Re: Bug#353277: ndiswrapper in main
On 21 Feb 2006, Steve Langasek verbalised:
> On Tue, Feb 21, 2006 at 10:40:06AM +1000, Anthony Towns wrote:
>> On Mon, Feb 20, 2006 at 05:36:13PM +0100, Robert Millan wrote:
>>> I requested that ndiswrapper and ndiswrapper-modules-i386 be moved
>>> to contrib.
>> While I would personally rather see the "contrib" demarkation cover
>> this, emulators, and clients for propietary protocols, I'm
>> disinclined to override both the maintainer and the ftpmaster that
>> accepted it, particularly on this single issue rather than as a
>> global policy change for those issues. I expect I'll either abstain
>> or vote against.
> I suspect I disagree with Anthony on where exactly the line should
> be drawn, but it does seem to me that the arguments used to justify
> ndiswrapper's presence in main are rather contrived. Nobody is
> going to want to run drivers under ndiswrapper in a production
> environment if there is a suitable free equivalent available for
> Linux; the only practical applications I see here are using non-free
> Windows drivers under Linux for otherwise-unsupported hardware, and
> using ndiswrapper as a tool for preliminary testing of drivers being
> written for Windows in an environment that doesn't require booting
> Windows. The former is what I use it for, and what every user I
> know uses it for, and doesn't justify a claim that ndiswrapper does
> not depend on non-free software. The latter, IMHO, would be grounds
> for shipping the software in main, but AFAIK this is purely a
> hypothetical at this point.
I think Raul pointed out that there are free windows only
drivers seen in the wilds out there, in which case ndiswrapper seems
to represent an implementation of a protocol that allows windows only
code to work with the Linux kernel. Whether or not such free code has
been ported to Linux natively should be immaterial, I would draw the
line at whether ndiswrapper falls in the category that "installers"
do, or in the category that "wine" does.
I think it tends to fall in the latter, since it does not seem
to be specific to any particular piece of code or driver; and there
is nothing that precludes a windows only piece of code that works
with ndiswrapper to be licensed freely.
> Either way, I do agree with Anthony that one-off overrides of
> maintainers don't seem like the best way for us to be spending our
While I agree with you both, it is because I thinik that
ndiswrapper actually belongs in main.
If you don't care where you are, then you ain't lost.
Manoj Srivastava <email@example.com> <http://www.debian.org/%7Esrivasta/>
1024D/BF24424C print 4966 F272 D093 B493 410B 924B 21BA DABB BF24 424C