Code reuse cuts the development time and brings down the defect rate. So it should be a good thing. Quite the opposite approach taken by the Mozilla guys. Those guys alsmost built their own everything (rendering engine, foundation classes, etc). There are a lot of open source libs there also, but i guess google wanted to focus on the minimal UI design and the per process architecture.
Also, consider when the original codebases were written. There was nothing to borrow when Netscape/Mozilla first took a crack at it, in many of these cases.
The hard parts of writing a web browser are the rendering engine, the javascript engine, the transport layer and security code. The rest, comparatively, is trivial and there are cross-platform OSS libs for all of those.
etal got it right though -- Mozilla began as an OSS project in 1998, long before many of those libs were available.
I disagree with such opposing view. 25 libraries would be a nightmare to get in just the right version, and specifically compiling a program with many libraries that aren't already compiled in Windows is usually a nightmare (at least with VC++, which I guess is what these guys are using together with Scons?), to compile the software. The idea that Chrome would automatically update in case one of such libraries changed, is probably not very convenient either as such updates could also break the code (because of changes in the interfaces for example, or in the behavior of libraries... while it's not very probable for each library, it becomes quite probable when you have 20+ libraries).
It does not use Python directly. Python is used for testing various parts of Chrome, such as code coverage, dependencies, measures page load times, compares generated html, etc.
