Aside from being a fairly cool -- if a bit undercooked -- new browser, Chome has another lesson for us. Google has very specifically NOT created an incremental bloatware version of existing code. They've built something new from nearly the ground up, embedding the proven technologies they needed -- javascript by V8, and rendering by Webkit. Only by starting from scratch were they able to get some of the size, memory management, and security enhancements that form the core of their promising new application.
I think Google's move here may just start a trend. The pendulum of software investment decisions has been moving for too long toward the ever increasing bloat. Vendors add on, extended, and graft on new features. Object oriented programming means you just subclass or overload without ever getting rid of anything. The down side to all this object extension is that it can be almost impossible to ever change direction once you've got your code code built.
The evidence for this is easy to find. Look at Websphere, Look at Vista, Look at Notes on Eclipse. All are powerful tools. All have a huge development effort associated with them. All suffer from a massive amount of legacy code embedded with little hope of really removing it.
What will be the next major software application to see the big do-over rewrite?
Comment Entry |
Please wait while your document is saved.
follows that ground-up is always the right way to go. The trick, IMNSHO, is to
have the underlying bloatware be more modular. Linux distros seem to be pretty
good at this so far -- giving the individual user a great deal of flexibility
in how far they want to go between barebones TTY consoles and
every-FOSS-project-known-to-man being installed and launched at startup.