I encountered two problems today, and report them in one thread, because they might have the same root cause:
-
When I search in the course’s database for a Hindi word, the search doesn’t find it. For example:
कितना is in the database, for sure. But when searching on it can’t find it. Search on English works just fine -
When Bulk adding words to a level, it adds also entries that already exist. For example, “अध्यापक teacher” is already in the database (spelled exactly like that). Using Bluk add, using exactly this stroke, it creates a new entry. ***
-
See also the following threat: Incorrect Hindi answers missing a ‘letter’, incorrectly marked as correct
The common denominator in the above three seems to be, that in each case the Memrise systems need to compare two items in the Hindi script - one entered by the user to one stored in the Memrise systems. It seems that that evaluation does function correctly.
Does anybody have similar experiences?
Would you know a solution/workaround?
*** an explanation on the database: when we Bulk add new entries, the Memrise-computer (or whatever) searches through the database to see if there are any entries that have the same value in the foreign language field (e.g. Hindi) and the English language field (or whatever the language of ‘origin’ is).
If the foreign/English combination does not exist, it creates a new entry (or record) in the database, and adds a pointer to the level pointing to the new record.
If the entry does already exist, it only creates a pointer in the level - the results in all other data, such as audio file, part of speech and gender, as well as any learner’s history for that record, are also present in that level.
I have tested that today to make sure that it indeed (still) works that way.