Source code

Comprehensive Study of Identity Management Systems

19 February 2020

This document presents a result of a comprehensive study of open source identity management systems. The study is unlike most other industry product comparisons. We have based this study on hard data and first-hand experience. We have downloaded, installed and configured every single evaluated product. And we have conducted a detailed source code analysis of every product. This is one of the unique aspects of open source: it is difficult to pretend or hide anything. The source code reveals everything.

The study has two parts
First part of the study compares project features. It is based on first-hand experience with each product. It summarizes and compared the products based on what they can do.
Source Code
The second part of the study analyzes source code of the products. We analyze not only the latest state of the source code. It also goes very deep into the source code history. It analyzes how the code was developed, how many developers actually worked on it, what is the actual development speed, etc.

The products evaluated in this study

The interesting fact is that although each part of the study was driven by a completely different methodology the results are similar. We consider this to be an indication that our results are correct. Another interesting fact is that the results of this study tell a completely different story than the studies conducted by business analysts. Therefore we have decided to follow an open source spirit and we are publising all the details of the study methodology including formulas, data sets and evaluation scripts. As the source code of all the evaluated products is already publicly available this means that anyone can follow our footsteps and re-do the study himself. Everybody can make sure that we have not cheated, that our model does not include any kind of magic constants or that it does not prefer one product over the other. We kindly leave the decision about trustworthiness of this study to the reader. Everybody can decide for himself whether to prefer the results of proprietary models and obscured data used by some industry analysts or whether to prefer completely open and transparent methodology used in this study.