Nick Kew wrote:
On Sun, 30 Dec 2007 09:27:48 -0600
"William A. Rowe, Jr." <[EMAIL PROTECTED]> wrote:
Since FTP doesn't provide us a clue of what flavor is which, for the
purpose of a proxy ftp representation, we'll be adding a directive
that you can apply site-by-site. Or globally, if you are convinced
that your FTP sites are mostly/nearly all utf8. That's fine.
Any objection to adding the attached (or equivalent) patch?
That would fix the regression the original introduces.
Quick comments, _ls_ is fuzzy - call it _list_ (or _nlst_ ;-)
Inheritance/default is broken in your config suggestion, you don't
have a fallback, but assigning a default value in create_conf would
hinder determining when to inherit vs. when to override. And third,
this patch was already cleanly written and available, in fact ...
Rudiger, would you please commit your patch, now that the original
patch is disclosed (we wanted them sequenced appropriately).
Oh, yes, here's why it's a regression. For clients that default to
ISO-8859-1, the patch is a no-op. For a client that guesses, and
guesses right, it may cease to work where it worked before. Users
who get hit by that will see a regression, and the least we can do
is make it fixable for them.
Ok - so nonconforming clients see a 'regression'. Sorry I just don't
buy that as a showstopper ;-) But given that this failure by a pmc
member to follow the security list, and the veto RFC behavior already
has delayed Jim's RM plans, let's get it over with and add the config
that Rudiger has authored for very, very fast discussion.
Bill
