[chef] Re: problem of route provider implementation

[Seth Falcon < >]

Hi there,

On Wed, Aug 18, 2010 at 8:41 AM,  
<
 >
 wrote:
> The route provider always runs route command regardless of the current state
> of route´s table.
>
> I have looked closely at the source code and it seems that the
> implementation contains several  errors
>
>
> a) in line 100 :  "is_running" is actually a local variable not the
> attribute "is_running" of the object so it will get lost not being
> used in method action_add
> this is fundamental error of programming, inside the instance method we must
> use @is_running to set the value of the attribute of the object
> b) the method load_current_resource has wrong logic, checking
> destination/target of the current and new route , because while
> destination/target can be
> the same, the gateway can be changed  and the route should be considered as
> changed and must be updated
> c) the method load_current_resource when pulling data from /proc/net/route
> shall take into consideration the header of the file and ignore it

This is all fantastic feedback.  Any chance you'd be willing to make
the changes you are suggesting inside a git checkout of the source
code and send along a patch?

+ seth