[Fwd: Re: [jcifs] CIFS specs used for developing JCIFS]

[Allen, Michael B (RSCH)]

Actually I think the "developer reference" which I believe their calling the SNIA document now is
pretty good but people have complained that it's not nearly good enough (including myself). In
reality the SNIA document go most of it's weight from the original Leach document:

http://ubiqx.org/cifs/rfc-draft/draft-leach-cifs-v1-spec-02.html

which is a very good start in itself. Unfortunately people in the CIFS community really haven't
stepped up to contribute what little bit's they've figured out. They also were obviously counting
heavily on MS participation but they seemed to back out of the discussion after a while. jCIFS
is written pretty much entirely from the original Leach version above although there's critical little
stuff (named pipes) in the SNIA doc. Also, despite the fact that this particular Leach version
above has 'v1' in the title, it has more information in it about transactions that Leach admitted
was inadvertently lost at a later date somehow (Please never "upgrade" this link Chris!). I
notified the SNIA guys of the discrepancy and I believe they changed it in the SNIA draft but I
have not looked at the latest.

You might also look at the jCIFS's source. There aren't too many comments but it's well
organised and should make sense with the "technical reference" in hand. I always update the
online code every time I upload a new version:

http://auditorymodels.org/jcifs/src/src/jcifs/

If you have any questions, I'd be happy to answer them here or post them to the Microsoft list
that Chris gave a link to below. I'm on that list too and will respond provided I know the
answer :~).

Mike

> -----Original Message-----
> From:	Christopher R. Hertel [SMTP:crh at ubiqx.mn.org]
> Sent:	Sunday, January 13, 2002 5:34 PM
> To:	jcifs at samba.org
> Subject:	[Fwd: Re: [jcifs] CIFS specs used for developing JCIFS]
> 
> "PALANIVEL,PRABHAKAR (HP-India,ex2)" wrote:
> > 
> > Hello,
> > 
> > Could you please send the CIFS specs used for JCIFS. I saw
> > the specs  in SNIA. But there are a few things which are left out and
> > also the formats for some calls differ.
> 
> Prabhakar,
> 
> We do this by reverse-engineering the protocols, and by exchanging
> information on this list, the Samba-Technical list, the Samba-TNG list, and
> the CIFS mailing list run by Microsoft.  There is also a conference once a
> year--we all get together and whine about how annoying this work is.
> 
> There *ARE NO* specifications for CIFS beyond what you have already seen. 
> Really.  I promise.
> 
> CIFS is a Microsoft protocol suite.  It is nothing like NFS.  There are no
> committees verifying changes to the protocols, no public comment, nothing. 
> Working with CIFS is ugly and painful and annoying and frustrating and it
> makes you grow a horn on your knee and fuzzy rings of green fuzz around your
> elbows.  And your nose will twist to the left.
> 
> Note, by the way, that many of the Samba Team members currently have
> consulting positions at HP.  HP is a great supporter of Samba (as are many
> other large companies including SGI, IBM, etc.).
> 
> Anyway, if you have specific questions you can ask them here, but the CIFS
> list is probably the better place for such questions.  People in the CIFS
> community tend to be very busy people, so be patient about getting answers.
> 
> Samba-Tech: http://lists.samba.org/listinfo/samba-technical/
>  Samba-TNG: http://www.samba-tng.org/mailinglists.html
>    MS-CIFS: http://discuss.microsoft.com/archives/cifs.html
> 
> I also recommend that you become familiar with Microsoft's NetMon packet
> capture program, and the excellent Ethereal sniffer.
> 
> Ethereal:  http://www.ethereal.com/
> 
> Good luck.
> 
> Chris -)-----
> 
> -- 
> Samba Team -- http://www.samba.org/     -)-----   Christopher R. Hertel
> jCIFS Team -- http://jcifs.samba.org/   -)-----   ubiqx development, uninq.
> ubiqx Team -- http://www.ubiqx.org/     -)-----   crh at ubiqx.mn.org
>