“Building secure Clojure web applications needs to be easier, and requires integrated security frameworks - not standalone libraries!” – John P. Hackworth, Clojure web security is worse than you think
An integrated security system for Clojure applications based on a set of components written to the specifications of Stuart Sierra's component.
Bolt has an ambitious goal: to separate all security-related concerns from Clojure-based applications so that they can be implemented as pluggable components.
The precise meanings of the terms component, system-map and system are those in component. In summary, a component is a map of data, usually implemented as a record with associated protocols specifying functions for start/stop and others. A system is a set of these components, with the inclusion of declared dependency references into each component.
In addition, bolt uses the following terms :-
- username - a user's short identifier, for example: bob
- email - a user's email address
- user - a map, containing entries that distinguish and describe a user
Bolt provides an integrated system of components, rather than requiring developers to roll their own from smaller libraries.
Functionality can be customised by interchanging components, providing necessary flexibility for bespoke Clojure applications.
Nevertheless, 'out-of-the-box' defaults should provide good security, on par with other languages and frameworks. That is what is currently missing in the Clojure landscape and the gap that bolt aims to fill.
The key difference between bolt and Friend is that bolt is designed for use with Component based applications.
Bolt is designed specifically for modular applications, where functionality can be added through the addition of extra components.
Stuart Sierra's component library provides a balanced, elegant and "essential" foundation for bringing all these parts together into a single system, so it's a natural fit for this problem. It is also straight-forward to decompose (and therefore reason about) the system (by understanding the role that each component plays). This is an important property of any security system - if the design is difficult to comprehend but 'just works' or works 'like magic' then it limits the number of people who can understand it and point out potential weaknesses.
To provide flexibility, bolt fully embraces and consistently adopts protocol polymorphism within Clojure, enabled by Stuart's approach. This will not to everyone's taste. Alternatives, such as the use of dynamic vars, are wholly avoided. Functional programming is a beautiful thing in the small, but presents practical challenges at scale. Polymorphism is one of the cornerstones of object orientation worth stealing.
Ultimately, whether bolt is right for you will depend on how you build your Clojure web applications. For smaller applications with a single set of Compojure routes, Friend is a better choice.
For larger applications, especially those with multiple modules and using Liberator or yada to provide a fuller REST API, bolt should be a good fit.
Bolt was formerly named Cylon.
Bolt is not suitable for production systems until it reaches version 1.0, which will indicate that bolt has been deployed into production elsewhere and has undergone thorough peer review.
Join our Google group bolt-discuss@googlegroups.com for discussion about how to improve bolt.
https://hackworth.be/2014/03/26/clojure-web-security-is-worse-than-you-think/ https://github.com/dhruvchandna/ring-secure-headers https://github.com/weavejester/ring-anti-forgery
Aaron Bedra's seminal ClojureWest talk in 2014 – http://www.youtube.com/watch?v=CBL59w7fXw4 - this was the inspiration behind bolt.
Mastodon C for sponsoring the development on bolt, and using it in their kixi projects kixi.hecuba and kixi.stentor
Also, to Neale Swinnerton @sw1nn for the original work in adopting Stuart's component library and showing how to migrate Jig components to it.
Yodit Stanton and the rest of the opensensors.io team for putting up with the original Cylon updates (Cylon being the former name of bolt) and being the first adopters of the OAuth2 features.
Juan Antonio Ruz for designing and developing the TOTP two-factor authentication support. Additionally Juan conducted the background research and co-authored the OAuth2 support, and many other aspects of the project. Also for providing a public example of how to use bolt.
Martin Trojer and others from JUXT for a continual stream of thought-provoking ideas and good advice.
Andrey Antukh for suggestions about integration with Buddy.
The MIT License (MIT)
Copyright © 2014 JUXT LTD.
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.