I think I see what it is: it's fine when you separate the phonemes, but not when they're run together, as no conversion takes place. But you're recommending that people not separate the phonemes for English, which means that the stress marks will come out wrong. kwami (talk) 20:41, 10 April 2009 (UTC)[reply]
Well, even if you don't separate the phonemes, you can still separate the stress marks, becuase I've defined those to work independently of language. So if you type an apostrophe or comma and separate it with pipes, then I think it should work right (as in the last example given on the template page). But if you don't separate it, then there will be no conversion, so you should use the IPA stress-mark symbol as you would with the other IPA templates (in fact there isn't much point using this template if you're not going to separate anything).--Kotniski (talk) 11:07, 14 April 2009 (UTC)[reply]
A separate point: I use AWB to cleanup IPA transcriptions, so that they're consistent with the IPA keys. Much of that is language specific, so it's convenient to have separate templates for each language, to customize the search, and to avoid excessive server load from preparsing, assuming preparsing is even working. But now this template transcludes to over 25,000 articles, which is the limit that AWB will download, so it can no longer be maintained. Not a problem if they're all parsed spellings, but many of them are actually entered as IPA, and not all of it's kosher.
Caught an IPA error in the auto spell conversion that went through the Polish articles: place names that begin with Tcz.kwami (talk) 16:56, 18 July 2009 (UTC)[reply]
As I think I mentioned before, I'm slowly working through the {{IPA-pl}} instances, converting those that use parsed spellings to {{IPAr|pl}}. When that's done, we should have two distinct sets: IPA-pl instances using explicit IPA, and IPAr|pl using parsed spellings. I hope that will solve the maintenance problems. What do you think is the problem with Tcz? I believe (and I've just asked a Pole who confirms this) that initial <tcz> is pronounced [tt͡ʂ], i.e. not the same as initial <cz>.--Kotniski (talk) 18:28, 18 July 2009 (UTC)[reply]
Hang on though, I've just noticed that at Gdansk, you've changed an IPAr instance with parsed spelling back into explicit IPA, using {{IPA-pol}}. Is there some reason for this? Is it because there's an audio file involved? (Incidentally, I don't think the stress mark is necessary for a monosyllable.)--Kotniski (talk) 18:38, 18 July 2009 (UTC)[reply]
Ah, I see. Sorry. I didn't realize you were doing this manually. I'm simultaneously pulling out the explicit pronunciations (the smaller set) to IPA-pol using AWB. Finally figured out how to scan for multiple pipes. Once that's done, we should be able to automate a mass conversion of the remaining IPA-pl to IPAr|pl, and then move IPA-pol back. (Hopefully it'll be less work this way.) I take it then there is no explicit IPA at IPAr|pl? This will solve the maintenance problem only as long as there is no explicit IPA at IPAr in any language, since AFAIK AWB cannot pull a limited list based on switches.
As for the Gdansk move, I've moved a couple over when they're accompanied by IPA-xx for other languages. I've moved them back.
I like added overt stress to stressed monosyllables for two reasons: it's more accurate, and I've noticed that people tend to omit stress even in phrases of monosyllables where stress info is required for pronunciation, as at Stanisław Lem, which was only stressed on the ni. I think they get used to leaving it out as understood when the word or name is in isolation, and don't realize they can't do that when it's not in isolation.
I left Gdańsk at Audio-IPA-pl, for ease of access of IPA with audio. I don't know if that will ever be needed, though, so go ahead revert further to IPAr|pl if you like. I think Danzig should be left at IPA-de, though, so it can be updated with the other German transcriptions if that IPA key is ever modified. kwami (talk) 20:21, 18 July 2009 (UTC)[reply]
OK, I think I see what's happening. Perhaps we should modify I've modified the documentation for this template to say that it is only to be used for respellings (and hence only for Polish at the moment), not for explicit IPA. Sorry, I didn't mean to imply I was doing the IPA-pl->IPAr|pl conversion completely manually (that really would take years) - I'm doing it with the help of my bot, along with a few other tweaks, but a certain amount of manual intervention is still required for the odd cases, which is why it's going to take quite a few more weeks until it's finished.--Kotniski (talk) 12:29, 19 July 2009 (UTC)[reply]
This edit request has been answered. Set the |answered= or |ans= parameter to no to reactivate your request.
Change the default option to generate an error, like at {{c-ja}} for Japanese. Will need to add trivial conversions (see similar recent edit to {{c-hu}}). — kwami (talk) 00:21, 14 February 2014 (UTC)[reply]
Undone: This request has been undone. Okay, I've undone the change you requested. Please put a working example in the /sandbox with working /testcases. I've deactivated this request until that is done and am watching this page now. — {{U|Technical 13}}(t • e • c)22:34, 14 February 2014 (UTC)[reply]
Done. Also needed Template:IPAc-pl/sandbox. Tested in mainspace, as sandboxes won't show if the error-tracking works.
I see a difference in the testcases page. I have my vector.css set to color redirecting links green, and the sandbox version does not link to the redirect. So, the link should be changed from Wikipedia:IPA for Polish to Help:IPA for Polish in the main template. — Eru·tuon05:33, 31 August 2016 (UTC)[reply]
Moved from Template talk:C-pl
