2010-02-26 SVN: new release * Release: 0.4.8 2010-02-26 11:13 cc@INF.ED.AC.UK * ChangeLog, lcfg.yml, specfile: supports F12 2010-01-22 10:23 squinney@INF.ED.AC.UK * ChangeLog, lcfg.yml: lcfg-om release: 0.4.7 2010-01-22 10:23 squinney@INF.ED.AC.UK * specfile: Added new build requirement on perl(Readonly) for the tests 2010-01-22 10:14 squinney@INF.ED.AC.UK * ChangeLog, lcfg.yml: lcfg-om release: 0.4.6 2010-01-22 10:14 squinney@INF.ED.AC.UK * t/01_load.t: Tidied up new tests 2010-01-22 10:12 squinney@INF.ED.AC.UK * ChangeLog, lcfg.yml: lcfg-om release: 0.4.5 2010-01-22 10:12 squinney@INF.ED.AC.UK * lcfg.yml, lib/LCFG/Om/Environment/NewAFSPAG.pm.cin, specfile, t, t/01_load.t: fixed syntax errors in the changes to the NewAFSPAG environment module 2010-01-22 09:22 squinney@INF.ED.AC.UK * ChangeLog, lcfg.yml: lcfg-om release: 0.4.4 2010-01-22 09:22 squinney@INF.ED.AC.UK * lib/LCFG/Om/Environment/NewAFSPAG.pm.cin: om did not work in the LCFG installer with the NewAFSPAG environment module. There is no easy way to set the om_environment for every component for specific contexts. Given all this we have gone back to doing a require() on the AFS modules and just warning if they are not loadable. 2009-12-04 09:09 squinney@INF.ED.AC.UK * ChangeLog, lcfg.yml: lcfg-om release: 0.4.3 2009-12-04 09:09 squinney@INF.ED.AC.UK * specfile: Turned off the automatic requirement finder and set them by hand to avoid having the AFS perl module be obligatory 2009-12-04 09:08 squinney@INF.ED.AC.UK * README.BUILD: Mentioned the additional required Perl modules and where to get them from 2009-12-03 11:55 squinney@INF.ED.AC.UK * ChangeLog, lcfg.yml: lcfg-om release: 0.4.2 2009-12-03 11:55 squinney@INF.ED.AC.UK * ChangeLog, lcfg.yml, lib/LCFG/Om/Environment/NewAFSPAG.pm.cin: The LCFG::Om::Environment::NewAFSPAG module now requires that AFS and AFS::Utils are available. If the platform isn't supported by the Perl AFS module then don't set om_defaults.environment to this module. Also we now check if openafs is actually running when the setpag fails. I can't see any better way to avoid spurious failures 2009-10-29 15:32 squinney@INF.ED.AC.UK * CMakeLists.txt, specfile: Made om setuid with cmake instead of the specfile 2009-07-06 10:40 squinney@INF.ED.AC.UK * ChangeLog, lcfg.yml: lcfg-om release: 0.4.1 2009-07-06 10:40 squinney@INF.ED.AC.UK * lib/LCFG/Om/Environment.pm.cin: Removed warn message which was left over from testing phase 2009-06-08 09:00 squinney@INF.ED.AC.UK * ChangeLog, lcfg.yml: lcfg-om release: 0.4.0 2009-06-08 08:59 squinney@INF.ED.AC.UK * CMakeLists.txt, specfile: Actually install the new schema and perl modules 2009-06-08 08:58 squinney@INF.ED.AC.UK * om.cin: Modified om to use the new environment initialisation module API 2009-06-08 08:58 squinney@INF.ED.AC.UK * lib, lib/LCFG, lib/LCFG/Om, lib/LCFG/Om/Environment, lib/LCFG/Om/Environment.pm.cin, lib/LCFG/Om/Environment/NewAFSPAG.pm.cin: Added new environment initialisation modules 2009-06-08 08:57 squinney@INF.ED.AC.UK * om_defaults.def.cin: Added new om_defaults schema to hold system-wide defaults for om 2009-06-08 08:56 squinney@INF.ED.AC.UK * om.def.cin: Added om_environment resource and removed ancient, deprecated om_user resource 2009-06-05 14:33 squinney@INF.ED.AC.UK * ChangeLog, lcfg.yml: lcfg-om release: 0.3.24 2009-06-05 13:33 squinney@INF.ED.AC.UK * om.cin: Replaced all the validation/untainting with checks that die if there is invalid data rather than silently continuing 2009-06-05 13:31 squinney@INF.ED.AC.UK * om.cin: The assembled ssh command is really a scalar not a list 2009-06-02 13:31 squinney@INF.ED.AC.UK * ChangeLog, lcfg.yml: lcfg-om release: 0.3.23 2009-06-02 12:49 squinney@INF.ED.AC.UK * om.cin: Properly configured Getopt::Long to pass_through options 2009-06-02 12:46 squinney@INF.ED.AC.UK * om.cin: Configured Getopt::Long to pass_through options 2009-06-02 12:43 squinney@INF.ED.AC.UK * om.cin: Fixed om() so that it actually uses the canonicalized values for the host names when the machine is not 'localhost' rather than throwing them away. This fixes https://bugs.lcfg.org/show_bug.cgi?id=145 2009-06-02 12:16 squinney@INF.ED.AC.UK * om.cin: fixed precedence issue with comparison of return value from index() 2009-06-02 12:13 squinney@INF.ED.AC.UK * om.cin: fixed problem with using index to search for a '.' in the machine.component string 2009-06-02 11:51 squinney@INF.ED.AC.UK * specfile: Also include pod in the lcfgpod directory 2009-06-02 11:49 squinney@INF.ED.AC.UK * specfile: man page is installed into section 1 not 8 2009-06-02 11:48 squinney@INF.ED.AC.UK * CMakeLists.txt, om.cin, specfile: updated copyright years. Package building now uses the new pod documentation 2009-06-02 11:43 squinney@INF.ED.AC.UK * om.cin: Improved command line handling. Added POD documentation. Added a --help option which prints out usage from POD. 2009-06-02 11:08 squinney@INF.ED.AC.UK * om.cin: Rewrote the command assembly part of remoteom so it is a lot nicer. Allowed digits in the authagent module name. Extracted mode for component file check to a readonly variable to make it clearer what is happening 2009-06-02 10:33 squinney@INF.ED.AC.UK * om.cin: Extracted '@LCFGCOMP@' into a readonly variable, this might be done better with LCFG::SysInfo to look up the information at runtime 2009-06-02 10:26 squinney@INF.ED.AC.UK * om.cin: Added svn/rcs info in header. All POSIX calls now show the OS error message if they die. Ran the code through perltidy 2009-06-02 10:09 squinney@INF.ED.AC.UK * om.cin: Fixed importing of EXIT_SUCCESS constant from POSIX module 2009-06-02 10:09 squinney@INF.ED.AC.UK * om.cin: Compare return values from POSIX functions to the EXIT_SUCCESS constant rather than -1. Various other minor tidying 2009-06-02 09:58 squinney@INF.ED.AC.UK * om.cin: Replace module-loading within an eval with UNIVERSAL::require. Also replaced all uses of built-in variables with their 'English' equivalents 2009-06-02 09:33 squinney@INF.ED.AC.UK * om.cin: Switched from double to single quotes where appropriate 2009-06-02 09:14 squinney@INF.ED.AC.UK * om.cin: Altered the usage of various modules so they do not automatically pollute the namespace 2009-03-12 15:05 squinney@INF.ED.AC.UK * om-wrapper.c.cin, om.cin, om.def.cin: Massive setting of svn:keywords property on component scripts, schemas, pod files and Perl modules 2009-03-09 16:43 squinney * lcfg.yml: Removed hardwired version-control type from lcfg.yml to allow future transfer to subversion 2008-07-30 09:50 squinney * specfile: Do not install README.BUILD as part of RPM 2008-07-30 09:44 squinney * README.BUILD: Updated package build instructions 2008-07-18 13:34 squinney * ChangeLog, lcfg.yml: Release: 0.3.22 2008-07-18 13:32 squinney * specfile: Forgot to add back in the om man page 2008-07-18 13:28 squinney * ChangeLog, lcfg.yml: Release: 0.3.21 2008-07-18 13:27 squinney * CMakeLists.txt, Makefile, README.BUILD, config.mk, lcfg.yml, om-wrapper.c.cin, specfile: Converted to the new build tools 2008-05-06 15:53 squinney * specfile: Replaced LCFGDEF with LCFGSERVERDEF 2007-08-08 14:15 cc * ChangeLog, config.mk: Release: 0.3.20 2007-08-08 13:43 cc * config.mk: sl5 support 2007-08-08 12:34 cc * ChangeLog, config.mk: noted SL5 support 2007-06-29 10:57 squinney * ChangeLog, config.mk: Release: 0.3.18 2007-06-29 10:57 squinney * ChangeLog, config.mk, om.cin, specfile: Added patch from Simon to allow underscores in component names 2007-02-13 13:58 squinney * ChangeLog, config.mk: Release: 0.3.17 2007-02-13 13:58 squinney * ChangeLog, config.mk: noted fc5 and fc6 support 2006-04-17 14:44 sxw * ChangeLog, config.mk: Release: 0.3.16 2006-03-28 21:45 sxw * Makefile, config.mk, om-wrapper.c.cin: Patches for Darwin support. Adds suid helper for os's which don't support suid scripts. Code by toby@ 2005-01-18 15:00 dcspaul * ChangeLog, config.mk: Release: 0.3.15 2005-01-18 15:00 dcspaul * ChangeLog, config.mk: - 2004-06-11 11:49 dcspaul * ChangeLog, config.mk: Release: 0.3.14 2004-06-11 11:48 dcspaul * om.def.cin: *** empty log message *** 2004-06-11 11:37 dcspaul * ChangeLog, config.mk: Release: 0.3.13 2004-06-11 11:37 dcspaul * Makefile, om.def, om.def.cin: We need to keep om_user around until all the clients have updated 2004-06-11 11:30 dcspaul * ChangeLog, config.mk: Release: 0.3.12 2004-06-11 11:30 dcspaul * ChangeLog: * Removed (obsolete) om_user * Added authorization reference to .def file * Added default methods to .def file 2004-06-11 11:27 dcspaul * ChangeLog, config.mk: Release: 0.3.11 2004-06-11 11:27 dcspaul * ChangeLog, config.mk, om.cin, om.def: *** empty log message *** 2004-03-31 10:23 sxw * ChangeLog, config.mk: Release: 0.3.10 2004-03-31 10:23 sxw * om.cin: Added support for ng_runuser and ng_rungroup, fixing Bugzilla #182 2004-01-05 22:31 sxw * ChangeLog, config.mk: Release: 0.3.9 2004-01-05 22:31 sxw * om.cin: d'Oh again 2004-01-05 22:22 sxw * ChangeLog, config.mk: Release: 0.3.8 2004-01-05 22:22 sxw * om.cin: d'Oh 2004-01-05 22:09 sxw * ChangeLog, config.mk: Release: 0.3.7 2004-01-05 22:09 sxw * om.cin: Fix perl's somewhat dubious marking of undefined variables as being tainted 2004-01-05 21:58 sxw * ChangeLog, config.mk: Release: 0.3.6 2004-01-05 21:58 sxw * om.cin: Avoid uninitialised variable warnings 2003-12-09 11:52 sxw * ChangeLog, config.mk: Release: 0.3.5 2003-12-09 11:49 sxw * om.cin: Fixes to allow remote om to work with strict taint checking 2003-11-13 11:11 sxw * ChangeLog, config.mk: Release: 0.3.4 2003-11-13 11:11 sxw * om.cin: Forcibly set a default umask - bug #103 2003-10-08 15:32 sxw * ChangeLog, config.mk: Release: 0.3.3 2003-10-08 14:18 sxw * Makefile, om.cin: Fixed 'clean' targets Added support for remote om (simple and ssh based for now) Blindly detaint all arguments (fixes bug #31) Don't attempt to resolve hostname if its 'localhost' (fixes bug #47) 2003-10-03 22:28 sxw * om.cin: Fix for Bug #31 Add = to list of characters which are allowed through the taint check. This is safe for us because we guarantee that these go nowhere near a shell at our end. It's arguably dangerous if the root-run client isn't sufficiently careful. 2003-08-31 21:15 sxw * ChangeLog, config.mk: Release: 0.3.2 2003-08-31 21:14 sxw * Makefile: Add ChangeLog stuff, and restrictions on committing 2003-08-31 21:10 sxw * ChangeLog, config.mk: Release: 0.3.1 2003-08-31 21:08 sxw * Makefile, config.mk, om.cin, specfile: Apply Alastair Phipps patches for Solaris support 2003-07-11 10:06 sxw * ChangeLog, config.mk: Release: 0.3.0 2003-07-10 15:06 sxw * om.cin: Add \/ and \. to arguments untainting 2003-05-29 11:39 sxw * om.cin: Update regexes 2003-05-29 11:33 sxw * om.cin: Fixes for aggressive taint checking 2003-05-26 17:41 sxw * ChangeLog, config.mk, om.cin: Release: 0.2.0 2003-05-26 17:40 sxw * ChangeLog, config.mk, om.cin: Release: 0.1.0 2003-05-26 17:38 sxw * om, om.cin: Made om configurable 2003-03-16 17:20 sxw * ChangeLog, config.mk: Release: 0.0.13 2003-03-16 17:20 sxw * om: Allow component names with numbers in them 2002-07-22 21:13 sxw * ChangeLog, config.mk: Release: 0.0.12 2002-07-22 20:41 sxw * om: Fixed bug which was preventing om operations with multiple arguments Patch from Paul Anderson 2002-07-17 11:13 sxw * ChangeLog, config.mk: Release: 0.0.11 2002-07-17 11:13 sxw * specfile: Remove requirement on dice-authorize 2002-06-11 08:32 sxw * ChangeLog, config.mk: Release: 0.0.10 2002-06-11 08:30 sxw * om: Fixed bug with loading LCFG resources. We previously tested the definedness of values - but with the new profile server, all values are defined, just not necessarily present. So, we need to do truth checks instead. 2002-05-06 11:07 sxw * ChangeLog, config.mk: Release: 0.0.9 2002-05-06 11:02 sxw * om: Fix silly typo 2002-05-06 10:55 sxw * om: Removed DICE::Authorize call. The name of the authorization system is now extracted from a resource 2002-05-06 10:25 sxw * om: Changed so that root can invoke components which haven't yet been given a correct set of resources Added assorted newlines to "die" commands 2002-04-30 10:32 sxw * ChangeLog, Makefile, config.mk: Release: 0.0.8 2002-04-30 10:16 sxw * ChangeLog, Makefile, config.mk, om, om.def, specfile: Release: 0.0.6 2002-04-26 21:14 sxw * ChangeLog, config.mk: Release: 0.0.5 2002-04-26 21:13 sxw * Makefile: Silly typo 2002-04-26 21:01 sxw * ChangeLog, config.mk: Release: 0.0.4 2002-04-26 21:01 sxw * Makefile: Fix mandir 2002-04-26 20:50 sxw * ChangeLog, config.mk: Release: 0.0.3 2002-04-26 20:48 sxw * Makefile: Fix typo 2002-04-26 20:47 sxw * Makefile, config.mk, om.def, specfile: Added om def file, as requested by Paul 2002-04-01 15:01 sxw * om: Fix another silly typo 2002-04-01 13:23 sxw * specfile: Fixed specfile bugs 2002-04-01 13:23 sxw * om: Fixed incorrect variable name 2002-04-01 11:45 sxw * om: Update to my current working version, which should fix problems with resource values as reported by Paul Also, fix callinguid check so that the value before setuid is stored Add simple usage() message 2002-04-01 10:04 sxw * config.mk: Fix man page install location 2002-03-26 17:29 sxw * ChangeLog, config.mk: Release: 0.0.2 2002-03-26 17:28 sxw * ChangeLog, config.mk: Release: 0.0.2 2002-03-26 17:24 sxw * ChangeLog, config.mk: Release: 0.0.1 2002-03-26 17:15 sxw * ChangeLog, Makefile, README, README.BUILD, config.mk, om, om.1, specfile: Added Build stuff 2002-02-18 18:42 sxw * om: Initial commit of om prototype 2002-02-18 18:42 * .: Standard project directories initialized by cvs2svn.