Source: http://news.ycombinator.com/item?id=5245576

The Opera Software I knew and loved was a technology company filled with 
some of the most brilliant engineers I have ever met. Despite being a 
tiny company halfway across the world from where the real action 
happened, it managed to conjure up some amazing stuff.

The financial margins were slim, the salaries crappy, and the the roof 
of our derelict office building was literally leaking. But that was 
_fine_, as we were all in this together. Even the new guy received token 
stock options, owning a tiny share of something big.

The company culture was egalitarian. Decisions were not made in some 
ivory tower, but the CEO himself would roam the hallways, discussing 
with individual engineers. Even as the company passed 500 employees, we 
managed to retain very much of the startup culture.

The company had its share of problems, of course. Problems that, if not 
solved, would doom the company long term. Opera tried to be all things 
to all people, completely lacking focus. It allowed itself to be pushed 
around by OEMs that didn't understand the browser game. Its greatest 
strength was that the engine was portable to any platform, any device. 
But this was gained at a huge penalty to the development speed and 
agility.

With the lacking focus, Opera also failed to understand the importance 
of design. Having engineers in charge is great. Having engineers in 
charge of _design_ is a terrible, terrible idea. Opera was a great 
engine - with a very mediocre UI.

However, most worrying of all was that Opera grew complacent from having 
unsurpassed standards support, the fastest rendering engine, and so on. 
When things got difficult, one would blame Microsoft, as if taking the 
moral high ground would make any difference.

The decline and fall of Opera has been a gradual process. Just praising 
the old regime and blaming the new is too easy, too simple.

As the competition came back to the game around 2003-2004, Opera wasn't 
too slow to react. It didn't react at all. We should have thrown out 
that ugly ad-banner instantly. Dropped the old paid model the moment 
Phoenix came into existence. We should have made the project open source 
the moment Safari/WebKit became public. Or even better: before.

I was just a 20-yearold brat when I joined in 2004, fresh out of high 
school. I don't blame management for not listening to me then. But I'm 
sad that they didn't figure this stuff out themselves.

From there on, we were playing the catch-up game half the time, and 
pushing the boundaries of the Web the other half. I remember the rush to 
get XHR ready to ship just after Google Maps made use of it in 2005, 
making our browser look pretty bad. But I also remember the scramble to 
support Acid2 (2006) first, the rush to remain on top of the layout 
performance benchmarks - and the crazy experiments we did. Opera 
Platform, which was essentially Firefox OS many years ahead of its time, 
was pretty cool.

The platform work we did was amazing, too. We had the full web available 
on the most crappy devices you can imagine. The release of the Motorola 
A1200 (2005) was one of my proudest achievements. As was the fantastic 
adaptive zoom we made for the Wii (2006). The iPhone, made public a 
little later, had the exact same feature as we had come up with.

Apple and Mozilla's products, at this time, struggled with the same site 
compatibility issues as we did in this Microsoft era. But unlike us, 
they were focused on shipping one thing well. Their products became nice 
and pretty, while ours looked like something out of a hobbyist shed.

I believe our engine was _better_ than the competition for a very long 
time, but we kept shooting ourselves in the foot. Over and over again. 
And by this time, the competition was picking up speed. They were 
loading the big guns, while we were not.

Even as the money started coming into the company in 2006-2009, we 
failed to make the necessary investments in our future. I became 
department manager in 2007, and remember losing people over petty salary 
issues. Minor to the company, but significant to the employee. Some of 
the people I lost I had to hire two new guys to replace - and train for 
a year before they could really step up. Man-years of wasted 
productivity, thanks to the holy quarterly numbers.

Not to mention the failed projects we poured resources into, draining 
all departments for the people they needed just to _catch_up_ to the 
competition. (Opera employees will know what I'm referring to. :)

Or the fact that when the testing systems I was in charge of became a 
major bottleneck of our development process, I had to scavenge discarded 
hardware from the trash to speed it up. I don't even want to know how 
many developer hours were wasted waiting for test results. I offered to 
have my recruitment budget cut to zero if I could just have the damned 
servers I needed to do my job, but burning out my team was apparently a 
better option.

These were just some of the issues I could see from where I was 
standing. Elsewhere in the organization, there were similar issues. Many 
of these things improved over the years, but it was too little, too 
late.

And as the old regime was slowly replaced by the new, they started 
making new mistakes. The way the 2008 reorganization was handled was 
atrocious. I kept my mouth shut then, to keep my department from 
worrying. The 2010 layoffs is a story of its own.

It eventually dawned on me that upper management didn't actually give a 
shit about the rank-and-file employees. We had become replaceable cogs 
in the machine. They abused the startup culture for a long while. 
Despite good margins, salaries were kept crappy, playing on the 
engineers' idealistic love for what they were doing. The token options 
were soon gone, too. The leftover money turned into impressive bonuses 
for the selected few.

We knew from the start that Opera Mini (2005) was a transitional 
technology, to be made obsolete as devices became more powerful. But 
there was no plan then, and I doubt there is any sensible plan now, for 
how to remain relevant after that era is over. We started with the best, 
most portable engine, and were surpassed by the competition while we 
were asleep.

The desktop browser has remained stagnant for years. We kept shooting 
ourselves in the foot with rushed, buggy releases. The same, crappy UI. 
The wrong, new features - instead of what people actually wanted and 
needed.

At some point, the cumulative mistakes made doomed Opera as a technology 
company. I knew we could have turned things around in 2007, if we wanted 
to - and knew how. Doing the same in 2009-2010 would have been very, 
very difficult, but possible.

By 2011, I knew it was game over and left.

You see, I'm not angry about the switch to WebKit. I'm disappointed that 
due to years of mistakes, this has become inevitable. Saying "18 years 
of development has been futile - our codebase is worthless, and will be 
dropped" is no victory. It's surrender. We lost, and it's our own fault. 
Stop blaming others - we brought this onto ourselves.

I'm not surprised about the layoffs. I'm angry that my friends and old 
co-workers, who I care a lot about, are treated like crap. The last half 
of the layoffs, mentioned in the Q4 report, seem to have been handled 
moderately well. The first half (affecting Core) seemed random and 
arbitrary, lasting weeks - and nobody knew who was next. Suddenly, 
without warning, the guy next to you - who had been there for a decade - 
would be leaving. WTF was going on?

And that's just the Scandiavian offices. Tokyo has suffered four rounds 
of this shit, without the benefits of Norwegian labour law.

I'm sad that the influential, aggressive Opera Software that used to 
push progressive agendas in standards bodies has become irrelevant. Its 
influence stemmed from having its own realm, its own voice. Now it will 
be playing third violin in the WebKit orchestra, following the lead of 
Google and Apple. Brilliant core engineers will be making a skin for 
Chromium, wasting precious talent.

I spent seven years of my life on this project. Now it's gone.