I would like to touch on the technology, understanding the work of which will increase the quality of anonymity in the network, especially when working for one anonymous browser identification (DS, Antic, etc.) – Fingerprintjs.The essence of the process is assigning a unique identifier to the browser, give a name to each user in the network. Some publications have expressed the opinion that this is the end of anonymity in the network, but not for us)
- Tracking individual users, what time they came, what they did before I came
- Personal advertising, just through user identification
- Powerful sites use internal Analytics through the use of a fingerprint to increase the quality of their work
The final task of the technology is to build a map of user actions. In a simple format, you could use a Cookie if the user the first time it comes to the site, it is assigned a number. If it was already there, the shop sees the ID that the fingerprint identified before. But in our time, Cookies can be deleted and the user becomes anonymous again, for example, incognito mode, which leaves no traces. How programmers solved the problems:
- the Evercookie project does not delete cookies, it does not just store information in a single storage, but uses a number of data
- flash cookies-previously not cleared, now this problem has been solved
- silverlight cookies – this is a special dedicated space on the user’s hdd, this data can not be deleted by clearing Cookies
And finally png cookies-the browser returns an image that contains encoded information, such as an ID, and then on a subsequent visit this information is restored even if the user has cleared the Cookie. Evercookie uses all available storage browser. For an ordinary user, it is almost impossible to delete all the data, But at the same time, Evercookie does not work in incognito browser mode. Accordingly, we came up with The fingerprintJS technology. FingerprintJS is a small library that tries to solve this problem.
Advantages (in our case disadvantages):
- Works in incognito mode
- Does not use the storage on the HDD
Used by baidu (6-7% of all popular sites use this technology).
How it works:
It polls the user’s browser for all specific and unique data and collects it in a single line, getting an ID (individual key):
- UserAgent is read first
- then the browser language
- time zone
- screen size and color depth
- reads the list of supported html5 technologies, all possible
- User-specific and platform-specific data is queried (doNotTrack setting, processor class, platform)
There are two more ways – it is to determine the list of all plugins (different versions, types, their number) + Canvas FingerPrint is used the text image is drawn hidden from the user’s eyes. it will be different on different machines a byte array, a strong technology for separating users, is also attached to the resulting fingerprint (string). That is everyone is going data is run through the hashing function and a common string is obtained – our ID. The actual detection accuracy is about 90-91%.