Hacker News Reader

Comment by substation13 | Hacker News Reader

Anyone else feel sad coming across something like this? So many hours spent for a product that is largely forgotten. Glad to see the pieces out in the open though.

View on HN

Symbian, the Ozymandias of smartphone operating systems.

They had a lot of the right ideas very early. They started building a high-resolution graphical mobile OS for 32-bit ARM in the mid-1990s, when most phones barely could display two lines of text and receive an SMS. Many concepts in iOS and Android today were pioneered by Symbian.

IMO, the biggest failing of Symbian was to disown the UI and focus on the lower layers of the operating system. Symbian's licensees like Nokia, Motorola and SonyEricsson wanted to control the GUI layer themselves, so they had annoyingly incompatible GUI libraries for Symbian. The platform was already hard to develop for, and then you had multiple vendors piling gunk on top like Nokia's Series 60 which was a terrible piece of work.

In a parellel universe Symbian would have acquired the BeOS team in 2003 and kept them in Silicon Valley, tasked with building a truly excellent mobile GUI on a five-year time horizon independent of Nokia's meddling. A well-designed high-level UI framework in a reasonable language could have made the embedded C++ underpinnings of Symbian mostly irrelevant — basically shipping Android a couple of years before the fact.

TIL: https://en.wikipedia.org/wiki/Ozymandias

Also one of several instances where the Breaking Bad writers put a lot of thought into amazing episode themes and titles: https://en.wikipedia.org/wiki/Ozymandias_(Breaking_Bad)

I think as long as it was companies like nokia who liked to create smartphones that where primarily phones with texting first and computers a very distant second Symbian did not have a chance.

I know that they made stuff like the 9x00 communicators but they where never intended as mass market products.

Nokia was never able to craft a desirable product that brought applications to the mainstream user, most people who got their smartphones only used them as phones.

Third-party apps for Symbian may not have been mainstream, but third-party games definitely were. Even before that, Java ME games were popular on feature phones.

Before the iPhone, mobile device companies were held ransom by wireless carriers. Carriers dictated everything about the phones, down the Verizon demanding Bluetooth file transfer was disabled to force people to use data minutes to send pictures.

This is one of the things that I think modern techies have forgotten. No matter what you think of Apple, they were the only ones with enough market power to force carriers out of that position (Apple having been newly minted at the king of digital music) . It’s why the iPhone was exclusive to AT&T — because Verizon was the market leader and AT&T had to make the deal to get the iPhone. And Jobs wouldn’t make the deal with any carrier that insisted on adding crapware to the phone.

This was true in the US. I don’t know about Europe. But it wasn’t true anywhere else. Even in the US it was dominated by Motorola (and BB) as a result. Motorola didn’t have much success outside of Razr in the rest of the world and Nokia and Sony didn’t have much success in the US. So it was really a tale of two worlds. I really liked the Sonys and Nokias of the era.

Symbian didn’t screw up the UI and it certainly was “too open”. It was so common to have malware and viruses on nokia phones. Bluetooth was scary.

The biggest issue of all was the incompatible app ecosystem. Each phone and OS and sometimes even models were different. Some were Java, others were Symbian and then the whole world of blackberries and windows.

It was not true in Europe either. You could of course use any device, and in many countries bundling a device with a phone contract was even expressly forbidden. Carriers had no control at all.

Insert the usual disclaimer that Europe has a lot of countries with different legislation, but the carrier stranglehold was a US specific problem.

I've been working for a carrier in Italy in the first half of 2000s. We designed some phones with manufacturers but any phone could work on our network AFAIK. There was not much choice at the beginning because we only had a 3G network and there were very few 3G phones (only one at launch.)

> was so common to have malware and viruses on nokia phones

I never experienced that and don’t know anyone who did. Was it really a thing?

I don’t think this is true. I worked extensively in the mobile industry back in the day and all Nokia apps had more painful signing process than iOS apps have today.

No, It's true, most my fiends who used to install pirated application (.sis file) from internet got affected by these viruses and at that time most popular antivirus companies (McAfee, Kaspersky, f-secure) had Symbian versions of their antivirus programs.

https://en.wikipedia.org/wiki/Cabir_(computer_worm) https://en.wikipedia.org/wiki/Commwarrior

I'm not sure what you expect here. Is it the manufacturer's fault if you sideload virus infested applications?

Note that these are not computer virii by the definition, as they cannot spread independently. User always need to download the application somewhere and install it themself. The closest analogue would be Remote Access Trojan (RAT) infected Windows PC pirated software.

This was very true in Japan where most of phone supported limited internet access even in 2001. Careers ordered phone manufacturer to what to develop, and most phones have zero compatibility for other career feature because every phones were locked.

It was half true in the mid 2000s in Australia. While you could put basically any device on to the network, Telstra and Optus definitely held a some sway over manufacturers.

There are different markets than US and phones aren't only sold by carriers. Even admitting what you say in US is true, it definitely was not in Europe or Asia thus I doubt any of your analysis of carriers impacting the tech.

It was a huge problem in Japan, South Korea and to lesser extent, the US. Japanese phone industry conformed to it but bled to death soon after iPhone happened.

You obviously didn't live at this age or were too young. Most people in the late 90s and the early 2000s got their phones subsidized by a carrier. There was a time when the carriers were the most important channel for selling phones and this was not only in the US.

I love your post.

NB tho, Symbian already knew (see 1998) how to make a "truly excellent mobile GUI" - sure, it was a PDA, but we were already making phone UIs with phone licensees then, and could happily have made a "well-designed high-level UI framework".

I'd've loved that "parallel universe".

Alas, the Symbian deal nixed it, and our new owners simply wanted us to make their kind of phone platforms for them: four whole new ones, in 18 months (heh right), mainly to their specs, while recruiting 100 new devs and 10 new designers.

So I feel it's a little harsh to say "the biggest failing of Symbian was to disown the UI" -- we never had the option to keep control of it. (It was a major failing of the Symbian deal itself, sure.)

Especially if it's something you used and liked. But yes, I'm a big fan of releasing these things to the public, for archival and for the curious. It's crazy how even big companies lose source code of old software, or at least only have it accessible in a completely chaotic state, with a missing build system, missing instructions, etc. Recent example would be 3D Movie Maker by Microsoft. People got it to compile eventually, but it apparently was a wild undertaking:

https://www.youtube.com/watch?v=gqXTzlDZmhU https://www.youtube.com/watch?v=mjtKHPwTWsc

Quite a few of those hours were wasted and contributed to Nokia's demise. They should have switched away from Symbian waaay sooner. I think clinging to it is a classic case of the sunken cost fallacy. By ~2005-2006 it was obvious, at least for some of us, developers, that continuing to invest in Symbian is a huge mistake.

By then, Nokia had been working on a Linux based OS for years. (That would become Meego.) But, if I'm not mistaken, at least a thousand people were working on Symbian and Symbian based products. So switching away from it was simply out of question. It would have been too bold a move. (Especially, since Nokia was dominating the market and these products brought in a very nice profit.)

wasnt this pain the premise of the infamous "burning platform" memo [1] and the eventual switch to Windows Phone as an OS?

[1] https://www.alexanderjarvis.com/memo-nokia-ceo-stephen-elops...

Maybe. But by then (2011 IIRC) it was way too late, sales had been going South, so one wouldn't need too much of a genius to figure out that Symbian was dead.

The problem was that in the mid 2000s things were going just too good and management didn't want to see how it could become problematic, or at least take the risk. We were selling smart phones, that weren't really phones anymore (heck, we called them mobile terminals!) but it was pretty hard and painful to develop software for these. This made hard to convince external, indie/hobbist developers (they ended up creating most of the apps for ios and android) and, of course made internal development pretty expensive. Worse, it must have made experimenting and thus innovation slow. Hence the UI was stuck. Not as if Nokia ever put too much effort into innovating the UI. Touch screens were somewhat frowned upon. Though that may in part have related to the existing UI, the effort needed to try out something new and the fact that those touch screens were still resistive back then. Which is really not a great UX.

Though I remember once trying a prototype phone at an internal conference with a touch screen providing haptic feedback (through some piezo efect). So you could kind of feel the on-screen buttons.

That memo made many people angry, I was at Espoo location the week after and not many happy faces around.

The Symbian teams had finally managed to bring app developers along with platform improvements on Carbide, Qt integration, better POSIX support via PIPS, Symbian was finally looking more modern and then management dropped the bomb.

Naturally it was the last drop for many of those small app developers that had enough of the Symbian SDK changes from the last set of years, now being asked to drop C++ and Java, and rewrite everything into Silverlight/XNA for WP7.

Instead of adopting WP7, most decided it was time to go elsewhere.

Its such a shame too, because Meego and the UI they built for the Nokia N9 was still ahead of it's time, and absolutely excellent to use. But it was dead on arrival, aside from us geeks who loved the idea of Debian on a phone :)

I vaguely recall Symbian being adopted/developed(?) by Nokia. It's incredible how early Symbian was, but what's most remarkable is that it wasn't even Nokia's only shot at a smartphone OS, they also had Maemo.

It's a shame Nokia is now throwing away everything they know about Symbian. The new nokia OS is shockingly crap. Some functions straight out don't work and other functions such as T9 text are awful. It's a shame when you realise that their old Symbian phones work perfectly well and there was no need to change it.

The even older DCT3 generation of Nokia phones worked much better than Symbian ever did. Nokia 6210 is still a great phone if you only need to call and receive SMS. But anything with a color display from that era is really showing its age by now.

Nokia hasn't been making phones for a long time. They license the name for HMD Global and the modern devices have nothing to do with Nokia itself.

Nokia earned billions from this code. I bet those hours were paid back many times over.

No, they probably didn't earn billions from that code. Their biggest cash cows were feature phones (non-Symbian) that sold 10-50 times as many as Symbian phones and were developed with many times smaller budgets. While some Symbian phones were pretty good for their area, there were probably at least as many failed projects as successful ones.

I don't have any data to back this up, but I think they sold hundreds of millions of high end (and expensive for the time) smartphones running Symbian. Remember the N line? The E line? Communicators? All running Symbian, all bestsellers for years.

No, I remember attending parties when 1 or 2 millions of a model had been sold. At the same time Nokia sold over 100 million of feature phones a year.

Edit:

As you write they were expensive. So while Nokia was the undisputd leader in the segment for a couple of years the absolute numbers were not that big. I remember some year the worldwide phone market was 400 million a year. Nokia had over 100 million of them, but smart phones (Symbian) only single digit millions.

One model, N95, sold 10 million units. It was priced at around $500. That's 5 billion revenue from one model alone. I'm sure feature phones sold a lot more, but that's irrelevant.

Yes, 10 millions for the N95 sound possible to me. But that was after 10 years of development with many 1000 developers. Revenue is not earnings. The orginal claim was that they earned billions.

Just before the iPhone, I think Nokia were shipping about 70 million Symbian phones a 1/4 and had about 50% of the market profit. Then, it didn’t

Indeed. Every time someone builds a walled garden, effort is duplicated needlessly, both in building it, and in using it.

It is the destiny for most code written.

Anyone who has been in this business for any length of time will look back at a long legacy of retired, obsolete projects. Very little in this space lasts. The high profile, long lived projects we see are just an tiny fraction of the software that's created, used, forgotten, and deleted.

I like to look back and note some of the companies I helped in that little window of time I could contribute, but those contributions are long gone.

We don't build bridges in this business, thats for sure.

I have friends who ask me how to set their kids on a path to a tech career like mine. I tell them they may not want that at all. This is a good job, today. But I don't think it is so different from building pickup trucks in Detroit in the 1960s. Nothing lasts forever. Including those trucks, and those jobs. We can't know the future, only that change is inevitable.

Software Manufacturing is a term that best captures my impression of the industry today.

Not really... it had its use. Code vanishes so quickly.

im still using workbench.

Reply on news.ycombinator.com