#wordpress 2023-03-29

2023-03-29 UTC
#
[snarfed] Ok. Any idea what's happening then? So weird since the HTML is exactly the same
#
GWG Not sure, have to look more closely
sebbu2, gRegor, RasAlGhoul_, [jeremycherfas] and [pfefferle] joined the channel
#
[pfefferle] [snarfed] have you changed some of your CSS? Since when have it changed?
gRegor, IWSlackGateway, [tantek] and [snarfed] joined the channel
#
[snarfed] [pfefferle] do you mean the CSS on Bridgy and Bridgy Fed? no, not for a long time
#
[snarfed] ^ I figured out part of this bug though, it's caused by the escaped # character in the source URL. if I double-escape it, ie %2523 instead of %23, the webmention works
#
[snarfed] which makes sense, since HTTP clients generally won't include fragments when they fetch URLs
#
[snarfed] so, not really a bug in the wordpress plugins
[dave] joined the channel
#
[dave] [pfefferle] any suggestions on next debugging steps for activitypub? No longer getting 406 error on follow attempts but follower still seeing ‘Pending’ or text about reversing follow attempt on Mastodon.
[pfefferle] joined the channel
#
[pfefferle] hey david, sorry for beeing so silent atm. but I have/had some final things to prepare before joining Automattic! I will come back to you as soon as possible to help you with you issues!
#
jeremycherfas GWG can you help me to understand the significance of this URL https://www.eatthispodcast.com/the-evolution-of-food-culture-in-mali/#comment-99748/ ? As far as I can tell, there are no comments on that page. Does WP simply ignore the #comment-no? Why?
#
Loqi [preview] [Jeremy Cherfas] The evolution of food culture in Mali Where eating alone is unthinkable
[manton] joined the channel
#
[dave] No worries, I was gonna wait but saw you message snarfed above and figured maybe you were settled in now -- not a big deal
#
[dave] [jeremycherfas] # and anything after are usually handled client-side by the browser and would typically be ignored if there's not a matching anchor in the page, unless there's some scripting that does something interesting with it. I assume that's the case here.
#
[dave] for example https://www.eatthispodcast.com/the-evolution-of-food-culture-in-mali/#notes would take you to the notes section because it has the id "notes" in the source code for the header, but https://www.eatthispodcast.com/the-evolution-of-food-culture-in-mali/#feralhogs would just take you to the page and leave you at the top since there's nothing with the name or ID 'feralhogs'
#
jeremycherfas OK. The reason I want to know is that that particular URL -- i.e. that page plus a seemingly random comment number that does not exist -- is being hammered on my server, and I have no idea why. I'm now trying to find out where they are coming from, but looking at logs is not my strong suit.
#
[dave] I think Wordpress adds the id "comment-(comment number)" to all comments and perhaps there was previously a comment there that was deleted
#
[dave] If you have access to the logs you might want to see if they all appear to be coming from the same IP/person and what they list as their user agent. you can filter by doing something like `grep "99748" ./log-file-name` to get just the relevant lines
#
jeremycherfas Nope, it isn't that. I've tracked down the vast majority of the requests to two servers in Russia, one in St Petersburg, one in Vyksa, Nizhegorodskaya oblast
#
[dave] oh weird
#
[dave] do they seem to hit any other URLs or just that one?
#
jeremycherfas Right now, this is just curiosity, and learning how to do such things.
#
jeremycherfas Maybe they hit other pages, but on that page, they hit a large number of random #comment-ids
#
[dave] I wonder if maybe they're using spamming software and trying to see how quickly their spam comments get deleted or if they manage to get through
#
jeremycherfas They're all GETs rather than POSTs
gRegor joined the channel
#
GWG Hmm...well, is there anything attached to the page? Webmentions, etc?
#
[pfefferle] it is used for threaded comments
#
[pfefferle] not completely true, the URL that is used by the threaded comments feature redirects to this URL
#
jeremycherfas Nothing. The episode dates to 2016, before I even added webmentions to the site
#
jeremycherfas I suppose there could be a bunch of comments that were automatically marked as spam by Akismet and then Antispam Bee
#
[snarfed] [jeremycherfas] the internet is the wild west, all kinds of undesired traffic like this happens all the time, sadly. [dave] is definitely right, blocking IPs etc can be useful, but otherwise it's not often worth spending too much time on any one instance like this to try to figure out its root cause
#
jeremycherfas Absolutely agreed, but I did want to get to the bottom of it, and it seems that comments automatically marked as spam are indeed the problem.
#
jeremycherfas Out of interest, because I've never done this, is it worth deleting spammed comments permanently, to free things up in the DB?
[tw2113_Slack_] joined the channel
#
[tw2113_Slack_] i find zero value in them
#
jeremycherfas Is that a yes delete or a no, no point deleting?
[KevinMarks] joined the channel
#
[tw2113_Slack_] delete
#
[tw2113_Slack_] get rid of in db
[schmarty] joined the channel
#
[KevinMarks] could the comments be linked from mastodon and it's servers trying to do link previews?
[jacky] and RasAlGhoul joined the channel
#
[pfefferle] GWG and I have nearly finished the merge of Webmentions and Semantic Linkbacks... There is one major change, that puts all facepiles on top of the comments instead of below... is this a show stopper for you?
#
[snarfed] [pfefferle] that's great news!!!
#
[snarfed] personally I'm ok with that
#
GWG I'm still trying to figure out the upgrade trigger..it isn't working in testing
#
GWG I think the other triggers should work, but,...
#
GWG I want the redundancy
[James_Van_Dyne] joined the channel