Difference between revisions 102774147 and 102820976 on enwiki

This is an example of my '''IPA Quickhelp Templates''' idea for templates to use rollover "status text" to make IPA more comprehensible for beginners. The idea is that each letter is a link. When the user rolls over the link, they should get a short (under 15-word) jargon-free text explaining how to pronounce that phoneme. This links to a redirect which links to the jargon-titled article about that sound.

(contracted; show full)ht, so I limited the number of characters to 50, instead of 99.  This is a major problem with [[Template:familytree|templates like these]], and its going to take a while to find the right balance.  If you have any more questions, let me know.  <font face="trebuchet ms"> &mdash;[[User:Wdflake|<font color="#555"><b>W. Flake</b></font>]] ([[User_talk:Wdflake|<font color="#555">talk</font>]])</font> 22:07, 23 January 2007 (UTC)


:::::Would it help to design the template to be used with recursive subst: , so that all the extra unused characters get deleted but the original template call (subst: and all) gets reproduced in a comment? Then, in order to edit, you could delete all the visible stuff and the comment opener and closer and edit the original template call, and it would subst itself back with your edits.... obviously somebody could sneakily or ignorantly edit the template call WITHOUT deleting or popping it out of the comment and that would not show up until the next person came and edited it properly... but a simple "Do NOT edit this without reading the [[missing instructions|instructions]]" would help avoid that...
:::::OK, what I mean to ask is not "would this help", the fact is it would solve this particular problem. It is more like "is there a precedent, and if so, what did people think?" Keep in mind that whatever solution we have for how to enter and edit these things - whether it be separate templates, pipes, or a big mess with a subst: in a comment at the end (the advantage of which is that we can at least put in wikitext comment warnings), is going to be showing up in the wikitext of a LOT of pages, some of which will inevitably be edited by people without the least template clue.
:::::I think that the max phonemes in a single IPA_popup template should be well under 50, maybe more like 20. Or, I just had another idea - have the template recursively call itself (through a redirect) if it gets more than a certain number of arguments (say, 10 at a time, with a maximum recursive depth of 6, for a total of up to 60), that way if it DOESN'T get so many arguments it doesn't build bulky #if statements around the missing ones. It would have to tell itself to use a different redirect for each level of recursivity, but that's simple. Also the place where it called itself would have to be tricked up so that it wouldn't evaluate unless the call was real - again, simple, once you recognize the problem. --[[User:Homunq|Homunq]] 02:50, 24 January 2007 (UTC)⏎
⏎
== Tradeoff: completeness vs. accessibility ==

The above "comments from a linguist" argue that no popup text can ever be complete enough to really get across unknown sounds, thus linking is needed. But having '''just''' a link, as is suggested there, seems to me totally to miss the point of this idea, which is to have at least some data on a given character available in the same place the character is used, without painstaking clicking through, looking up, and reasse(contracted; show full)
:With this in place try hovering over this "{{IPA_hover|a}}". [[User:Jimp|Jimp]] 07:57, 23 January 2007 (UTC)

:: What about case by case? Do the javascript and hovers for simple sounds, and let the article authors decide to break some sounds out of it and provide links to that IPA symbol's article —[[User:Random832|Random832]]<small>2007-01-23T22:56:12[[User talk:Random832|UT]][[Special:Contributions/Random832|C]](01/23 17:56EST)</small>