gamernetwork / yafa Goto Github PK

To avoid this dark traffic problem where referrer isn't sent.

The YAFA ads are being blocked on EG by uBlock Origin when used in Firefox, thanks to this addition: https://github.com/uBlockOrigin/uAssets/blob/10d0289b32549f1996df5983e518d5800e52c790/filters/filters.txt#L585

This is fixable using the MutationObserver class in JavaScript, but older versions of IE don't support it, and it also prevents people (like us!) from being able to edit styles via the element inspector, which could prove to be a pickle!

The basic code to do so looks something like this:

var observerOptions = {attributes: true, attributeFilter : ['style'], attributeOldValue: true}
    ,observer = (typeof MutationObserver !== 'undefined') ?
        new MutationObserver(function (mutations) {
            mutations.forEach(function (mutation) {
                if (mutation.attributeName == 'style') {
                    observer.disconnect();
                    mutation.target.setAttribute('style', mutation.oldValue);
                    observer.observe(mutation.target, observerOptions);
                }
            });
        }) : null
});
if (observer) {
    observer.observe(target, observerOptions);
}

based on image server logs?

It's messing and uses wrong zone. See VG for inspiration.

Given 2 campaigns:

• A with priority 10
• B with priority 1

The HIGHEST priority will always show. This allows ops to traffic a 'fallback' house ad for when nothing else is booked without having to explicitly enable and disable campaigns.

How does this combine with #1 ?

• A priority 10 weight 5
• B priority 10 weight 10
• C priority 5 weight 100

A and B will show with 1:2 probability ratio. C will never show until A and B are deactivated.

• EG
• USG
• VG247
• RPS
• ModDB/IndieDB
• Nlife/Pushsquare
• Gamesindustry.biz
• Metabomb

It's just a cloud right now, despite the model being hierarchical

Say 3 campaigns are trafficked to eg-net:

• Monkey Ball weight 2
• Trackmania weight 4
• HItman weight 6

Each would appear with a probability of weight/sum(weights)