On Mon, 2002-03-18 at 04:45, Buchan Milne wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Brad Felmey wrote: > | On Fri, 2002-03-15 at 04:53, Buchan Milne wrote: > | > | > |>It is related to code-pages etc, so just because it works for does not > |>mean it is fixed. I have not been able to reproduce the problem either. > |>Brad, have you tested with samba-2.2.3a-7mdk or later? > | > | > | Where? IINM smbfs doesn't require or even interact with Samba. > | Clientside is not all Mandrake, and I know for a fact that this still > | oopses against Debian. Why is a newer Samba going to make a difference > | to the kernel? > > smbfs uses smbmount (part of samba) to negotiate the connection. The bug > is with character translations of smbfs, and -7mdk added new > codepages/character sets, which someone has reported as fixing their > problem with smbfs. I don't think this is the problem (trying the > 2.4.18-7mdk kernel from Juan's site should be the real fix), but since I > can't reproduce your problem, it would be nice to know for sure.
Actually, according to samba.org, smbmount is _not_ part of samba, it's just bundled for convenience. They want no part of it, and are pushing their smbsh solution (which hasn't been developed beyond the sucking stage yet). AFAIK smbmount would work with or without samba. (?) Regardless, 8.2 final is being released with this dumbness intact, so I'm going to hold off any installs in our production environment until the updates to fix this are released. I've already spent some considerable hours working on this and around this to give the developers info so they could fix it before 8.2. Since they've (unwisely, IMO) chosen not to, I'm not going to even mess with it until they take businesses seriously and fix this glaring flaw. -- Brad Felmey
