On this page you will find explanations regarding the repository organisation (especially categorisation), the selection process, and the rating and popularity calculations.

Choice of Categories

To organise the repository, we started by making a list of categories that would be present. It was essential to make choices, as open source is so wide. We have kept all areas relevant for the company, and where Smile has already implemented projects and has a real expertise.

We divided over forty categories into three "dimensions"

• Infrastructure, for example: firewalls, VPN, monitoring, virtualisation, operating systems, http accelerators, etc.
• Development and intermediate layers, for example: corporate directories, databases, ESB, web and mobile frameworks, search engines, MOM and EAI, etc.
• Applications, for example: CRM, business intelligence tools, CMS, document management tools, portals, e-commerce solutions, etc.

Solution selection

For each category, we called upon our referral technical, functional and business experts to help choose the best solutions, those on which any business can rely on to build with confidence, their most ambitious projects.

We tried to be as complete as possible. We have identified more than 300 open source tools, providing a wide choice that will allow you to build the most relevant architectures.

Rating calculation

Every solution has a global rating (if ratings have been submitted by visitors and by Smile) out of 5.

This global rating corresponds to the average of the Smile rating and the visitor rating (global rating=  [Smile rating + rating grade] / 2).

The visitor rating corresponds to the average of all the ratings submitted by visitors (Visitor rating = [rating n1 + rating n2 + rating n3 + ... rating Nn] / n).

Finally, every rating (be it from Smile or a visitor) is out of a total of 5 points (ranging from 1 to 5) and is an average of ratings for 6 criterias (noted on a scale of 1 to 5): Reputation, dynamic, quality of technical base, functional scope, extensibility, available resources; (rating = [criteria 1 + criteria 2 + criteria 3 + criteria 4 + criteria 5 + criteria 6] / 6). 

Popularity calculation

On each page, we present the popularity of the selected solution in relation to its category and the entire site. We therefore have two values. These calculations are based on the number web page views by the visitors on the solutions page (the number is limited to counting one view per visit / Google-type engines are excluded from the statistics)

To calculate the "popularity on the website", we compare the page views for the solutions in question with the average page view for all the solutions on the site. If a product has the same number of views as the overall average for all solutions in the repository, it gets 50%. Then the percentage of popularity varies proportionally.

To calculate the "popularity in the section", the method is the same but the number of page views of the product in question is compared to the average page view of all products in the category (eg Drupal, CMS).

Google Pagerank calculation(PR)

To calculate the Google Pagerank of a website for a solution, we use the API provided by Google and we audit the main URL of the product website. The PR value of a site is not the final indicator, but provides a good indication of the popularity of a website (visits, number of inbound links, influence, ...).

Retrieving information from Ohloh

From the API provided by Ohloh (open source directory connected to the solutions repository), we retrieve information such as Ohloh user ratings (also out of 5), the number of votes, the number of lines of code for the solution, the number of contributors over the last 12 months, programming languages ​​detected, the address for downloading the tool and the link to the complete Ohloh sheet. As such Ohloh and the Open Source Guide are complementary.