On Wed, Feb 21, 2001 at 12:39:08PM +1000, Anthony Towns wrote: > > > > E!: non-FHS-directory > > > E-: missing-manpage > > > E?: standards-version-uses-4-digits-not-3 > > when I rewrite lintian (started yesterday) the lintian messages will match > > policy: > > Error (E:) -- violate a MUST > > Warning (W:) -- violate a SHOULD > > XXXXXXX (?:) -- a MAY is not followed > > Currently, aiui, lintian uses E: for problems that it's sure are mistakes, > and W: for problems that it's only guessing are mistakes. I think that > division is still useful. >
no, it tries to do this based on 2.x level MUST/SHOULD and the authors beliefs of severity. Has nothing to do with the sureness of the test. > katie or testing could legitimately automatically reject packages with > E! lintian errors, but not E- or W!, eg. > lintian will never be able to return a sure judgement. Manoj's packages confuse it thoroughly, but on hand inspection I am sure they follow policy. Every message lintian outputs should be checked manually and by a re-read of policy. It is trying to discern what a human meant. In the realm of coding, people do all kinds of crazy things and lintian can only cope so well. Assume every message is 'X-:'. A Package with an E: should be marked for human inspection at best. James Troup has stated that when I trust lintian he will consider hooking it into dinstall. I think this is a good thing. It is my hope to have lintian to a sane state by summer (July-ish). Wichert wants something in 3 months for the FSG. Not sure if the code base will make that, but I will try.
