So a third competitor appearing and taking the marketshare from both of them doesn't count? Such a great logic to get promotions. Of course it would come from google.
You are again thinking in terms of winning, and in literally the first sentence i wrote that we were not trying to win anything.
The goal was to get SF to not be a shitty monoculture.
Ideally by SF being forced to compete with us, and making their product less shitty. It does not require destroying or harming SF to achieve this goal.
They chose, ironically, not to compete, and were destroyed as a result of their own choice.
It also happened that third parties came along and helped make it both not shitty, and not a monoculture.
Goal achieved. Why does marketshare have to enter into any of it?
Nobody was trying to destroy anything.
The person who started the project (dibona) came from VA linux and slashdot, and was very good friends with the people who ran both (still is!).
He also started summer of code and lots of other things.
If you send troops to block the enemy from advancing, if those troops block the enemy from advancing, they've succeeded. Even if those troops didn't "win the war". Even if they all died in combat. The mission was a success.
If you want to say this is revisionist history from a googler... sure - make that case. But simply deploying a service to try to 1) prevent a competitor from gaining marketshare and/or 2) get the competitor to suck less... it's a valid move.
Personally, I don't think google code alone made much of an impact on SF directly, but google code and ms codeplex together probably did get people to start considering using something beyond SF.
Because the truth matters? it is quite annoying to see people trying to bend reality to make themselves or their projects more important than they really were. Google code sucked, like many google projects.
Github wasn't successful because Google Code made people "start considering using something beyond SF". Github succeeded because git is great and social network features allowed it to reach a much bigger audience.
Google code did not at all suck, compared to the alternatives - essentially just SourceForge! - at the time it launched.
A lot of what the OP wrote rings true to me, Github obviously hit on a better model, but they clearly had different goals.
I'm sympathetic to people who came into this landscape after Github was already around feeling like google code was a lame also-ran, but as someone who thought "why is everything hosted through this shitty SourceForge website" when I first started using open source, it was a huge improvement.
