Hello,
A 7-Year-Old Bug Still Unresolved
If a bug report has been open for 7 years, that typically signals one of three things:Low prioritization by the product team.Complexity or ambiguity in reproducing or resolving the issue.Inadequate ownership, possibly because the feature depends on a third-party or outdated module.t's fair to ask the vendor:
Has this bug been triaged or put into a backlog?
Is there any plan to rework or replace the underlying language detection engine?
2. Vague Definition of "Short Message"
A reliable language detection engine should work even on short texts, or gracefully fallback when confidence is low. The fact that it fails without a clear guideline on what constitutes a "short" message suggests:
The algorithm relies too heavily on heuristics (like word frequency).
There's no minimum confidence threshold or fallback to English.For example, a message like “Hi, please see the attached” might be mistaken for Japanese simply because it’s too short to match against enough English features.3. False Positives with No Japanese Characters
This is the most critical failure. A language detector misidentifying English as Japanese without any Japanese characters likely stems from:Misweighted language models, especially for character sets that overlap with Latin punctuation.A default fallback to Japanese when detection confidence is low (possibly due to regional settings or past behavior). e-zpassnh com
