[11:41:11] Okay why did Mediawiki have to have such good awful whitespace handling semantics? [11:41:24] https://en.wikisource.org/wiki/Railways_Act_1921#86.1 Seventh Schedule [11:42:16] It seems to be if see random whitespace where not expecting mangle the output of a table [11:42:22] This is UNACCEPTABLE [11:43:50] I've tried VERY possible way I can think of to get these two pages to transclude cleanly [11:44:02] https://en.wikisource.org/wiki/Page%3ARailways_Act_1921_(ukpga_19210055_en).pdf/90 [11:44:09] https://en.wikisource.org/wiki/Page:Railways_Act_1921_(ukpga_19210055_en).pdf/91 [11:44:23] And it's starting to be beyond annoying to play hunt the semantics [11:44:43] ON EVREY SINGLE fix [11:44:57] I've been trying for a long time to get this working [11:45:35] And I'm getting to the point where I may scrap the whole thing because Mediawiki is too stupid to understand it's OWN markup [11:45:48] I'd like an apology [11:45:54] for my wasted time [11:46:05] do we know who invented wikisource? [11:46:17] Dysklyver: Sorry [11:46:21] I'm ranting again [11:46:24] xD [11:46:33] but having had the same issue crop up [11:46:55] when the "Fix" approach that worked in other places doesn't work in the exact same situation [11:47:09] seriously it sounds like a major design flaw in the way wikisource handles table transclusion over multiple pages [11:47:14] Expanding out the code in Special:ExpandTemplates gave no hints [11:47:28] Dysklyver: That is known issue [11:47:48] any fix planned? [11:47:58] Not with a recognisable ETA [11:48:14] humpf [11:48:24] i guess it's not an easy fix [11:48:28] Patches are complex to integrate [13:14:24] Dysklyver: https://phabricator.wikimedia.org/T185750 [13:14:31] Your input would be appreciated [13:15:06] The issue of Table continuations need to be resolved in my mind by having explictly specficed terminations [13:16:19] that would be a good idea imo [13:20:35] Currently a lot of wikitext markup is whitespace dependent [13:20:44] Thats fine on a site like Wikipedia [13:21:25] where you have single page and "realtively" simple templates.... [13:22:06] On Wikisourcre where you have several generator templates, like cl-act-paragraph, and many of the ones that generate US Satute formatting [13:22:48] havign markup that's whitspace dependent introduces subtle interactions that can be a pain in the &^% to pin down. [13:23:20] For Wikisource and Templates there needs to be a "whitespace" independent markup method [17:33:22] Hi [17:33:27] Anyone awoke? [17:33:40] And willing to talk to a "difficult" user? [17:33:44] (I.e me) [17:34:20] always [18:19:48] Dysklyver: The eventual issue I found I was having was essentialy a line of table markup that ended up reading as follows [18:20:30] |item1|-|item2 which given how template syntax works WILL break [18:20:53] dang [18:20:58] adding nop/nopt is not going to solve the fundamental problem of a lack of a terminator [18:21:23] I.e forcing the |- at the start of the template to be on a newline in every circumstance [18:21:43] Mediawiki markup in this instance isn't whitspace independent [18:22:13] I don't see an easy soloution to this given the way the relevant template currently works.. [18:22:27] I don't see an easy solution either [18:22:43] It may be that all it needs is {{nopt}} to have a line feed. [18:23:11] But it would be better to my mind if {{#nopt:}} {{#nop:}} where actually parser hints [18:23:18] rather than "hacks" ;) [18:23:47] I think the developers are aware of this though [18:25:05] yes [18:26:33] There's also the issue of how to do sidenotes that can be formatted like DIV's [18:26:58] the current ones are spans... which isn't necessarily a good thing for lengthy ones [18:27:55] HTML5 doesn't have an easy method for breaking a DIV out a P or SPAN [18:28:22] and an