Check whether the outcome improved
Check the supported character matrix
Exercise accented text, non-Latin characters, punctuation, and symbols across the encodings the product documents. Include a plain Latin control record. Compare actual characters, not just rendered width or a successful row count, because silent replacement can leave both of those apparently correct.
Carry the choice through confirmation
Where encoding selection exists, change it in preview and confirm the chosen value reaches execution. Retry the same fixture with automatic detection. Inspect the final record and a supported retrieval or export surface to establish whether preview and persisted content share the same decoding behavior.
Exercise invalid byte handling
Import a fixture containing an invalid sequence for the declared encoding. Confirm the error identifies a useful location and correction path. A warning should not disappear while the importer silently substitutes a character that the user never approved as part of the source content.
What to carry forward
Accept when supported text remains intact and invalid sequences follow the documented failure behavior. Record the exact encoding matrix and retrieval surfaces exercised. Keep untested producer formats explicit; passing one accented name does not establish support for every byte convention or language used by future imports.
Technical background: PostgreSQL documentation.
Keep the decision with the work.
Use a Work Item in Aglet to record the problem, the evidence you have, and the next decision. Add an owner and priority, then keep updates in the discussion so the next person can follow the reasoning.
Create an account See the product workflow