view-source:http://w3c-test.org/web-platform-tests/master/conformance-checkers/html/parser/004-novalid.html The  character should be flagged as an error. Other test cases that exhibit the same problem are at: http://w3c-test.org/web-platform-tests/master/conformance-checkers/html/parser/ e.g., view-source:http://w3c-test.org/web-platform-tests/master/conformance-checkers/html/parser/001-novalid.html (there's a U+000B character in the body of the document)
The basic problem here is that in Gecko, the code for reading a character is mere array access without checking the character for bogosity. I supposed we could look into using template magic to compile a different "read character" code path for View Source. But in that case, we'd then need a way to attach errors to individual characters. So this is intentional in the sense that I knowingly didn't take the time to implement all this, but in principle, it would be nice to have code for this.
(In reply to Henri Sivonen (:hsivonen) from comment #1) > So this is intentional in the sense that I knowingly didn't take the time to > implement all this, but in principle, it would be nice to have code for this. OK, understood. I raised it in spirit with the same nice-to-have principle you mentioned, so I don't feel strongly about it (certainly not strongly enough to take time to write up a patch for it myself So, given all that, I think moving this to resolved=wontfix would be fine. I don't think it's worth keeping open if there's not a need/activity to refine the behavior any time soon to add it.
You need to log in before you can comment on or make changes to this bug.