What is HTTP ETag?

You possibly have used ETag on WordPress Cache Plugin. But What is HTTP ETag? ETag is an identifier assigned by a web server for versioning. There are differences between Uniform resource locator (URI) and Uniform Resource Identifier (URL). ETags are like fingerprints. This fingerprint is used for data deduplication purposes. ETag stands for entity tag.

What is HTTP ETag in Short?

ETag is part of HTTP. HTTP is the protocol for the World Wide Web or WWW or W3.

Going a bit off topic, that www is the thing which wrongly added before the domains and still there are morons use it; sometimes forces others via tutorials to change the default newer path on web server software, despite it holds no value. Only SEO importance of WWW subdomain is that, it must be 301 redirected towards the main IP address. Naked domains carry higher weightage – we can not point a naked domain with only CNAME. It should get resolved to an IP address. ANAME Record is not any standard. Nginx, for a reason uses /usr/share/nginx/html as public directory and it should not be changed to /var/www in tutorials without mentioning the history or link towards the history.

ETag is one of many methods, HTTP provides for web cache validation, thus you used ETag for W3TC Plugin for WordPress. ETag allows a client on client-server model to perform conditional requests. This method allow the caches to be more efficient, because the web server does not need to send a full response if the content has not been modified. In other sense, this ETag can be used for concurrency control, to prevent simultaneous updates of one resource from getting overwritten on multiple nodes.

So this ETag is kind of identifier which is dynamically assigned by a web server software and corresponding dynamic programing language like PHP to a specific version. If the resource content at a certain URL changes, a fresh new, different ETag is assigned. But nothing more can be interpreted from their comparison.

The wordings must be understood rightly – usage of strong ETag (terminology, see later) is impractical for a web server to generate with dynamically-generated content by PHP or other dynamic programing language. But the cached version is purely HTML. The decision, which one will be served (the cached or the dynamic) is an internal logical decision by the web server from the settings done by the webmaster.

One can check Header for Etag using cURL :

If-None-Match: “1d30-4c993ec28581d” means; from your command line, on again doing a cURL, the server may now compare the client’s ETag with the ETag for the current version of the resource. If the ETag values match (resource has not changed) then the server may send back a HTTP 304 Not Modified status. “Use the old version you have”, in other sense.

What is HTTP ETag in the Context of Validation?

ETag can be strong or weak. Strong and weak are labelled by validation. A strong ETag match means that the content of the two resources is byte-for-byte identical. Strong ETags permit the caching and reassembly of partial responses, as with byte-range requests. Weak ETag match indicates that the two resources are semantically equivalent and used in dynamically-generated content.

W/ is the prefix for weak ETag. ETag are generated using a collision-resistant hash function of the resource’s content, a hash of the last modification timestamp or the revision number. Historically, CRC32 and CRC64 suffered from hash collision problem.