I did something "amazing" on a website and my narcissistic side wants proof for posterity that I managed to actually do it, even if the website in question decides to change it's rules or goes off-line.
For example, what inspired me to ask this question was 100 year long streak on GitHub https://stackoverflow.com/questions/20099235/who-is-the-user-with-the-longest-streak-on-github/27742165 which could disappear without proof at any time since GitHub usually simply makes smaller modifications to their back-end without informing anyone. And screenshots that I upload myself are obviously trivial to falsify.
But I can see this having more serious applications, like serving as legal evidence that is simpler than subpoenaing server logs, or when those are inaccessible (e.g. foreign country). Although it has already been declined in court once.
If only there was a web service, in which I input an URL, click a button, and then it would immediately:
- fetch the HTML and all of its dependencies
- make the original webpage viewable on a public URL of the mirroring website. Storing a screenshot only instead of the original HTML would be acceptable, but not ideal.
- associate a timestamp and an URL to it, in a way that I couldn't change it afterwards. E.g., it could encode that information in the URL as
http://snapshot.com/view/2015-01-01/http://github.com, or simply add a custom header
Non-mandatory features that would be nice:
- Authentication. Not sure if technically feasible.
- send an arbitrary HTTP request instead of just a simple GET
- retrieve other media types HTML, in special PDF. archive.org does this.
robots.txtwhen I do a manual request. It's not cool to crawl those websites, but I don't see why not take a snapshot on request, which is just like me loading the page. Original motivation: prove that a website was at position N on a Google search: https://web.archive.org/web/https://www.google.com/search?q=x86+paging+tutorial archive.is currently does this: https://archive.is/vkc8B
- archive the page even if it gives 500. Can be used as proof that I've found a bug. archive.org does not seem to do this.
Of course, the entire system would hinge on the belief that of the web service in question is not cheating, which although not ideal, is acceptable to me.
What I have found so far:
- Wayback machine: http://archive.org/web/ . If only there was a button that I could press to generate a snapshot whenever I want it, it would solve my problem
- many websites that take a URL and convert it into an image, but none that stores the image on their server and assigns a timestamp and origin URL to it