Paul Biron

A temporary work around is:

• manually edit wp-includes/version.php
• set the value of $wp_version to 5.4-beta3
• go to Tools > Beta Tester (or Network Admin > Settings > Beta Tester on multisite) and select Beta/RC - Bleeding edge

You should then be able to update to 5.4-RC1.

It leads to the following page full of dropped table errors and a site that was previously working before, completely breaks.

View post on imgur.com

Based on that 2nd screenshot something is definitively wrong, because the tables shown there as not existing haven’t been a part of WP since version 2.2 (since about 12 years ago).

Thanx for providing the info about the version you updated from and the answer to @howdy_mcgee ‘s question re: plugins.

Sorry to hear that you’re having this problem but I haven’t been able to replicate it.

I created a small site with WP 5.1.1, about 10 different plugins (the only one from your list being Classic Editor). I then created various posts with post_type != ‘post’ and created a “custom” table (did that part manually, i.e., it wasn’t actually created by one of the installed plugins).

I then updated to 5.2.2 in /wp-admin/update-core.php without problem: post_type’s were not changed and the custom table is still there.

I then repeated the process but updated to 5.2.2 via WP-CLI, with the same result…all was fine.

Since the release of 5.2.2 yesterday I’ve also updated ~30 other sites (some via automatic update, some via WP-CLI) with no problems.

As unlikely as you think it is I, too, suspect that the problem is with one of the plugins you’ve got installed.

Sorry I can’t be more help.

What version did you update from?

Thanx for the review, I’m glad you like it.

thank you!

I’m aware of that other option.

I should note that my use of Live Traffic may be different from that of others/the intended purpose. I don’t generally use Live Traffic to see what’s happening on my sites “right now”. Rather, I view Live Traffic once a day to look for suspicious accesses that were not block by the WAF and that maybe should be manually blocked.

I often hit the front-end of sites I manage when I’m not logged in, hence, even with Don’t log signed-in users with publishing access turned on
many of my hits would still be shown in Live Traffic. Of course, seeing my hits in Live Traffic is not bad, but it’s easier to quickly scan things if I don’t have to see them.

Turns out the problem is in Gutenberg.

I’ve opened an issue on GB Difference between GB’s and classic editor’s interpretation of TinyMCE’s style_formats option, so am closing this as resolved.

thanx. I emailed them yesterday (and got the confirmation response).

Meant to add a comment here saying I did so but got caught up in other matters and forgot about it.

Forgot to mention: I’m a free customer.

OK. I’ll experiment more and see if I can figure out a workable strategy for myself. Thanx for you quick replies!

Conclusion, for now you can keep the version built by your Linux system in version control, which works on windows systems too as far as I know.

yes, as far as I can tell, the rules built on the linux host do work on my windows box.

You just don’t save the settings, or overwrite the file with a save from your Linux system before committing it.

Did you mean to say “overwrite with a save from you Windows system before committing it”? If so, then yes, I guess I could VC the linux built .htaccess and just never change any settings on my windows box.

I do have a dev version of the site on the linux host, but it is far easier to experiment with how changes to a given setting effect what w3tc does on my windows box.

Are there specific setting(s) that control the output of those 2 lines in .htaccess? Or is it the simple act of clicking “Save Settings”? If it’s specific setting(s), then I can just make sure I don’t change that setting(s) will being free to experiment with other settings on the windows box.

Forgot to mention: w3tc 0.9.7.

Let me know if there is any other info you need to help figure this out.

I just released v1.1. In that version I’ve removed the code sample (and other Dev Notes) from the readme and replaced it with a link to the documentation on github, which has been updated as you suggested.

Also, that github documentation was been rewritten to hopefully be more clear.

Thanx again for the report.

Please accept my appologies for the delayed response! For some reason I was never notified that you posted this.

You are absolutely correct. I will be releasing a new version shortly and will be sure to correct that.