[Fsf-Debian] few arguments to FSF

Adam Bolte abolte at systemsaviour.com
Fri Aug 10 17:42:33 UTC 2012


On Fri, Aug 10, 2012 at 01:15:58PM +0200, Paul van der Vlis wrote:
> Op 10-08-12 12:09, Adam Bolte schreef:
> > On Fri, Aug 10, 2012 at 11:23:41AM +0200, Paul van der Vlis wrote:
> > I'm not comfortable with the idea that any hardware not supported by free
> > software should be thrown out, if it can work with free software 99%+. Sure
> > it's not good from a freedom POV and should be avoided wherever possible, but
> > in situations where the alternative is to throw it out, it's potentially very
> > wasteful.
> 
> There are many people using Windows or Mac, you don't have to trow it
> away. Maybe you can even swap a device with somebody.

Ideally, we would be able to convince people to use free software by
explaining why it is so important. We would actually be doing the exact
opposite of this by trading hardware known not to be compatible with non-free
operating systems.

I would rather the hardware be used, and used to run software as free as it
possibly can - even if that means (hopefully only temporarily) running a blob.

 
> > It's also very difficult to be 100% sure of *all* the components in any device
> > - particularly when it's not possible to test hardware with a LiveCD or some
> > such in advance.
> 
> Information would be good. It's nonsense that everyone needs to test.

Information is only useful to people with a technical skill-set. Most people
are not technical. I work with web developers, but I very much doubt even
those guys would know what chipsets their hardware uses.

I think it's reasonable to expect that we can steer the masses who are
interested in free software towards certain brands, however I can't see many
people researching purchases much beyond that level of detail. Not everyone
knows a geek to help them out.

 
> > Previously I would just build my own desktop machine from
> > whatever parts I think would play nice, however laptops and tablets don't give
> > you such flexibility. In many cases, there is no way to know which chipset a
> > particular laptop will have. 
> 
> Boot with a liveCD or USB stick, and do "lspci".

That contradicts your previous statement. I agree that this can be useful and
I would if given the opportunity, however it is rarely possible. This is
especially true on tablets that are not easily bootable from an external
device due to requiring custom kernels, bootloaders, etc - and if you already
have that level of detail about the device, you wouldn't need to be looking in
the first place.  :)

 
> > The model names can be identical but use
> > different hardware in different countries or for different revisions... no
> > hardware database is ever going to fully resolve such issues. 
> 
> Not all hardware manufacturers are like that.

That might well be true, but it's pretty common. I recall trying to purchase
laptops compatible with coreboot from ebay a while back, and seeing many
different kinds of X60 and X60s models and giving up because I considered the
chance of success to be too low for the price.

Say somebody recommends a Sony Vaio 13" laptop because it works great for
him/her with free software... there are potentially hundreds of laptops with
that description, even though that's how stores will refer to such laptops as
(where the exact model VPCYB16KG or some such may not even be mentioned).

 
> The opposite would be good possible: somebody can make hardware what
> works 100% OK with free drivers/firmware. I think that would be really
> good.  And we all need to help selling them.

I agree, and I'd love to see that. If such a laptop were available where I
live, I would purchase it next time I'm in the market for a new machine. As it
stands, I *really* tried to get a Yeeloong Lemote, but did not succeed. The
only distributor that supposedly would deliver to my country made me wait a
month to send something that was in stock, after which point I had to get my
money back from PayPal for non-delivery.

However, my understanding is that even Yeeloong - the only manufacturer of
laptops that run 100% free software - also sell laptops that don't work with
100% free software due to an ATI graphics chipset or some such. Again, this
would confuse purchasers.

 
> Yes, there are many sites I cannot use or understand without flash. The
> alternatives are not good enough at the moment, except for Youtube and
> some older sites.

In my case, I simply go without. If I can't browse web pages using free
software, I simply visit other websites that I can browse fine. The user may
not have much choice about what hardware they have, but they surely have a
choice about what sites they visit.

 
> And without flash I can only use Youtube for video. I can't e.g. watch
> the evening news in my country, you need flash or silverlight for it.

I can't watch flash video news either - but I can read it just fine.

 
> > but some non-free driver that's
> > required to use a machine isn't? I can't agree with you there.
> 
> There are many good alternatives for such a machine.

Alternatives which I may or may not have access to. However, as long as the
web remains free, I will always have access to alternative websites.


> > If it's just a matter of changing:
> > 
> > deb http://ftp.debian.org/debian wheezy contrib non-free
> > 
> > to say:
> > 
> > deb http://ftp.nonfree.org/debian wheezy contrib non-free
> > 
> > I don't think it is all that much of an improvement 
> > - not as long as the Debian project plays any part in it.
> 
> It's only a little step, but it's important. E.g. to become FSF free in
> the future.

I don't agree that it's important at all - unless the Debian project distances
itself *completely* from nonfree.org.

 
> > And I do believe that the Debian project
> > will need to be doing the right thing by the FSF for endorsement - and not
> > just the distribution. There will perhaps need to be that level of trust that
> > the distribution will not change its views down the road.
> > 
> > To take this to the extreme, if Microsoft released a 100% free software
> > distribution tomorrow (while still pushing proprietary software for everything
> > else), I think the FSF would be reluctant to recommend that distribution also.
> 
> You say it: "while still pushing proprietary software..."/

Yes. Microsoft distribute Windows, which is proprietary software, just as the
Debian project distributes lots of proprietary software in non-free right now.

 
> > There are alternatives. Most probably, these would be a lot better today if
> > more people cared about them - and more people would care about them if more
> > people actually used them.
> 
> You will not become glad to use them, when they do not work.

Then I simply wouldn't use that software if it wasn't that important to me, or
I would contribute to making it work to fit my needs.

 
> > But I disagree it's always possible (or at least practical) to buy hardware
> > (assuming you are the one doing the buying) that you know for a fact will with
> > 100% with free software - not until manufacturers and/or local stores readily
> > publicly commit to the cause. In many situations, the best you can do is make
> > an educated guess.
> 
> I know all about it, I sell computers with Debian pre-installed. And I
> buy the hardware for it. And what I say: we need a manufacturer who
> would make a "free" line.

So you agree that it's quite a burden? More the reason we can't expect
end-users to always get hardware purchases right, at least not until such
manufacturers come along that will sell to any country.

Regards,
Adam
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 490 bytes
Desc: Digital signature
URL: <http://lists.alioth.debian.org/pipermail/fsf-collab-discuss/attachments/20120811/5ebc7671/attachment-0001.pgp>


More information about the Fsf-collab-discuss mailing list