ongoing handset development tracks within Nokia that will keep the market flooded with Symbian devices for the next few years, even if nobody particularly wants them in 2012.

(Nokia has recently introduced new devices with Symbian S60 3rd Edition -- that was their cutting-edge smartphone OS in 2006! They run a slow-moving ship...)

By the way, the EU-funded Symbian-centric research project is aimed at embedded devices. I don't think it will produce anything that benefits handsets much...

Some research kids in Eindhoven and Helsinki will manage to load Symbian onto some kitchen gadget, connect it to a vacuum cleaner over Bluetooth, and the result will be presented as "Euro-Home 3.0: Vision of an interconnected domicile" at some Symbian conference in 2015. Everyone involved will be pleased with such concrete outcome from this ambitiously forward-looking international research project.

> Everyone involved will be pleased with such concrete outcome from this ambitiously forward-looking international research project.

How I wished you weren't 100% spot on with that.

Well, at least we will end up with a few kids that can build such a thing. If it's spend on education it isn't all wasted. I worry more about the part that is spent on administrative meetings in Brussels.

It may well not be that much money. But why should the EU put in money there? Imagine it in the hands of a Y-Cominator type fund aimed at embedded OS development instead.

Of course, I know that is beyond the current funding apparatus which sits on this type of money in the EU, but maybe it is time we change that.

That's a different argument, you imply that Symbian will be around because of that finance. I've pointed out this is not the case. To put this in perspective a commonly known metric is that when Nokia paid 264M Euros for Symbian that, that value was the cost of two years worth of phone releases (dev. cost wise).

On this argument I suppose it depends if you automatically dislike "big guv'munt" and resent the funding, think it could be done different or for some reason think Symbian is a bad focus for this cash. Who knows.

I primarily think it is a bad focus for this type of cash. You say it is not worth much for Symbian, in light of other money spent on it. So why give it the money then?

If you spent the same amount of money on other embedded software efforts, where 22M Euro would be a lot, then I predict you would get a lot more value for money.

I don't resent big government money as such being spent on technology, but this seems badly placed. Who in their right mind would start developing new embedded systems on Symbian today? (I am by no means an embedded OS expert, but talking to my friends who develop on mobile phones, Symbian is decidedly long in the tooth these days. I can't see that it will be that much better as an embedded system. Also, see enjo's comment below on developing in Symbian.)

> The browser was always a second class citizen, a third party component

Before the iPhone, the browser in Nokia phones was almost cuttting-edge, based off a version of Webkit shortly after it was open-sourced. Unfortunately, it stagnated as Nokia neglected to track the Webkit trunk.

My Nokia E71 could render the iPhone version of Google Reader, with all it's AJAXy goodness, perfectly and far better than (say) a Blackberry. That's how I read most of my feeds for about two years. (I now have an Android.)

Nokia is now pouring a lot of effort into QtWebkit and the Symbian mobile browser, which should resurrect that part of the platform and bring it up to this decade's standards.

They're also basing their JavaScript application platform ("WRT") off QtWebkit, with hooks to allow access to native O/S functionality, and it's being heavily promoted as a development platform on equal footing to Qt C++. Much of this development is going on in the open (http://bugreports.qt.nokia.com/secure/BrowseProject.jspa), so you can track this progress.

For all the flak Symbian has recently gotten, it was actually great OS at a time. I had a phone in 2005 which did multitasking, copy-paste, ssh, python scripts and web browsing. Yes, it was dog slow, and the browser wasn't that great, but still it was quite a feature set, and the phone wasn't even that expensive.

No it wasn't.

It was a marginally functional OS full of unbelievably quirky behavior. The underlying subsystems were so fundamentally broken (and poorly written) that Symbian itself couldn't actually fix them. There is a reason messaging never quite worked right. The installer was perpetually broken as well. I never seen a project that introduced more regression from release to release than Symbian.

Fundamentally it was a C++ based OS that was introduced before C++ really existed as a standard. They independently invented and implemented a number of features that would eventually become standard within C++ (exceptions spring to mind).

Either way, Symbian is a mess. It's seriously a warning about the dangers of Object Oriented development gone terribly wrong. Huge class hierarchies? Yep, Symbian's got em. Convoluted multiple inheritance with indeterminate behavior? Check. The list goes on and on.

Symbian was (and is) so fundamentally flawed that while your phone did all of those things, it also failed far to often at installing apps, providing usable interfaces, and actually getting messaging and things to work.

*note: I'm a bitter refugee from the Symbian world having been principally involved in one of the most widely distributed apps written for the platform.

I never actually wrote Symbian code, but there was a time where it was on the horizon at the company I worked at, so I did some reading to prepare me for it. The only horrors I actually remember are the string classes and the exception handling, but even if the rest of the system were perfect, I would still agree with you.

Your comment is from a developer perspective, while zoiker was talking as a user. I agree with both of you, but there is a difference.

    iPhone too, uses FreeBSD at its lowest level ...

Wait, what? Is this just a reference to OSX's BSD heritage or something that I've missed all this time?

OS X has a lot of FreeBSD code, but it isn't a OS running on top of FreeBSD.

Lowest level of user-land perhaps. Plus some sections of the kernel were adapted from FreeBSD, IIRC.

It's close enough for tech-blog work, I suppose.

OS X (and derivative iOS) are essentially proprietary GUIs put on top of a FreeBSD base system on top of a hybrid Mach/XNU kernel with a bit of FreeBSD mixed in.

"Breaking - Nokia and Symbian making "major announcement" at 1530 GMT" tweeted by BBC's Rory Cellan-Jones.

"That "major" announcement from Nokia - "Nokia reaffirms its commitment to the Symbian platform"

Also, wrapping the Symbian Foundation into a 'licensing operation': http://www.symbian.org/news-and-media/2010/11/08/symbian-fou...

Reply on news.ycombinator.com