2011-02-08 16:44 AEST by Arthur Barrett - Add SSPI/GSERVER support to native CVS Eclipse plugin, this is a feature request on the Eclipse bug site:
Bug 41097 - [Connection] Support for SSPI & gserver CVSNT protocols
https://bugs.eclipse.org/bugs/show_bug.cgi?id=41097

2011-11-28 22:00 AEST by Arthur Barrett - *** Bug 6313 has been marked as a duplicate of this bug. ***

2011-12-04 06:41 AEST by Arthur Barrett - Modification to previous comment (now redacted).

The patch from "Jene Jasper" on the Eclipse bug is not an SSPI client at all, it simply supplies an SSPI 
connection that wraps the existing EXT connection - ie: it still requires extnt.exe to function.  He 
explains that this was written for compatibility with sandboxes that already had :sspi: in CVS/Root...

The patch from "Alex Blewitt" titled "100% Java plugin that allows GSSAPI access via GServer for any OS 
with Java 1.4" appears to be an attempt at a 'real' GSSAPI client - it should work with gserver protocol 
from CVS or CVSNT servers - I've not tested this.  Since I've now got the redhat4 VM working as both a 
CVSNT Server (gserver) and CVSNT Client (gserver) I should test it there using Eclipse 3.1...  Authors 
notes:
http://alblue.bandlem.com/2006/02/eclipse-gserver-support-for-cvs.html


The patch from "ian" titled "Modified version of Jacob Barret. It now compiles in VC++6 and does not 
require VC++ 7 " appears to be an attempt at a 'real' SSPI client for 'NTLM' protocol only (also there is a 
patch in the 'text' of the bug from 'ben_o_it' to change the codepage from 'ASCII' to 'ISO-8859-1'.  Since 
this is not very secure, and specific to CVSNT I think it's not a particularly interesting patch.  I've no idea 
what the VC++ 7 stuff refers to - the jar file is pure java and I cannot see any references to JNI...

There is also a plugin for SSERVER support - again Eclipse 3.1 era:
http://home.arcor.de/rolf_wilms/cvsssl/cvsssl_help.html