nostalgia_chrome

Cross-platform Chrome History Analysis

Self tracking

There is a movement of self tracking. Monitoring pulse, heartbeat and so on. But the most important is not being tracked: our online behavior.

Making sure we can self document, we need the following things.

1. Chrome only keeps its history for a max of 90 days, so we need to start saving history.

2. We need to collect HTML data from the pages we visit.

3. We need to extract and analyze data from the HTML, such as code snippets, links, microdata, images, events.. anything really. This is done in Nostalgia Core.

4. Allow plugins (and make them configurable, please contribute). The first example is that it will additionally track which videos you watch.

What can you expect (Data overview)

In ~/nostalgia_data/meta.jsonl an index will be saved per visit:

{
  "path":"/home/pascal/nostalgia_data/html/1576317113.7_httpsgithubcomnostalgiadevnostalgia_chrome.html.gz",
  "url": "https://github.com/nostalgia-dev/nostalgia_chrome",
  "time":"1576317113.75019"
}

In ~/nostalgia_data/html the source HTML will be stored as .html.gz (reaching about 8x compression).

In ~/nostalgia_data/videos_watched.jsonl the data for events on HTML5 video elements will be stored (on stop playing/close tab):

{
  "playingSince": 1576273573.08,
  "seekTime": 0,
  "playingUntil": 1576273599.977,
  "duration": 26.8970000744,
  "totalClipDuration": 3510.301,
  "pageLoadTime": 1576266470.316,
  "loc": "https://www.youtube.com/watch?v=Zz-bhLjVS5o",
  "title": "Lost Frequencies | Tomorrowland Mainstage 2019 (Full Set) - YouTube",
  "likes": 24137,
  "dislikes": 946
}

Installation

1. Clone this repository: git clone [email protected]:nostalgia-dev/nostalgia_chrome.git

2. In Chrome click the settings button and click "More tools" and navigate to "Extensions". Click "Load unpacked". Navigate to the chromePlugin folder and click "Open".

3. pip install nostalgia_chrome

4. To test it out, run nostalgia_chrome run_server. This will run the web server in the foreground so you can see that it works.

5. Visit a (non-file / localhost) URL so that you can verify it works. The data will be stored in ~/nostalgia_data/meta.jsonl, ~/nostalgia_data/html.

6. To make sure nostalgia_chrome gets automatically run on boot:

On Linux (systemctl based):

pip install sysdm
sysdm create "nostalgia_chrome run_server" --extensions ""

On Windows awaiting contribution for how to do this #2

On OSX awaiting contribution for how to do this #1