[patch] remove wireless support

Owen Fraser-Green owen at discobabe.net
Wed Jun 2 08:16:29 PDT 2004


On Wed, 2004-06-02 at 16:17, Robert Love wrote:
> On Wed, 2004-06-02 at 10:32 +0200, Owen Fraser-Green wrote:
> > No, the Linux system needs iwlib but what about all the other operating
> > systems that HAL could (and hopefully one day, will) run on?
> This argument does not fly: the same thinking can be applied to iwlib.

But iwlib (or, more correctly wireless-tools) is the tool of choice for
wireless configuration on Linux. BSD already has it's own tools (I think
they just use ifconfig). OSX, Windows and all the others presumably have
their own tools too so the reality is that iwlib will always be specific
to Linux. This means that, in today's environment, every wireless
network-aware program which aims to be cross platform must basically
implement it's own abstraction layer over the different libraries. That,
in itself, is surely bloat.

> > Surely pretty much everything HAL does could theoretically be done much
> > better using dedicated libraries for each of those tasks.
> Not device discovery, exported a hierarchy of your devices, and so on.
> It does those well.

Yet, even for these things, we could write many bloated applications
which meet all our needs.

> > Another important feature of HAL though is it's potential to provide nearly
> > everything under one roof.
> I don't see "everything under one root" as a feature at all!  But, then,
> maybe that is why I don't use Emacs.  ;-)

But surely the point of any abstraction layer is to collect many things
under one roof but, granted, "everything under one roof" was taking it a
bit far though. The art, really, is in finding how much to fit under the
roof, and where to put the chimneys (the fewer the better).

> > I agree the two code path thing is odd but I think a better goal is to
> > strengthen HAL to provide the missing second one instead of crippling it
> > by not providing any.
> That takes us back to the previous argument, which I feel leads us down
> a slipper slope toward feature bloat, unclear scope, and potential
> sudden brutal death.

I think a sudden death lies on both sides of the path, both just as
brutal as each other - one caused by gluttony and the other by
starvation. I don't think, however, we've explored every possible
solution to the previous argument.


hal mailing list
hal at freedesktop.org

More information about the Hal mailing list