There a few claims you have that I think are slightly off.
First, push is very different than inlining, because individual resources are cacheable by the client, and inlined resources are not. There's a little handshake at the start of a push and if the client has (or doesn't otherwise want) the resource, it can cancel the push.
In my team's experiments with using push for HTML Imports, push ends up being better than inlining (using Polymer's vulcanize tool), even with a cold cache.
Also, push is relatively easy to implement if you use the response header approach, and was availably early on in mod_spdy.
Another method would be to add a headers to the response:
Server-Push: /relative/url
Then have an output filter pull them out and push them -- but AFAIK the APIs to do that don't yet exist in mod_h2 to push like that. Adding headers in the response is nice because it works with any 'backend', and the web server doesn't have to parse the HTML/css/etc going the through it.
I believe this could even avoid the need to include cache-busters in the URL since you could push with an ETag and the client can terminate if it already has the appropriate bytes without the incredible performance hit of using ETags with HTTP 1.
I think this is unsustainable (any non-trivial change to your html resources would require updating and restarting httpd). This is not DRY.
Jetty's approach to this problem is really interesting. Using HTTP referrer headers, it builds up a tree of resources that are often fetched together. Then, it optimistically pushes companion resources down.
https://github.com/eclipse/jetty.project/blob/master/jetty-s...
You'll note that I said “simplest” – this could work on a $2/month shared hosting plan (not sure where you got the “updating and restarting httpd” bit from with a couple decades of prior art showing otherwise).
The Jetty approach is clever, too, as would be things like scanning rel=prefetch links but simple declarative approaches have their moments, too. One thing the Jetty approach doesn't appear to offer is the ability to manage resources if not all resources are appropriate for every client so it would appear to do things like ship every size of a responsive image to every client.
If the server can parse served html, it could preemptively also send along other required assets (css, images, etc) before the client has time to parse and request them. This would help remove necessity for optimizations like image bundling.
Apache version numbering is similar to what the linux kernel used before 2.6.
In a.b.c, odd b indicates the unstable or experimental branch, and even b's are used for major and backwards-compatibility breaking changes.
Backwards-compatibility preserving changed can be introduced in c.
So 2.2 and 2.4 are pretty different versions, while 2.3 and 2.4 are closer except 2.3 is unstable. 2.4.0 is theoretically compatible with 2.4.17, except the latter has a lot more functionality and features.
If all it does is add new functionality, and doesn't modify or break any existing functionality, then that seems to be good software engineering to do in a minor upgrade doesn't it?
Seems valid: MAJOR.MINOR.PATCH, with the minimum changes being increment "patch" for backwards compatible bug fixes, "minor" for new/improved functionality that does not break backwards compatibility, and "major" for any API/feature change that is not backwards compatible. They could of course chosen to increment MAJOR instead, but they don't have to in this case.
(I'm assuming they are following "standard" semver rules here)
Compatibility in this case is referring to published APIs and not changing documented behaviour. There are a large family of changes that are bug fixes that don't change correct use of an API - many security fixes for example.
What's sad is that this is yet another github page with #777 colored text on white background. The resulting contrast is too low to read comfortably!