LWN.net Logo

safety-critical systems can use ROM

safety-critical systems can use ROM

Posted Oct 18, 2006 11:51 UTC (Wed) by smitty_one_each (subscriber, #28989)
In reply to: safety-critical systems can use ROM by stevenj
Parent article: FSF should separate GPLv3 changes (Linux.com)

>As Moglen and Stallman have repeatedly pointed out, in conditions where non-modifiability is critical for safety, or because of legal constraints, you can always use a ROM.

For what values of 'always'? Among the long-term challenges here is the one where an idealogue fundamentally tweaks the 'spirit' of the GPL in questionable ways, and makes a big show of soliciting input, all the while shunting all input to /dev/null.

While sympathizing with where the FSF is trying to go, I'd argue that the market, as opposed to the license, is a preferred feedback loop.


(Log in to post comments)

free market is not the answer

Posted Oct 18, 2006 12:42 UTC (Wed) by man_ls (subscriber, #15091) [Link]

The free market has already spoken: use Windows, forget about Linux. Join AOL. iPod rules. Do you really want to leave your decisions to the market?

free market is not the answer

Posted Oct 18, 2006 17:25 UTC (Wed) by smitty_one_each (subscriber, #28989) [Link]

Your reply doesn't seem to take into account the increasing success of FOSS, and companies built upon it.
AOL is TU.
iPod rules what?
I'm not arguing blind faith in the market; it is certainly subject to manipulation.
However, over time, the court of public opinion, as expressed in the market, tends towards reasonableness.

free market is not the answer

Posted Oct 18, 2006 22:14 UTC (Wed) by man_ls (subscriber, #15091) [Link]

Yeah, the court of public opinion has chosen Windows, iPod (73% market share) and AOL (19% market share, #1 ISP with difference). Have you bought your copy of Oracle yet?
Your reply doesn't seem to take into account the increasing success of FOSS, and companies built upon it.
Yeah right. The most visible "FOSS" (what a horrible acronym), i.e. libre program is probably Firefox, and the most extended; it has been competing with Internet Explorer for ages, it is free (as in beer) as the competition, but also libre, there is a buoyant company behind it and it is the direct descendant of Netscape which in 1994 was the only reasonable choice. And IE has had the most horrible track record in performance, bugs and security that you can imagine; as to marketing Microsoft is now a convicted monopolist because of it. Yet last month our "FOSS" champion had 11 or 12% market share, depending on who you believe. Great success in the "court of public opinion". Reasonableness indeed.

The free market is an economic mechanism, which must be nurtured and kept within bounds constantly through regulation. Its effect is to lower prices and improve efficiency. It doesn't choose technologies or morals; it just goes with the cheapest option. When two $0 options meet, convenience seems to win every time.

Let's stop attributing human qualities to an economic phenomenon, please. We might try to make something other than money drive us for a change.

free market is not the answer

Posted Oct 19, 2006 13:51 UTC (Thu) by smitty_one_each (subscriber, #28989) [Link]

>Let's stop attributing human qualities to an economic phenomenon, please. We might try to make something other than money drive us for a change.

Fair enough, but I have some difficulty with trying to view economic phenomena as pure abstraction: economics is a derivative of human behavior, subject to all of the usual chaotic influences. How does one _not_ anthropomorphize a fundamentally anthropomorphic thing?

I do agree, on a personal level, that "something other than money" would make a great driver. Externalizing my own non-monetary motives would probably bring in howls of derisive laughter, however. Money, OTOH, is the low-common-denominator metric that everyone at leasts understands, possibly without liking. Abstract motives attenuate rapidly; cold, hard cash drives far more people.

If you step back and look at the last couple of decades from a distance, FOSS has gone from nowhere to at least somewhere, and its adoption has a positive slope. Proprietary stuff, OTOH, is growing increasingly painful on all levels. I'll venture a guess that Vista proves to be the last OS excreted by Redmond.

Give it a couple of decades. Future so bright, we gotta wear cheap sunglasses.

safety-critical systems can use ROM

Posted Oct 18, 2006 14:10 UTC (Wed) by cventers (subscriber, #31465) [Link]

> For what values of 'always'? Among the long-term challenges here is the
> one where an idealogue fundamentally tweaks the 'spirit' of the GPL in
> questionable ways

I feel saddened that anyone sees things this way, especially given that
the 'spirit' they identify in the GPL is an incidental property rather
than an engineered end. But this issue has probably been beaten into the
ground now, though it's only fair to point out that the paranoia over
'changing spirit' is a long ways from being universally shared.

> ...and makes a big show of soliciting input, all the while shunting all
> input to /dev/null.

Pardon me, but do you have any evidence that this is what is happening?
Because there are a whole lot of people who have voluntarily stepped
forth and involved themselves in the process, both from the private user
community and corporate community, that have publicly said otherwise.

In fact, the only parties I've ever heard complaining about this
so-called "open sham" is kernel developers - the one party who has
repeatedly refused invitations to participate.

(Ironically, the one kernel developer that /did/ participate was seen
here asking his colleagues why they didn't join him, and I don't recall
him attributing the feedback process to a 'sham'...)

Personally, based on the statements of some kernel developers, I'm sad to
say that I'm having a hard time distinguishing much of what they say
publicly about GPL and RMS and FSF from FUD.

safety-critical systems can use ROM

Posted Oct 18, 2006 18:24 UTC (Wed) by nim-nim (subscriber, #34454) [Link]

> While sympathizing with where the FSF is trying to go, I'd argue that the
> market, as opposed to the license, is a preferred feedback loop.

ROTFL:

1. your licensing terms are part of your market offer
2. DRM proponents trust the "market" so much they've massively invested in changing legislation worldwide

safety-critical systems can use ROM

Posted Oct 19, 2006 9:54 UTC (Thu) by drag (subscriber, #31333) [Link]

"While sympathizing with where the FSF is trying to go, I'd argue that the market, as opposed to the license, is a preferred feedback loop."

Well I don't know if you noticed, but FSF has invited many different people and many different corporations to join in on the discussion and I haven't seen a person or company actually involved in the proccess bitch about how the FSF is operating or complaining about the GPLv3 (although people say that there are still ways to improve it.) Many of them are major propriatory software makers and holders of large amounts of patents.

The only bitching I see is from people that refused to join in on it in the first place. I mean it would be different if they joined in and became disillutioned about it and dropping out and THEN complaining. Then it would have a meaningfull impact.

I mean seriously folks, come on. It's not like they are making it _difficult_.
http://gplv3.fsf.org/

After all the GPLv3 isn't even released yet? So how can you gauge the market based on a vocal group hating a beta version of the product (ie the license draft.)

Bitching over GPLv3?

Posted Oct 19, 2006 20:46 UTC (Thu) by vonbrand (subscriber, #4458) [Link]

The only bitching I see is from people that refused to join in on it in the first place. I mean it would be different if they joined in and became disillutioned about it and dropping out and THEN complaining. Then it would have a meaningfull impact.

Said people have repeatedly stated that they objected to the whole "anti-DRM" idea well before the whole GPLv3 process started. They have had rows with RMS (and the FSF) for ages. Why should they now forget all that and just run with the crowd, expecting to be listened to this time around?

Bitching over GPLv3?

Posted Oct 19, 2006 21:18 UTC (Thu) by khim (subscriber, #9252) [Link]

Why should they now forget all that and just run with the crowd, expecting to be listened to this time around?

They sure don't. And I respected Linux and kernel developers while they said "If you want politics, go over to the projects which do politics. We do code" and acted like it. But when they started doing politics I've lost most of the respect and then the question why arose.

I mean: you "have had rows with RMS (and the FSF) for ages". Fine. You ignore the whole GPLv3 process since you don't believe your participation can change anything. No problem. But then you suddenly start "doing politics", you raise big stink, etc. Why ? If you can answer the question - I'm all ears.

And I remember previous "row" very well indeed. When Linus started using BitKeeper RMS said right away that it's bad decision and it's so very sad that kernel developers are using non-free software for essential things. And RMS was right back then BTW. But have you seen RMS's loud and noisy campaign titled "Let's force kernel developers to drop BitKeeper" or anything like this ? I sure don't.

It's one thing when you disagree with someone - and it's totally different thing when you slander and bad-mouth someone, sorry. Kernel developers crossed this line 2006.09.15...

Bitching over GPLv3?

Posted Oct 19, 2006 21:57 UTC (Thu) by bronson (subscriber, #4806) [Link]

Erm, where are the kernel devs bad-mouthing the FSF? Behind that link I see mostly glowing praise for the FSF, the GPL, and many of the GNU software products they have enabled. I think the kernel devs have expressed their opinion very professionally and clearly and they seem to have a good point.

Khim, if you consider this slander and bad-mouthing I do hope that you slander and bad-mouth me some time soon! :)

Bitching over GPLv3?

Posted Oct 20, 2006 7:36 UTC (Fri) by khim (subscriber, #9252) [Link]

Erm, where are the kernel devs bad-mouthing the FSF?

On Groklaw (mostly Linus), comments on other sites (including LWN), interviews, etc. But there too. You don't need dirty words to slander and bad-mouth - it's possible to do it with great politeness.

I think the kernel devs have expressed their opinion very professionally and clearly and they seem to have a good point.

May be, but after that piece (and you are right - it's quite "politically correct") it become clear that they are not content with coding. They want to play politics - but they don't like open confrontation where FSF or RMS can respond. Do you know anyone who's using (or used) such tactics ? Right: it's our dear McBride. Do you like his approach ? Do you feel it's correct approach ? Why it's good for kernel developers then ? Because they are good guys ?

Sorry but while all talks about GPLv2 is were internal discussion between kerel developers I was content and respected their right. It was open to the world - but that's just how linux development works. But when you start using "SCO tactic" (talk to the press first, to the offending party second) - all previous talks are changing colour: was they really only internals talk which become public since all talks about kernel development are public ? Or were they intended for the general public from the start ? And why they only talk to the general public - never to the FSF or RMS ?

When they accuse FSF in "a fundamental violation of the trust" I'd like to see more the just "while we may argue forcefully for our political opinions, we may not suborn or coerce others to go along with them". GPL was always political: if "The licenses for most software are designed to take away your freedom to share and change it [but GPL is different]" (the very first sentence of the GPLv2) is not political then what is political ? And promise a user who needs changes in the system will always be free to make them himself, or hire any available programmer or company to make them for him was written years before GPLv2 (let alone GPLv3) so it's kind of hard to argue that "DRM-clause" constitute "violation of trust" - I already find it quite hard to fullfill this promise without such clause ("source code for all modules" plus "the scripts used to control compilation and installation of the executable" are not enough in today's world - or so it seems).

P.S. The best summary of my thoughts:
Would I trust RMS to write a successful OS kernel? God No (see HURD)
Would I trust kernel developers to choose (let alone write) successful license? God No (see BitKeeper).

And that's about it: track record for RMS's OS kernel development is poor but track record for writing good licensed is good (not excellent due to GFDL, but if you'll compare GFDL with CCL...).

Bitching over GPLv3?

Posted Oct 20, 2006 11:59 UTC (Fri) by smitty_one_each (subscriber, #28989) [Link]

>When they accuse FSF in "a fundamental violation of the trust" I'd like to see more the just "while we may argue forcefully for our political opinions, we may not suborn or coerce others to go along with them". GPL was always political: if "The licenses for most software are designed to take away your freedom to share and change it [but GPL is different]" (the very first sentence of the GPLv2) is not political then what is political?

Are not licenses are about regulating issues of legality? If the GPL is political, are you contending that the FSF is, in fact, a political party? I have always, in my 5 or 6 years of membership, considered it a Foundation, and a *community*. I really don't buy off on RMS's philosophical/political views. Am I a bad person?

>Would I trust RMS to write a successful OS kernel?
Trust? While I may disagree with the gentleman, I wouldn't consider him dishonest or dishonorable in any way. I'll venture that, given his Lisp-sih predeliction and disinterest in mult-threaded applications (as seen in Emacs) that kernel coding simply is not to his taste. Does it strike anyone as odd that the HURD has never achieved much popularity? Maybe the GPL is a necessary, but insufficient, part of the FOSS diet, and a few more pragmatic elements had to be mixed in to achieve robust nutrition. Just sayin'.

>Would I trust kernel developers to choose (let alone write) successful license?
How about some pragmatism? BK worked, and then it didn't, and then there was git. No blood, no foul. Much publicity was enjoyed by all. Woo hoo.

Copyright © 2013, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds