Plugin Directory: Include link to changelog.txt if exists

[Ipstenu]

Given that readme files over 10k act weird (they're too large), we've been recommending people move their old Changelog updates to a changelog.txt file

This helps a lot, but it also makes the changelog LESS discoverable for people who care.

If possible, having a link to the changelog.txt file off the developers tab would restore this usefulness.

[joyously]

Would this affect what is shown in the admin when clicking on View Details for a plugin update? That's where it's so useful to see the changelog, to know if the plugin should be updated for that site or not.

[Ipstenu]

No. This is not related.

Or it shouldn't be. The changelog.txt should be for the OLD changelog data (like if you're on version 3.0, you can move 2.0 and older to the other file).

[dd32]

Given that readme files over 10k act weird (they're too large)

This was a limit of bbPress 1 - the serialised readme content would overflow the field it was stored within.

[Ipstenu]

Things still don't show properly if they get too long per @otto42

[dd32]

FWIW Here's two of the largest plugin readme changelog's:

​https://wordpress.org/plugins/woocommerce-jetpack/#developers
​https://wordpress.org/plugins/easy-digital-downloads/#developers

They're both 200K+ changelogs that display perfectly ok. There's >4,000 plugins with readme's over 10K, 200 over ~50KB, 15 over 200K.
Adding support for this and encouraging plugin authors to put extra effort into yet another thing doesn't seem worth anyones time, especially when it doesn't cause issues.

If we want to encourage changelogs to be cleaned up, I'd rather just suggest plugin authors delete changelogs for any releases done over a year ago, it's not like it's going to stay relevant.

[Otto42]

I would not define having a 382k HTML for the plugins page to be "proper" in any way at all.

For speed considerations alone, I still would recommend reducing the readme.txt. And I'm fine with recommending people ditch old changelogs entirely, because most users don't care.

[obenland]

I'd rather just suggest plugin authors delete changelogs for any releases done over a year ago, it's not like it's going to stay relevant.

Replying to Otto42:

I'm fine with recommending people ditch old changelogs entirely, because most users don't care.

Let's go with that then.