When comparing Tornado vs Bottle, the Slant community recommends Bottle for most people. In the question“What are the best general-purpose Python web frameworks usable in production sites?” Bottle is ranked 4th while Tornado is ranked 7th. The most important reason people chose Bottle is:
Being a small one file distribution it includes almost every vital thing you need to support little websites (routing, templating). Everything else can be implemented using plugins.
Ranked in these QuestionsQuestion Ranking
Pros
Pro Designed for performance
Tornado can scale to several open connections at the same time. This makes it perfect for applications where performance is very important.
Pro Amazing asynchronous functionalities
Pro Flexible
Being a small one file distribution it includes almost every vital thing you need to support little websites (routing, templating). Everything else can be implemented using plugins.
Pro Single-file distribution
Bottle works around the one-file approach, everything is done in a bottle.py
file. This means that it's extremely easy to share and upload your application since it practically is just one python file.
Pro No need to install
It is so little there's no need to install, it is included in the standard libs python.
Pro Async, *let friendly
Using it with gevent is a breeze. It's a WSGI app so it's easy to make it work with anything.
Pro Truly magnificent
Cons
Con It's usually useful only when building real-time web applications
Because it's basically a framework for building real-time applications it has great support for things like WebScokets and uses non-blocking I/O. But it lacks in features that are used to build general-purpose websites like blogs, e-commerce sites, web portals etc...
For example it lacks ORM integration (at least out of the box) and it doesn't have a very powerful template engine.
Con Small community. Difficult to find online docs and examples
Con Very hard to develop projects that are not smaller than 1000 lines
While Bottle is a great framework for building small applications (generally less than 1000 lines of code), it starts getting very hard to manage your application if you want to go even a bit larger than that.
The fact that it follows a single-file distribution model and that it's missing something like Flask's blueprints only make this problem worse.