On Ubuntu in 2008, printing from OpenOffice failed on Tuesdays and worked on every other day. The cause was one line in the `file` command's magic database. A rule meant to match the twenty-nine byte string "Tue Jan 22 14:32:44 MET 1991" at offset 4 had its spaces left unescaped, so the comparison stopped after three bytes: any file with "Tue" at offset 4 was reported as an Erlang JAM file. A print job carrying the day name near the start of the file was therefore misidentified one day in seven, and the printing path acted on the wrong type.
Editorial judgements on a 0 to 10 scale, argued in the notes below and nowhere else. They order a list. They are not evidence and they never appear in a claim.
Never let a sniffed type decide behaviour when a declared type is available, and when you must sniff, treat a match as a hypothesis: log what matched, and fail closed on an ambiguous one rather than acting on the first rule that fires.
WHAT IS ESTABLISHED. The defect, from the bug report itself: the magic line `4 string Tue Jan 22 14:32:44 MET 1991 Erlang JAM file - version 4.2`, the unescaped spaces, the effective three-byte comparison, the minimal reproduction `echo "1/2 Tue" >> file && file file`, the affected versions (present in 4.21 and 4.24, absent in 4.23, introduced in 4.20-4), and the fix in Debian file 5.01-1 by escaping the spaces. Fix Released across four Ubuntu series, linked to Debian #514056. WHAT IS ATTESTED BUT NOT MEASURED. That the symptom was a weekly printing failure, which rests on the bug title and on users in the duplicate who observed it before knowing the cause, one of whom demonstrated it to a sceptical spouse. Good testimony, not a measurement. WHAT IS NOT ESTABLISHED HERE, and is the reason this record does not go further. The exact byte at offset 4 of the file OpenOffice handed to the printing path in 2008. The mechanism requires the day name to sit there, which is consistent with a date written near the head of a generated file, but no 2008 spool file was examined for this record. A modern reconstruction is not evidence about a 2008 one. WHY IT IS RATED reproduced ANYWAY. The experiment reproduces the part the claim turns on, that an unescaped magic rule matches on the day name alone, with a positive control, the escaped rule, and a negative control, a file carrying a different day. The reproduction covers the defect, not the deployment. SCORES, opinions. wtf 10: the highest in the catalog. A program that works six days in seven is the purest form of this subject, and the user who thought they had gone mad is the correct reaction. denial 7. damage 4.