[00:03:09] <MirahezeRelay>	 <mudscape, replying to orduin> That's probably not needed now that we aren't directly storing json
[00:04:03] <MirahezeRelay>	 <mudscape> Storing json data type on MySQL is stupidly finicky
[03:23:38] <MirahezeRelay>	 <orduin, replying to mudscape> looks like it might still be necessary, I got `Wikimedia\Rdbms\DBQueryError: Error 3144: Cannot create a JSON value from a string with CHARACTER SET 'binary'.` when testing with mysql, but not with mariadb. Looks to be inconsistent behavior between the two applications.
[03:45:03] <MirahezeRelay>	 <orduin> swapping from using the json data type to longtext should fix the issue
[09:16:27] <MirahezeRelay>	 <thewwrnerdguy> JSON handling is one of the few areas where mysql and mariadb differ, I've bumped into that issue before lol
[10:01:07] <MirahezeRelay>	 <originalauthority> #LongLivePostGres
[15:52:56] <MirahezeRelay>	 <thewwrnerdguy> <https://issue-tracker.miraheze.org/T15584> is the drive link here even usable
[15:57:13] <MirahezeRelay>	 <pskyechology, replying to thewwrnerdguy> no
[16:26:59] <MirahezeRelay>	 <notaracham> ...did they convert the images into a hash?  I am confused.
[16:27:08] <MirahezeRelay>	 <notaracham> Oh, revision metadata.  Nvm.
[16:33:00] <MirahezeRelay>	 <whostacking> hi, would like to bump https://issue-tracker.miraheze.org/T15767. editors are now unable to edit anything without it being glitched
[16:33:25] <MirahezeRelay>	 <whostacking> i suspect it could be due to the wiki reset on the wiki that happened not long ago
[17:16:38] <RhinosF1>	 dmehus: has that image purged now?
[17:20:43] <MirahezeRelay>	 <rhinosf1, replying to whostacking> We are aware of that
[17:22:52] <MirahezeRelay>	 <rhinosf1> @paladox is memcache safe to restart to clear it? I'm fairly sure that's completely broken and corrupt revision cache which is a memcache thing
[17:23:56] <dmehus>	 RhinosF1: ack, yes. It seems that in preview mode, it was perennially stuck in the previous iteration, which, I'm guessing, is as intended. When I submitted the edit, it still showed the previous iteration, but refreshing the image page and purging seemed to fix it
[17:24:06] <dmehus>	 thanks :)
[17:24:36] <RhinosF1>	 Cool
[17:24:45] <RhinosF1>	 I gave it some gentle encouragement
[17:24:56] <RhinosF1>	 Which we can do if needed
[17:25:09] <RhinosF1>	 But generally waiting a while is expected for thumbnails now
[17:25:59] <dmehus>	 RhinosF1, sounds good thanks :)
[17:29:12] <MirahezeRelay>	 <paladox, replying to rhinosf1> yeh but remember if it's parser cache it'll be in the DB so restarting memcache without dealing with that first would be meaningless
[17:36:09] <MirahezeRelay>	 <rhinosf1> @whostacking did that help?
[17:36:30] <MirahezeRelay>	 <rhinosf1, replying to paladox> I think revision cache is more likely tbh
[17:38:04] <MirahezeRelay>	 <rhinosf1> I will purge as well though
[17:39:28] <MirahezeRelay>	 <whostacking> yes it did! thanks for the help. editing works as intended now
[17:40:47] <MirahezeRelay>	 <rhinosf1, replying to whostacking> Good
[17:40:53] <MirahezeRelay>	 <rhinosf1> I set off a purge for good measure
[17:41:03] <MirahezeRelay>	 <rhinosf1> But it was probably revision cache which was cleared
[17:46:05] <MirahezeRelay>	 <whostacking> so re: doors import issue, some pages appear to be present, but templates and other namespaces are inconsistent?
[17:52:20] <MirahezeRelay>	 <rhinosf1, replying to whostacking> That I can't help with
[17:52:24] <MirahezeRelay>	 <rhinosf1> Because I didn't do the import
[17:52:55] <MirahezeRelay>	 <aeywoo, replying to whostacking> Yes, I explained that in the ticket for the import, but I believe the import wasn't successful.
[17:53:38] <MirahezeRelay>	 <thewwrnerdguy> the import is probably possible now
[17:54:21] <MirahezeRelay>	 <aeywoo, replying to thewwrnerdguy> Well, the pages that were already imported will need to be wiped won't they? Unless the import can recontinue?
[17:54:38] <MirahezeRelay>	 <thewwrnerdguy> I think the import can recontinue
[17:54:42] <MirahezeRelay>	 <aeywoo> But isn't it possible for us to run into stale revisions?
[17:54:57] <MirahezeRelay>	 <thewwrnerdguy> if not then the wiki will probably need resetting once again so might as well try
[18:15:39] <MirahezeRelay>	 <rhinosf1, replying to thewwrnerdguy> Imports can be resumed
[18:15:45] <MirahezeRelay>	 <rhinosf1> Mediawiki should do the right thing
[18:15:58] <MirahezeRelay>	 <rhinosf1, replying to aeywoo> What stale data?
[18:38:15] <MirahezeRelay>	 <thewwrnerdguy> [1/2] does anyone know why the lovenikkiwiki refreshlinks stuff is taking so long
[18:38:15] <MirahezeRelay>	 <thewwrnerdguy> [2/2] I started on mwtask151 and mwtask161 9 hours ago and they've both only done about 3500 pages
[19:48:00] <MirahezeRelay>	 <wikitideharej> https://meta.miraheze.org/wiki/Extensions says that MultiUpload was declined and that as an alternative you can use UploadWizard (among other options). But I don't see UploadWizard mentioned anywhere else on that page as an enabled extension
[19:48:23] <MirahezeRelay>	 <wikitideharej> (To be clear I don't want MultiUpload; I found this while checking to see if UploadWizard was on the list)
[19:52:31] <MirahezeRelay>	 <thewwrnerdguy, replying to wikitideharej> it's spelled "Upload Wizard"
[19:59:41] <MirahezeRelay>	 <notaracham> It's there, or at least used to be
[20:00:05] <MirahezeRelay>	 <notaracham> We definitely used it on my project in the past, even though it was a little cantankerous
[23:32:00] <MirahezeRelay>	 <wikitideharej> Does anyone know how this could happen? https://commons.miraheze.org/wiki/Commons_talk:Policies?action=history