User talk:JohnC5/Sandbox4

Could y'all check these?
, hey, could y'all check these for me? I haven't finished the consonant stems yet, but this is a start. It is based primarily on Whitney. Also, any ideas you have about the display would be great. — JohnC5 08:12, 6 October 2017 (UTC)
 * This is amazing... mādhavpaṇḍit (talk) 08:44, 6 October 2017 (UTC)
 * The final things that remain are these declensions:
 * Root nouns (stem form: -VC)
 * Participles in -ant or -at (stem form: -at)
 * Possessives in -mant or -vant (stem form: -mat and -vat)
 * Perfect Participles in -vāṅs (stem form: -vas)
 * Comparatives in -yāṅs or -yas (stem form: -yas)
 * As you can see, there is some ambiguity in stem form between these declensions. I would prefer to enter alternative stem forms into the template to disambiguate, namely:
 * Root nouns (remains -VC)
 * Participles in -ant or -at (now -ant)
 * Possessives in -mant or -vant (now -mānt and -vānt, or -ant and 1)
 * Perfect Participles in -vāṅs (now -vāṅs)
 * Comparatives in -yāṅs or -yas (now -yāṅs)
 * This will mean that the user will be required to manually enter this alternative stem form, as opposed to having the page auto-generate. There still an issue that some root nouns can overlap with the other declensions. This will probably need to be handled with some sort of override parameter (e.g. 1, etc.). We will also have to create a separate system for irregular nouns, and I'll need to modify these to correctly do the adjectival forms correctly as well. I would really appreciate your feedback. — JohnC5 10:09, 6 October 2017 (UTC)
 * This is absolutely fantastic... I don't see any errors atm. —Aryaman (मुझसे बात करो) 12:07, 6 October 2017 (UTC)
 * Any ideas about this proposal?The alternative proposal would be a series of parameters like 1, 1, 1, 1, 1 along with pattern recognition from the "normal" stem forms. I think this might be better. This would mean the system would look like:
 * Root nouns (stem form: -VC and 1)
 * Participles in -ant or -at (stem form: -at and 1 or 1)
 * Possessives in -mant or -vant (stem form: -mat / -vat and 1)
 * Perfect Participles in -vāṅs (stem form: -vas and 1 or 1)
 * Comparatives in -yāṅs or -yas (stem form: -yas and 1)
 * I think this might be more intuitive and I can force the user to enter a parameter when making a decision. What do y'all think? — JohnC5 02:08, 10 October 2017 (UTC)
 * I agree, the user should have to input a parameter for those special cases. This is really great, thanks for all the work you put into this (and grc-decl)! —Aryaman (मुझसे बात करो) 13:58, 10 October 2017 (UTC)
 * I trust your judgement! Your efforts are appreciated. I love this... -- mādhavpaṇḍit (talk) 16:32, 10 October 2017 (UTC)