Forum Replies Created

Viewing 15 replies - 1 through 15 (of 138 total)
  • Thread Starter bjf2000

    (@bjf2000)

    It could have been worse: 20 years, for example. I have no explanation.

    Anyway, I also wasn’t aware of that particular filter, but it works great, thanks:

    add_filter( 'simple_history/add_admin_bar_menu_item', '__return_false' );

    bjf2000

    (@bjf2000)

    It’s unclear if we’re meant to act on the messaging though (if we can–I’m unsure). What do you think? And what’s the worst that can happen if we can’t?

    Thread Starter bjf2000

    (@bjf2000)

    I did that, and they said that the feedback and feature request were passed on to the team, so that’s promising.

    I had forgotten that the audit log, which I only mentioned in passing, was Premium. It’s not the problem though, and I just mentioned it to provide a fuller picture.

    The overall feature request pertains to everyone, Free included: make the alert emails more informative, for the reasons outlined.

    Thread Starter bjf2000

    (@bjf2000)

    Solved may be taking it too far, as it still does happen occasionally, just not constantly like it gotten to before, only “corrected” by a purge. Sure would be nice to actually solve it.

    Thread Starter bjf2000

    (@bjf2000)

    I believe that I’ve finally cleared this up.

    The 403 problem: I think it’s resolved, but it has gone a long time between events before. Still, it hasn’t happened once since the reset in this higher-frequency period, so that’s a good sign. It could have been the existence of the page rules, or it could just have been that my particular combination of settings didn’t agree with the major revamp of the plugin several months ago. We’ll probably never know, but if anyone hits this problem, I suggest simply resetting the plugin and heeding the following section, if it applies.

    The color scheme problem: When troubleshooting why the ability to purge cache from the menu was missing after the reset despite being On in Settings, I made an incredible discovery. To explain, I’ve been using the Dark Reader browser extension for about the last year (long after I originally configured your plugin). When it’s active, this is what the toggles look like for your plugin:

    View post on imgur.com

    I thought nothing of that (it’s obviously disabled) until I started noticing that some settings showing in the JSON export didn’t agree with what I was seeing in the GUI, like the feature that lets you purge cache from the menu. As you know but I didn’t, that above screenshot is NOT how you show which way it’s set. When you disable Dark Reader for the site (which I’d rather not do since it then disables it for the rest of admin and the public site), it then looks like this:

    View post on imgur.com

    So, now realizing that I’d made numerous changes yesterday based on false information, I reset again, and not surprisingly, things went much better, such as:

    View post on imgur.com

    And yes, your Cache Rule does appear in CF now. It wasn’t before the reset. That missing rule in combination with Disk Cache being On (when I thought it was Off!) might have been the 403 issue in a nutshell.

    The response header still shows HIT, the numbers on the dashboard are better, etc. It’s working fine now that I got out of the way.

    Suggestion: It’s great that you have your own dark theme for the plugin, but if possible, can you somehow work around the Dark Reader problem? If for nothing else, it might save someone else going down this same rabbit hole.

    Thread Starter bjf2000

    (@bjf2000)

    I backed up settings and reset, since it’s been many years and I wanted to go back to defaults. This also gave me the opportunity to use a token.

    Too early to know if it made a difference, and I know the Test function has its own problems, but I still can’t get it to look better than this:

    View post on imgur.com

    If a Cache Rule is supposed to be visible over on CF, that statement is correct, none was created. Only one related to ACME Challenge exists, and we created that.

    And Disk Page caching certainly does have an issue, since it’s disabled. When enabled, it forces “x-wp-spc-disk-cache” to disabled. If I keep Disk Page caching off, I get HIT (as before). Furthermore, Cache Engine goes to Disabled when I enable Disk Page caching, so it really doesn’t like us.

    Enable Cloudflare CDN & Caching is On (as before).

    I’ve now disabled our three legacy Page Rules on CF, though I don’t see a word there about any deprecation. Where did you see Page Rules are deprecated? In any case, I hope these were the problem with the 403’s.

    For the record, the Page Rules were (in this order):

    oursite.ca/preview=true*
    Cache Level: Bypass, Disable Performance

    oursite.ca/wp-
    Browser Integrity Check: On, Browser Cache TTL: 30 minutes, Security Level: High, Cache Level: Bypass, Disable Apps, Disable Performance

    https://oursite.ca/*
    Cache Level: Cache Everything

    Incidentally, I see nothing in the Page Rules section about any deprecation, which is very unusual for CF when they’re deprecating something.

    Thread Starter bjf2000

    (@bjf2000)

    I should add that Disk Page cache is disabled, as it always has been because I wasn’t aware of it before (if it existed in the old UI, I don’t think it was as prominent as it is now, or there was phrasing to discourage using it). Would enabling it help?

    The header when it’s happening (as it did again today–no idea why the frequency has increased):

    {"Age":["47"],"Alt-Svc":["h3=\":443\"; ma=86400"],"Cache-Control":["max-age=1333600, must-revalidate"],"Cf-Cache-Status":["HIT"],"Cf-Ray":["990a7a722bbadcce-FRA"],"Content-Encoding":["br"],"Content-Type":["text/html"],"Date":["Sat, 18 Oct 2025 19:35:20 GMT"],"Server":["cloudflare"],"Server-Timing":["cfCacheStatus;desc=\"HIT\"","cfEdge;dur=8,cfOrigin;dur=0"],"Strict-Transport-Security":["max-age=31536000; includeSubDomains; preload"],"Vary":["Accept-Encoding,User-Agent"],"X-Content-Type-Options":["nosniff"],"X-Frame-Options":["SAMEORIGIN"]}

    Thread Starter bjf2000

    (@bjf2000)

    @freshpromo No, but yours appears to be another kind of warning. As long as it doesn’t start doing it in heavy volume, it may just be a one-off or occasional thing.

    Thread Starter bjf2000

    (@bjf2000)

    Please just tell me what the nature of the particular file is and its importance. Also, what particular Elementor settings, if any, pertain to it. I need to make this stop and I don’t feel it’s fair that you’re not revealing anything about such a small file that should have a very clear purpose that you could relate to me in a couple sentences.

    50,000 lines in the log in 7 days is not harmless.

    Purely hypothetically, why would one think it’s correct to have a PHP file with an undefined array key when presumably array keys should be defined?

    Thread Starter bjf2000

    (@bjf2000)

    @wfpeter

    Hi, OK, so to take this from another angle, is there a Wordfence setting that you’re aware of which would preclude the sending of just the verification mails as opposed to others relating to Wordfence? Because if so, I should probably check that, since the others do come in and have been reliable for years.

    While I understand that all “Wordence” emails come from us and not you, these particular emails do involve an outside party (Google), as opposed to the normal variety of Wordfence emails. So I assume the Wordfence plugin is dealing with Google in some way before sending the email. Given that I wonder what would happen if there was a problem between the plugin and Google. It might prevent those emails from sending. Is there a log of this sort of thing buried somewhere?

    The site has no plugin related to sending of emails, so it’s just the normal method that the host uses for sending on the back end. I guess it’s possible that they’re filtering out just these, though that would be pretty wild and I wasn’t aware that was even a possibility. I’ll try to check though. I do know, at least, that the ones that do come in from Wordfence/Wordpress are perfect in terms of SPF/DKIM/DMARC, so the problem shouldn’t be in that direction.

    As for the scoring, I think that’s working well, as I really have to botch a login repeatedly before Google steps in. That’s good. We’re on the default of 0.5.

    Thanks

    Thread Starter bjf2000

    (@bjf2000)

    @josklever

    Wow, so it was as you described, and I hope this does the trick (won’t know right away, since a few days sometimes separate these emails). I have no idea how I did that originally, and it’s so subtle that I would have never noticed that was then a way to unsubscribe, especially given that it looks like nothing happens when you click on it (as you say, two letters in the URL change). Thanks.

    Thread Starter bjf2000

    (@bjf2000)

    That would seem like the place, but:

    View post on imgur.com

    I suppose one of those default rules might be kicking in, but I have no connection to any tickets, unless somehow creating a thread in this forum makes a connection. This did start at basically the same time as my creating a thread here about JetPack.

    Thread Starter bjf2000

    (@bjf2000)

    @fierevere

    Those are forum topics though, which is not what I’m talking about (I mentioned actually wanting those). The only other thing on the page is listed at the bottom:

    Subscribed Topic Tags
    You are not currently subscribed to any topic tags.

    Subscribed Plugins
    You are not currently subscribed to any plugins.

    Subscribed Themes
    You are not currently subscribed to any themes.

    Thread Starter bjf2000

    (@bjf2000)

    They have the same return address as ones from this forum: [email protected]

    But the “From” name is different. For ones from this forum, it’s “ww.wp.xz.cn Forums.”

    For the WordPress Plugin emails, the “From” name is omitted.

    Thread Starter bjf2000

    (@bjf2000)

    Incredibly, hours after writing that, Jetpack 13.1.3 was released:

    “Backup: write helper script to ABSPATH by default to avoid backup failures.”

    It took some detective work, but I traced that back to this:

    https://github.com/Automattic/jetpack/pull/35508

Viewing 15 replies - 1 through 15 (of 138 total)