Release 4.1.0

[dd32]

v4.0 was released on Jul 24, 2021, it's probably about time for a new release, especially with the PHP 8.2 fixes.

This PR contains the changes needed to the Readme & Plugin file to make this happen.

The changes that would be included in the release are this: 4.0.0...master (plus this PR).

I've drafted a release: https://github.com/Automattic/wp-memcached/releases/tag/untagged-4d644ee8bdcb0576e6dc

I've requested a review from the folks who have contributed to the repository for review of this + the release notes above.

Fixes #163

[dd32]

Probably worth pulling in #160 and #136 first

[lucas-oliver-trondsen]

Thank you @dd32 for this new version! This plugin is pure gold and it should never be abandoned. So many developers are using it. Only those who have never experienced the page loading speed gains of using memcached, and had fallen for Redis' marketing techniques, are not using memcached and this fantastic plugin. If they hadn't been so inclined to imitate others, they would have gained so much from the simplicity and efficiency of memcached and of this plugin, since memcached is so lightweight and even faster than Redis.

[edneville]

@lucas-oliver-trondsen, much of IT is like this - those with marketing departments get more users, probably because enterprises feel happier to put money behind it as they look like they have more investment and more likely to succeed. Marketing departments will point at the others as "hobby". Why isn't this plugin part of the stock install though? I think the #159 might have stopped it, but also the unique salt on install needs to be guaranteed. I wonder if the salt isn't present, a key hash of siteurl+db_user+db_pass+db_name+db_host would be enough to guarantee a unique site? Maybe then this could be part of the stock install? It seems a lot of energy computing dynamic pages could be saved. Are there other problems?

[lucas-oliver-trondsen]

I think a hash of the site's domain as salt would work. WP_CACHE_KEY_SALT has to be unique for each website connected to the same Memcached server and since each site's domain is unique, the domain's hash should be ok (something like hash('sha256', 'somedomain.com')).

It would be nice to have this plugin added to the WordPress core. If the developers of this plugin wish, they can add a pull request to WordPress, so that this plugin becomes available by default to anyone who installs WordPress and wants to enable object caching using Memcached. However, I don't know if it will be accepted.

@dd32 are there any news about publishing this new version ?

[edneville]

@lucas-oliver-trondsen

In #166 I've used the DB parameters + blog_id, in theory there could be a situation where two WP installs on the same machine present an intranet version and an internet version, for the same hostname, one both to internal network and the other to the external, yet sharing the same memcache. They should have different DB sources though.

In my experience, where there are multiple sites on the same machine, there's a much higher risk of two sharing a copy/paste error than running into a collision from this.

Let me know if I've missed anything on that PR please.

[lucas-oliver-trondsen]

@edneville Indeed, in rare cases there can be 2 websites with the same hostname, one site served inside the LAN, with a separate DNS server, and the other served on the Internet. I looked at #166 and and it looks good to me. Since the salt has to be unique, and $blog_id seems to be unique even in multisite setups where multiple websites share a common wp-config.php file, the uniqueness is there.