Bug in Tools: 1. Pharo Image
• You are subscribed to this case
This is not an issue.
The first ByteArray is correctly encoded in uft8. You can get the same result evaluating.
'Les élèves français' asUTF8Bytes.
or
ZnUTF8Encoder new encodeString: 'Les élèves français'
The second ByteArray is the Latin-1 encoding of the same String. 'Les élèves français' asByteArray returns this byte array, but this is a just a coincidence, as this depends in a set of many things. Starting the asByteArray uses the code point of the character and tries to put it in the resulting ByteArray.
You should never convert Strings to ByteArray (or back) without using an encoding.
The inspector is trying to do the best to recognize some data in the ByteArray to show it, so it tries with two common encoding (latin-1 and UTF8).
The second ByteArray is not properly encoded in UTF8, so it is ok not to show.
Please refer to the documentation of ZnEncoders or general String Encoding techniques. If there is any doubt please ask in the mailing list.