Difference between revisions of "OWASP Passfault"

Demo Site

Does the Demo Site capture or log passwords?

No, of course not

How can I be sure the Demo Site doesn't capture or log passwords?

You can't, There is no way to verify what is uploaded to appspot (google is hosting the demo site) However, you can look at the code: https://github.com/c-a-m/passfault/blob/master/jsonService/src/main/java/org/owasp/passfault/web/PassfaultServlet.java We took the following steps to ensure the passwords don't get logged:

• GETs are blocked so no urls will have accidental passwords stored in the logs
• passwords are read directly from the input stream to prevent parsing into Java Strings
• the memory is cleared as soon as analysis is complete.
• HTTPS is required on this URL (using the appspot domain)

To be extra cautious, download the code and execute it locally. (See the readme) https://github.com/c-a-m/passfault/blob/master/README.txt

Why do you need to pass the password over the wire?! Isn't that insecure? Others do it in javascript-client why don't you?

Passfault's mission is to replace password policies, not just be a cute strength meter. It was intended to be used by the sites that already take and use your passwords. With that use-case, adding Passfault doesn't lessen security in any way. Plus it has the added benefit that we can store LOTS of password lists and do some in-depth analysis that can't be done client-side.

Does 2FA (two-factor authentication) make Passfault obsolete?

No, not if one of the factors is password authentication. 2FA lessens the risk of passwords, but if you no longer care about password security then you shouldn't use passwords for authentication at all. If you still use passwords, you should have an effective password policy.

Some argue that password policies make passwords less secure, does that apply to Passfault?

Researchers have found that, of all techniques used by password policies, only the required length had an effect on the overall strength, and even that claim is dubious. Passfault works different and we claim we can do better, a lot better, than any traditional password policy.

How does Passfault compare to zxcvbn?

Passfault is very similar to zxcvbn in it's approach to password analysis. In fact it is the only comparable tool that we know of, and the only alternative we endorse. zxcvbn presents the strength in units of "entropy", this measurement could be derived from passfault's "pattern size", however we feel that the "time-to-crack" help convey to the end user it's real risk (the downside to this is that really large numbers don't mean much to users, 10 years, or 10 million years, still feels like a long way away. Entropy is logarithmic so it shows this better. However entropy units are not intuitive to users.). We also search for a few more patterns that we think are valuable.

Why java - I hate java. If it were only in language x I'd use it.

Passfault is packaged up in docker as a microservice. You'd probably want to run passfault as a microservice anyway, so forget that it's java and just run with it. That said, we welcome any ports to typescript or any other language.

Discuss with us more on twitter [[1]] or join the email list: [[2]]

Volunteers

OWASP Passfault is developed by a worldwide team of volunteers. The primary contributors to date have been:

• Cam Morris
• Bernardo Araujo Rodrigues
• Ray Stone
• New Jersey Institute of Technology students contributed to release 0.8 (Highlander):
  • Michael Glassman
  • Georgina Matias
  • Scott Sands
  • Brandon Lyew
  • Kevin Sealy
  • Llina Ljoljevski
• University of Florida Students contibuted to release 0.7 (Gator):
  • Neeti Pathak
  • Carlos Vasquez
  • Chelsea Metcalf
  • Yang Ou

Others

• Partnet Inc. has donated paid labor on OWASP Passfault
• JetBrains has donated professional licenses for IntelliJ IDEA. If you are developing on OWASP Passfault contact the project leader and be sure to get a license!

• Join the OWASP Passfault Mailing list
• See the Roadmap
• Peruse the open issues
• Fork the code on github.
• If contribute significantly to the project contact the project leader for a professional license of IntelliJ IDEA by JetBrains (Thanks JetBrains!)

Release 0.8

Goal: preparation for ESAPI

• More meaningful word lists
  • Frequency lists: build lists of the most common words, names. (Done for English, Spainish)
  • Improved configuration of finders and wordlists

Release 0.9

• UI improvements
• Fix backlog of issues
• experiment with configuration of wordlists

Release 1.0

Goals: Enterprise Ready - UI improvements for learning better password strategies - Easier to configure and run, not requiring a developer to wire things up.

Other Important Goals

• Javascript library generated by GWT and GWT Export. Do you know GWT? Please help us build a javascript version of passfault using GWT Exporter: https://code.google.com/p/gwt-exporter/
• OS system integration:
  • Linux
    • running passwd on linux runs passfault
    • apt-get install passfault
  • Windows
• Document each pattern finder on the OWASP wiki.
• JQuery Plugin: A JQuery plugin that will let a web site use either the passfault applet or a passfault JSON Service to analyze a password.
• Wordlists: We can always use better word lists. Contact us on the mailing list if you want to help.

For current bugs and smaller tasks see the issues list on github: https://github.com/c-a-m/passfault/issues?state=open