3

A lot of times I notice a link in Hot Network Questions, or Linked Questions, or Related Questions, or the regular question list, that looks familiar but isn't marked as visited in my browser. Usually when I go to it, it is in fact familiar, in that I already visited it but with a URL that ends in a different source-tracking query parameter, such as lq=1.

Example:
Screenshot

Is there a script or other program to automatically mark (most of*) these links as visited?

*There's also share URLs with their trailing user IDs and missing slugs. That's a bit of a different issue, but still worth including if possible.

6
  • The only way a userscript could do this is to keep track of every visited URL you browse on SE sites! That could get real messy, real fast. ... But a proper browser extension (Add-on) could leverage the protected history API. Both approaches would have to make judgements about what URL-parameters were equivalent though. For example, this and this are obviously the same, but this is not the same at all. Commented Jan 19, 2015 at 22:33
  • @BrockAdams: I had a few ideas about clever trimming to get the equivalent URL and then monkeypatching in the right thing when you click, but that would be fragile and a bit hard to get right. (Like most clever solutions.) Commented Jan 19, 2015 at 22:36
  • I forgot about that approach. Since you aren't trying to fool the user, or clickjack for revenue or otherwise be Evil (like that site that rhymes with "Google"), that approach may actually be relatively easy. Still have the judgement issues though. Just trimming links is sometimes a disservice. Commented Jan 19, 2015 at 22:41
  • Oops, never mind. That above approach won't work, because the URL in the history-list will still not match the displayed URL unless you perma-sanitize all links (messing with SE's tracking and statistics). Commented Jan 19, 2015 at 22:45
  • @BrockAdams: There might still be a solution to that (silently double-opening the link in a hidden iframe?), but like I said: fragile. Commented Jan 19, 2015 at 22:51
  • @BrockAdams: I think the simplest solution to that problem would be to strip the tracking parameters from the URL as soon as the page loads, as if it had been redirected. Do that with @run-at document-start and you shouldn't even have much extra latency. I've thought of maybe adding this to SOUP at some point. Commented Feb 3, 2015 at 14:24

0

You must log in to answer this question.

Browse other questions tagged .