Comment on AdNauseam is a uBlock fork that goes further: it actively attacks marketers by auto-clicking every ad before blocking

• lime@feddit.nu ⁨1⁩ ⁨month⁩ ago

  the part that’s safe is in the browser. it’s a basic fact of how http requests work that you can just request data and then not read it.

  also, “task managering the popups”? unless i’ve missed some very weird development that has literally never worked, because popup windows are part of the parent process.

  source

  • medgremlin@midwest.social ⁨1⁩ ⁨month⁩ ago

    Back on Windows 98 through XP, each individual window was a process that could be killed in Task Manager, and popups opened in a new window.

    source

    • lime@feddit.nu ⁨1⁩ ⁨month⁩ ago

      really? sounds like a weird span of systems considering they share so little code. i’d like to read on how they did that.

      source
• doodledup@lemmy.world ⁨1⁩ ⁨month⁩ ago

  What makes you think uBlock is safe without checking relevant code sections?

  source
• techt@lemmy.world ⁨5⁩ ⁨weeks⁩ ago

  Here you go, from the repo:

  function timeoutError(xhr) {
        return onVisitError.call(xhr, {
          type: 'timeout'
        });
      }
  
      const url = ad && ad.targetUrl, now = markActivity();
  
      // tell menu/vault we have a new attempt
      broadcast({
        what: 'adAttempt',
        ad: ad
      });
  
      if (xhr) {
  
        if (xhr.delegate.attemptedTs) {
  
          const elapsed = (now - xhr.delegate.attemptedTs);
  
          // TODO: why does this happen... a redirect?
          warn('[TRYING] Attempt to reuse xhr from ' + elapsed + " ms ago");
  
          if (elapsed > visitTimeout)
            timeoutError();
        }
        else {
  
          warn('[TRYING] Attempt to reuse xhr with no attemptedTs!!', xhr);
        }
      }
  
      ad.attempts++;
      ad.attemptedTs = now;
  
      if (!validateTarget(ad)) return deleteAd(ad);
  
      return sendXhr(ad);
      // return openAdInNewTab(ad);
      // return popUnderAd(ad)
    };
  
    const sendXhr = function (ad) {
  
      // if we've parsed an obfuscated target, use it
      const target = ad.parsedTargetUrl || ad.targetUrl;
  
      log('[TRYING] ' + adinfo(ad), ad.targetUrl);
  
      xhr = new XMLHttpRequest();
  
      try {
        xhr.open('get', target, true);
        xhr.withCredentials = true;
        xhr.delegate = ad;
        xhr.timeout = visitTimeout;
        xhr.onload = onVisitResponse;
        xhr.onerror = onVisitError;
        xhr.ontimeout = onVisitError;
        xhr.responseType = ''; // 'document'?;
        xhr.send();
      } catch (e) {
        onVisitError.call(xhr, e);
      }
    }
  
    const onVisitResponse = function () {
  
      this.onload = this.onerror = this.ontimeout = null;
  
      markActivity();
  
      const ad = this.delegate;
  
      if (!ad) {
  
        return err('Request received without Ad: ' + this.responseURL);
      }
  
      if (!ad.id) {
  
        return warn("Visit response from deleted ad! ", ad);
      }
  
      ad.attemptedTs = 0; // reset as visit no longer in progress
  
      const status = this.status || 200, html = this.responseText;
  
      if (failAllVisits || status < 200 || status >= 300) {
        return onVisitError.call(this, {
          status: status,
          responseText: html
        });
      }
  
      try {
  
        if (!isFacebookExternal(this, ad)) {
  
          updateAdOnSuccess(this, ad, parseTitle(this));
        }
  
      } catch (e) {
  
        warn(e.message);
      }
  
      xhr = null; // end the visit
    };
  

  That’s pretty much it! Let me know if it doesn’t make sense, I can annotate it

  source