similar to: [PATCH server] don't add nil key to failures hash for PartialSuccessError handling.

I've pulled out the no-longer-necessary remnants of the old way of handling auth and before_filters as well as fixing a couple bugs that had crept in along the way. Unit test fixes for the refactoring will follow in a subsequent patch. Signed-off-by: Scott Seago <sseago at redhat.com> --- src/app/controllers/application.rb | 46 ----------------

There's still room for additional refactoring as we may want to eventually split out nic, bonding, and ip address manipulation into their own controller/svc module, but that would be a much more extensive redesign than we want to deal with for the overall service layer creation refactoring. Signed-off-by: Scott Seago <sseago at redhat.com> --- src/app/controllers/network_controller.rb

A bunch of controllers were not extending ActionController::TestCase, and instead were just using the basic test/unit testcase, so we were not getting some of the rails baked-in stuff. Signed-off-by: Jason Guiditta <jguiditt at redhat.com> --- src/test/functional/host_controller_test.rb | 2 +- src/test/functional/nic_controller_test.rb | 2 +-

Also includes a couple minor model changes to support better error reporting. This method was added so that a client could submit multiple vms to the service layer in one call and get back appropriate messages showing any failures without those failures also causing the entire call to fail, or forcing the client to implement handling (begin/rescue blocks) of these errors to continue through the

This isn't yet functional, but I'm submitting the patch so Ian can continue work on it. Signed-off-by: Scott Seago <sseago at redhat.com> --- .../lib/ovirt/controllers/network_controller.rb | 26 ++++++++++++++++++++ .../physical_network_impl_controller.rb | 10 +++++++ .../lib/ovirt/controllers/vlan_impl_controller.rb | 9 +++++++

I have what I feel is a standard habtm association to manage user/roles/permissions. However, I''m having a small problem wondering why ".include?" isn''t working like I thought it should. I''m attempting to use this to determine whether a user has access to a particular permission: if session[:user].role.permissions.include?(''Lead List'')

created/edit network information is all specified on one simpler form --- src/app/controllers/network_controller.rb | 27 +++++++++++++++---- src/app/models/bonding.rb | 2 +- src/app/models/network.rb | 2 +- src/app/models/nic.rb | 2 +- src/app/views/network/_form.rhtml | 15

The recent refactoring didn't work properly for new smart pools, so I've cleaned jup the parameter handling for the various pool 'create' actions. Signed-off-by: Scott Seago <sseago at redhat.com> --- src/app/controllers/hardware_controller.rb | 5 +---- src/app/controllers/pool_controller.rb | 6 ++++-- src/app/controllers/resources_controller.rb | 6

From: Scott Seago <sseago at redhat.com> This adds some of the network API stuff to ovirt-agent. It's still not complete but what is there is functional. Signed-off-by: Ian Main <imain at redhat.com> --- src/ovirt-agent/lib/ovirt.rb | 3 + .../ovirt/controllers/hardwarepool_controller.rb | 2 +- .../lib/ovirt/controllers/network_controller.rb |

This revised version of the patch incorporates jguidditta's sorting fix as well as some other changes required to make this work fully on the cloud UI side. --- src/app/controllers/pool_controller.rb | 2 +- src/app/controllers/resources_controller.rb | 4 +++- src/app/controllers/smart_pools_controller.rb | 5 ++--- src/app/helpers/application_helper.rb | 14

This revised version of the patch incorporates jguiditta's sorting fix as well as some other changes required to make this work fully on the cloud UI side. --- src/app/controllers/pool_controller.rb | 2 +- src/app/controllers/resources_controller.rb | 4 +++- src/app/controllers/smart_pools_controller.rb | 7 +++---- src/app/helpers/application_helper.rb | 14

This revised version of the patch incorporates jguiditta's sorting fix as well as some other changes required to make this work fully on the cloud UI side. --- src/app/controllers/pool_controller.rb | 2 +- src/app/controllers/resources_controller.rb | 4 +++- src/app/controllers/smart_pools_controller.rb | 7 +++---- src/app/helpers/application_helper.rb | 14

There's still a problem with this working on the cloud side, but that can be resolved in a subsequent commit. --- src/app/controllers/pool_controller.rb | 2 +- src/app/controllers/resources_controller.rb | 7 ++++++- src/app/controllers/smart_pools_controller.rb | 5 ++--- src/app/helpers/application_helper.rb | 14 ++++++++++++++ src/app/models/vm.rb

Add of a button destroy for disabled hosts. This button behave in a similar way than the delete button of a VM. Signed-off-by: Sylvain Desbureaux <sylvain.desbureaux at orange-ftgroup.com> --- src/app/controllers/host_controller.rb | 5 +++++ src/app/services/host_service.rb | 15 +++++++++++++++ src/app/views/host/show.rhtml | 17 +++++++++++++++++ 3 files changed,

Hi All, I get this strange problem with newly scaffolded apps - I''d really appreciate any help in this regard. /usr/local/lib/ruby/gems/1.9.1/gems/activesupport-2.3.4/lib/ active_support/message_verifier.rb:46:in `block in secure_compare'' /usr/local/lib/ruby/gems/1.9.1/gems/activesupport-2.3.4/lib/ active_support/message_verifier.rb:45:in `each''

This resend includes the rebase, but fixes a problem I just noticed with nil/empty checks.

This is a resend of the 2-patch series from a week or so ago, needed to be rebased due to some commits on next

The following two patches should be applied in the order: * [PATCH server] Add svc_vm_actions method to VmService. * [PATCH server] Cloud UI layer to initiate actions on vms. The service layer patch should work fine on its own (and not break anything), but cloud patch relies on service being there.

Hello all, I''m having a strange issue with a model that belongs_to two other models. The one association is polymorphic and the other is normal. ================================ class Role < ActiveRecord::Base belongs_to :user belongs_to :permission, :polymorphic => true end class User < ActiveRecord::Base has_many :roles has_many :programs, :through => :roles end

An app I''ve been developing has been chugging along nicely, but around the time I added a new join model a couple of commits ago, my development environment has been acting screwy. 1. When there''s an error, the error messages and tracebacks are not being returned to the browser. The browser keeps the connection alive indefinitely. 2. Stopping the request from the browser