[pkg-wpa-devel] Proposal: Split the package

[Felix Homann]

On Tuesday 11 April 2006 12:37, Kel Modderman wrote:
> I have had great feedback from many people for the alternative approach
> that I brought to the table; ifupdown integration.

No doubt at all. It is great. But there still remain things that can not be 
done with this approach. Therefore working setups should not be broken.

> The ifupdown scripts are self contained and lay dormant unless you
> activate them with the special commands in /e/n/i, therefore they can be
> installed by default regardless.

True.

> The init script, as I understand it, really has its limitations:

For sure it has, but the new approach makes no exception. Why not support 
both? As I see it it's the only way not to break working setups on upgrade. 
What's the problem with including it and not enabling it by default?

>     * it can be used with only one interface as defined in its conffile

Those who need more will surely profit from the new approach.

>     * it conflicts with new low-level services, such as network-manager,
>       because it is always active

Those who want to use network-manager can easily turn it off.

>     * it was never intended to be used as the backbone of a "roaming
>       daemon solution" (in my opinion) but was just used that way out of
>       convenience

Here are some notes from man wpa_supplicant:

"In addition, it [wpa_supplicant] controls the  roaming  and  IEEE 802.11 
authentication/association of the wireless LAN driver." 

"wpa_supplicant  is  designed  to be a "daemon" program that runs in the 
background and acts as the backend component controlling  the  wireless 
connection."

> I have no objection to the daemon personally, but its deprecation was
> discussed and decided upon before I arrived on the scene. 

OK, but now that the upgrade has broken working setups of people not involved 
in the discussion, it might be worthwhile to rethink the decision.

Regards,

Felix