When strings are copy+pasted from weird sources like Office Documents, the string encoding will sometimes be ASCII-8BIT and causes 500 errors when endpoints try to process them. We should be able to fix this by converting values of params to UTF-8 strings in a new rack middleware that runs early in the chain (e.g. right after we remove non-whitelisted params)
When strings are copy+pasted from weird sources like Office Documents, the string encoding will sometimes be ASCII-8BIT and causes 500 errors when endpoints try to process them. We should be able to fix this by converting values of params to UTF-8 strings in a new rack middleware that runs early in the chain (e.g. right after we remove non-whitelisted params)