Summary

• SWISH stores HTTP logs and saved programs on the server. If the user is logged on, the connecting IP addresses and a basic profile is stored based on information provided by federated login and entered by the user.
• Stored information is used to remember user choices and restore sessions. The HTTP logs are used for analysis and debugging purposes.
• Public saved programs may be searched and retrieved by anyone. HTTP logs may be shared with research partners under a non-diclosure agreement. Other data is never shared.
• HTTP logs are deleted after 6 months. Saved programs are never deleted. User profiles may be deleted by the user when logged in.
• If you do not agree with these policies, disconnect now.

Details

Below we specify which information is stored where and how it may be used. Details depend on the installation. Notably login may or may not be supported and logging may or may not be enabled.

User preferences

Preferences such as semantic highlighting or default profile, etc. are stored in your browser's local store and passed along to the server when applicable.

User profiles

If the user logs on a basic profile is established from data provided by the identity service provider (ISP, e.g., Google) and details added by the user. The given name and avatar URL are used in the chat service and associated with saved programs. The email is used for user-enabled notifications only. This profile may be deleted at any time using the Profile menu associated with the login avatar.

Stored files

Saved files are stored on the server along with the IP address, a time stamp and the information provided by the user in the Save dialog. If the file is marked public, it may be searched for, opened and edited by anyone, honouring user selected access restrictions. Saving a file after editing creates a new version. All versions are kept and never deleted.

Cookies

If the user is logged in SWISH uses a session cookie to verify the login status.

HTTP logging

HTTP interaction is logged with full detail. This information is used to reproduce and fix stability issues. The data may be used for analysing the user interface and aspects of Prolog programming such as program evolution, debugging, etc. This information will not be made public. Server interaction includes:

• If semantic highlighting is used, the server maintains a mirror of your source while it is being edited.
• Queries posted
• The program against which a query was posted
• Interaction with the query such as asking for more answers and debugging interaction.