We have a few functions that are either only defined on char16_t versions of string classes or only on char versions. So for example sometimes 'Find' takes a string and offset that defaults to -1, but other times 'Find' takes a string and then a bool param indicating whether to use a case insensitive comparison or not and *then* an offset. So something like: > int32_t Find(const nsTString<char>& aString, bool aIgnoreCase = false, > int32_t aCount = -1) const; > > template <typename EnableIfChar16 = IsChar16> > int32_t Find(const self_type& aString, int32_t aOffset = 0, > int32_t aCount = -1) const; I've found a few bugs where users were expecting just the `Find(str, offset)` version and were calling it with something like: > // expects to use the offset version, but hits the `aIgnoreCase` version. > nsCString aCStr; > aCStr.Find(str, itr - start); I think we'd be better off removing the `aIgnoreCase` version and adding something like `FindIgnoreCase` that only works with char params.