[03:38:19] Krinkle: In T40417 you write "the edit page cannot vary based on data in a session", but that's not realy true. It's sort of true for anon views where there is no session, but then varying based on session data is impossible when there is no session anyway. For logged-in users, the edit page contains a hidden form field with the edit token, which varies based on session data. [03:38:20] T40417: MediaWiki's anonymous edit token leaves wiki installations (incl. Wikipedia) open to mass anonymous spam we can't block - https://phabricator.wikimedia.org/T40417 [12:44:16] RoanKattouw: true. I’ll double check but I think I meant we can’t make GET action=edit start a session with set cookie [12:44:35] Per the restrictions [12:46:52] Can -> can make, can not -> may not be made to [15:41:44] paladox: Apparently the comment textareas (inline and not) in the new Gerrit UI have some weird event handling going on. If I use JavaScript or browser addons (e.g. GhostText) to manipulate the value of the textarea, Gerrit doesn't notice the change when I hit "Save" or "Send" unless I manually type something in the textarea after. [15:42:33] anomie yeh, that would be because the way polymer is built [15:42:37] and the binding [15:43:09] so because GhostText did not use the binding that the commit msg uses [15:43:17] gerrit would think it's ok to submit [15:44:28] also it uses your local storage [15:44:36] so it caches what you write in there too [15:44:45] paladox: When "Save" or "Send" is clicked, it should recheck the textarea to pick up any changes its event handlers missed. [15:45:25] anomie hmm, im guessing that could possibly be fixed in 2.15 / 2.16 or 3.0 [15:45:48] Could as easily not be fixed [15:46:28] anomie multi binding does recheck though [15:46:45] ie if you did [15:46:59] this.test = 'test' [15:48:33] and then properties: { test: { type: String, notify: true } } [15:49:03] and then the element before being [15:49:12] that indicates it expects data to go both ways [15:49:25] [[]] would mean it going one way [15:49:49] anomie do you notice it on https://gerrit-test.wmflabs.org/gerrit/q/status:open / https://gerrit-new.wmflabs.org/r/ ? [16:02:16] paladox: The bug exists on gerrit-test.wmflabs.org, at least for making inline comments. [16:02:24] ok [16:02:45] paladox: Is there an "anonymous" account on gerrit-new? [16:02:52] anomie nope [16:03:01] Looks like I can't test there then. [16:03:10] it uses https://ldapauth-gitldap.wmflabs.org [16:04:27] paladox: Ok. The bug exists on gerrit-new too, at least for making inline comments. [16:04:31] ok [16:04:38] anomie i can submit this upstream [16:04:43] could you file a task in phab [16:04:49] and i can forward this upstream :) [16:05:09] upstream are currently re doing the ui again (they are now onto the header / admin ui / change ui) [16:05:18] paladox: What project in Phab? [16:05:20] (they have been updating the ui all week) [16:05:22] anomie gerrit [16:12:53] paladox: https://phabricator.wikimedia.org/T192665 [16:13:10] ok :) [16:13:15] forwarding upstream [16:18:18] anomie https://bugs.chromium.org/p/gerrit/issues/detail?id=8811 [16:19:26] paladox: Is the annoying "Ctrl-Z doesn't work in the Reply popup" bug already filed upstream? [16:19:51] anomie nope, does it work in https://gerrit-new.wmflabs.org/r/ or https://gerrit-test.wmflabs.org/gerrit/q/status:open ? [16:19:59] i can file it upstream [16:20:08] though file in phab and i will forward it upstream :) [16:20:35] anomie oh wait [16:21:00] nope never mind [16:21:04] thought it was https://gerrit-review.googlesource.com/c/gerrit/+/169050 [16:27:44] paladox: https://phabricator.wikimedia.org/T192666 [16:27:49] thanks [16:31:03] anomie https://bugs.chromium.org/p/gerrit/issues/detail?id=8812 [16:54:34] https://phabricator.wikimedia.org/T189215 [16:56:42] apparently visualeditor can cause sites to go down. [16:56:43] for example using visual editor on https://meta.miraheze.org/wiki/Miraheze (that specific page) [16:56:52] causes everything to be muddle up and then things take a while to load