[00:02:17] <tgr>	 maybe a simpler alternative is to convert to raw HTML, use regex to figure out where cuts start/end, replace those parentheses with something very unique, convert back to DOM, and then you can discard in a single pass
[00:02:55] <tgr>	 it's more work internally, but most of it is done by domino
[00:04:21] <tgr>	 parentheses in attribs etc. would still mess up the identification of ranges but the only failure would be too much / too little text removed
[00:08:21] <cooltey>	 tgr: I've noticed the description of the "Picture of the day" in zh-wiki only shows "<" , is that excepted? https://zh.wikipedia.org/api/rest_v1/feed/featured/2018/01/10
[00:12:32] <tgr>	 cooltey: probably weird EXIF data in the image
[00:13:00] <tgr>	 should be caught somewhere along the way though
[00:13:29] <tgr>	 ideally in the MediaWiki metadata extraction logic
[00:19:14] <cooltey>	 I see. Should we fix it? 😮
[00:19:23] <tgr>	 cooltey: either broken by https://gerrit.wikimedia.org/r/#/c/402151/ or a longtime bug that only surfaces with some weird image content
[00:19:52] <tgr>	 wanna file a bug about it?
[00:20:56] <cooltey>	 OK! I will file a ticket, thanks!
[00:31:59] <cooltey>	 tgr: a dumb question: what kind of tags should I tag for the ticket? 😅
[00:33:48] <tgr>	 mcs since that's the service that fails, multimedia (or mw-file-management) since it's file related
[00:35:49] <cooltey>	 tgr: thanks a lot!
[18:24:33] <Volker_E>	 jdlrobson: why do we have such constructs in mobile CSS: `@media print and (max-device-width: 720px)`
[18:27:46] <Volker_E>	 view-source:https://en.m.wikipedia.org/wiki/Jade_Terrapin_from_Allahabad