undocumented(7) and lintian

[Chris Waters <xtifr@dsp.net>]

Hi,

After some of the discussions here, it occurred to me that having
lintian be quiet about the use of undocumented(7) might not be a good
idea, since such use *is* a bug.  So I filed a bug report against
lintian (#55081), asking for it to issue a warning about the use of
undocumented(7).

Gecko promptly rejected this and closed the bug report, saying that
policy allows the use of undocumented.  I responded by pointing out
that, while policy allows it, it requires you to have a bug report on
file, so it's still a bug.

Of course, both of us are correct.  (This is the *only* case where
policy blesses something it calls a bug.)  And it's his package, so
I've left the bug report closed for now, even though I'd like to see
it made a wishlist item if nothing else.

He also suggested that I bring the matter up on this list.  

Now, it's true that we have a proposal before the group already to get
rid of undocumented(7) (#39830).  And I think that Gecko may be hoping
that this proposal will pass (as am I).  But I'm not confident -- even
though all the objections have been answered, every time the proposal
is mentioned, someone else jumps in to object (usually with the same
tired old objections).  Which makes it a little difficult to
demonstrate any sort of concensus.  The proposal is currently marked
as "stalled".

I thought that having lintian warn about undocumented(7) would be
useful as an interim measure, and might help raise the percentage of
packages that actually *have* a bug report on file when they use
undocumented.  Might, in fact, help raise the general awareness that
the use of undocumented(7) is not a final and complete solution to the
problem of no man page.

So, I thought I'd bring the matter up here.  If enough people on this
list agree with me, then maybe Gecko can be persuaded to change his
mind.  Or, contrariwise, if enough people agree with him, maybe I can
be persuaded to shut the <bleep> up!  :-)

So, what do you all think?

cheers
-- 
Chris Waters   xtifr@dsp.net | I have a truly elegant proof of the
      or    xtifr@debian.org | above, but it is too long to fit into
http://www.dsp.net/xtifr     | this .signature file.