Re: [NM-openconnect PATCH] Allow 'lasthost' and 'autoconnect' settings.

[David Woodhouse <dwmw2 infradead org>]

On Tue, 2009-03-31 at 09:44 +0100, David Woodhouse wrote:
> One thing that's tempting is to make nm-openconnect-service _not_ bail
> out when it sees a configuration key it doesn't like. At least it's
> easier to add stuff in the auth-dialog then, and the patch in $subject
> would have purely cosmetic, rather being necessary to restore
> functionality:
> 
> diff --git a/src/nm-openconnect-service.c b/src/nm-openconnect-service.c
> index 68fcd15..bae04ec 100644
> --- a/src/nm-openconnect-service.c
> +++ b/src/nm-openconnect-service.c
> @@ -134,11 +134,7 @@ validate_one_property (const char *key, const char *value, gpointer user_data)
>  
>         /* Did not find the property from valid_properties or the type did not match */
>         if (!info->table[i].name) {
> -               g_set_error (info->error,
> -                            NM_VPN_PLUGIN_ERROR,
> -                            NM_VPN_PLUGIN_ERROR_BAD_ARGUMENTS,
> -                            "property '%s' invalid or not supported",
> -                            key);
> +               nm_warning("property '%s' invalid or not supported", key);
>         }
>  }

Hm... after making the auth-dialog store even _more_ stuff that
nm-openconnect-service doesn't know about (cached group/username entries
in the authentication form), it occurs to me that there's no particular
reason for me to put them in /system/networking/connections/$N/vpn/ -- I
could put them in another 'directory' altogether, and then they wouldn't
offend nm-openconnect-service at all.

Would you hunt me down if I did that?

-- 
David Woodhouse                            Open Source Technology Centre
David Woodhouse intel com                              Intel Corporation