Hacker News Reader

Comment by zeveb | Hacker News Reader

An interesting idea, but I'm concerned about the privacy implications. Does it send all 404 URLs to the Wayback Machine to see if it has a copy? That reveals one's browsing habits without one affirmatively doing anything. Does it strip URL query parameters? These are commonly used for session IDs and auth tokens. Does it strip URL path parameters? Conceivably these might have private information in them as well. For that matter, paths themselves may contain private information.

If all it does is offer a link to the Wayback Machine, then it sounds great. But I worry that folks might want it to do more, without realising what that might mean.

View on HN

The No More 404s experiment, like all of Test Pilot, is totally opt-in. The whole idea is to let us try things with Firefox, get feedback and iterate quickly.

No More 404s Telemetry ping is only gathering data about how often the Add-on is fired and how often it is clicked. We (Mozilla) don't know anything about URLs. The Wayback Machine has it's own privacy protections in place that you can learn about here: https://blog.archive.org/2013/10/25/reader-privacy-at-the-in...

Also worth noting, every Test Pilot experiment comes with a brief explainer of all data collections. Here's the explainer for No More 404s:

In addition to the data collected by all Test Pilot experiments, here are the key things you should know about what is happening when you use No More 404s:

* We collect basic usage on how many times you encounter a Page Not Found error (code 404), how many times a cached version of that page exists from Archive.org, and how many times you choose to view the cached version.

* To provide cached versions of pages, we send 404 error page URLs to Archive.org. Archive.org discloses its privacy policy here (https://archive.org/about/terms.php).

* We do not collect URLs of the pages you request or the URLs we send to Archive.org.

* We may share survey results you submit to us and aggregated telemetry data related to this experiment with the Internet Archive.

> To provide cached versions of pages, we send 404 error page URLs to Archive.org.

Why not only do that on a user action, instead of on every single 404?

I think that's exactly what the text you quoted is saying. When you actually request the cached page, it sends the URL to Archive.org.

It could also legitimately mean that they send every 404 page URL to archive.org — that's why I'm asking for clarification.

The article (and the screenshot) seem to imply that it doesn't actually check with the Wayback Machine to see if the URL exists. It only displays a prompt telling you that you can try to view it there. So theoretically it would only send the URL to the Wayback Machine if you chose to use its service.

Reply on news.ycombinator.com