Hello,
http://arstechnica.com/tech-policy/2015/10/volkswagen-pulls-2016-diesel-line... states:
The VW scandal came to light when researchers from the International Council on Clean Transportation and West Virginia University performed all kinds of tests on VW vehicles, discovering that when the vehicles were on the road, they polluted substantially more than when they were being tested for pollution emissions.
If this software was open-source (not even free!) the process to find the problem woudl be much easier and less costly. And most probably the issue would not occur at the very first place - VW engineers woudl not have temptation to do what they did.
Do you think VW case can be used as an example of the benefits which free software might provide to the society?
On 09/10/15 07:55, Vitaly Repin wrote:
Hello,
http://arstechnica.com/tech-policy/2015/10/volkswagen-pulls-2016-diesel-line... states:
The VW scandal came to light when researchers from the International Council on Clean Transportation and West Virginia University performed all kinds of tests on VW vehicles, discovering that when the vehicles were on the road, they polluted substantially more than when they were being tested for pollution emissions.
If this software was open-source (not even free!) the process to find the problem woudl be much easier and less costly. And most probably the issue would not occur at the very first place - VW engineers woudl not have temptation to do what they did.
Do you think VW case can be used as an example of the benefits which free software might provide to the society?
Did you see my blog[1] about this?
The key point here is that it is a conspiracy: (a) it involves /deliberate/ dishonesty, (b) lots of people must have known, (c) closed-source, non-free obfuscation was used to hide it and (d) it was used to deceive government(s) and now they all look stupid.
In the law, people convicted for conspiracies are usually given bigger punishments than people convicted of doing something bad on their own.
I just feel this is an opportunity too good to miss for free software advocates.
There is a revolving door between industry and the regulatory agencies and they won't ever fix themselves without outside pressure.
With consumer organizations making lawsuits, it could also be a good opportunity for FSFE to forge alliances with other groups and share the free software message more widely.
1. http://danielpocock.com/the-only-way-to-ensure-the-vw-scandal-never-happens-...
Hi Vitaly, Daniel and other subscribers,
I'd like to bring some nuance to this issue. I was told in college that such practices were already going on, that was years ago, so for me it didn't come as a surprise. I was however surprised by the strong reaction, which in hindsight makes sense.
Car emissions requirements are very strict, and it is a real challenge pushing technology to meet these ever stricter demands. For the EU these emissions are verified based on a predefined tests the car has to go through. If there is plenty of headroom in the technological solution you can just pass any test flawlessly. If this headroom is lacking and you want to pass the test the first step is to optimize the system for the test-curves. The next step, we've now observed, is to detect the chain of events in the test and handle the situation differently.
Would not any modern car optimize its performance for different sets of driving habits and environmental circumstances? Were the tests or regulations too naive not to set (I assume) requirements for optimizations or related software auditing?
Making the point that an open source development model would have prevented this issue seems a bit overblown to me. Yes it would be easier to spot, but would it? As I've read this specific issue was discussed on a technical forum between regulators and engineers months earlier which did not directly trigger any greater discussion. Furthermore as I've described earlier the industry has been aware of such practices for many years, so I guess this is just the first real case made against such practices.
Spotting such specific code in a software project as complex as a modern car would not be simple. But of course having the code at hand combined with real outside-in tests would be surely help.
Don't get me wrong, this is an environmental crime which is serious and a through investigations and regulatory adjustments are needed. But to me this issue has more to do with regulations, requirements, company ethics, and less to do with the licensing or development model of the source code.
I'm more concerned about the security implications of hidden source code (VW cars apparently had an unresolved security flaw in their lock system [1]) and the limited possibility of executing maintenance or modifications due to software restrictions (John Deere). I'm not sure how this should be regulated, but having decent laws in place to provide freedom for car drivers and car owners would be a great step forward.
[1] http://it.slashdot.org/story/15/08/13/1853228/banned-article-about-faulty-im...
Glad to hear others' thoughts on this.
Kind regards, Nico
On vr, 2015-10-09 at 08:02 +0200, Daniel Pocock wrote:
On 09/10/15 07:55, Vitaly Repin wrote:
Hello,
http://arstechnica.com/tech-policy/2015/10/volkswagen-pulls-2016-diesel-line... states:
The VW scandal came to light when researchers from the International Council on Clean Transportation and West Virginia University performed all kinds of tests on VW vehicles, discovering that when the vehicles were on the road, they polluted substantially more than when they were being tested for pollution emissions.
If this software was open-source (not even free!) the process to find the problem woudl be much easier and less costly. And most probably the issue would not occur at the very first place - VW engineers woudl not have temptation to do what they did.
Do you think VW case can be used as an example of the benefits which free software might provide to the society?
Did you see my blog[1] about this?
The key point here is that it is a conspiracy: (a) it involves /deliberate/ dishonesty, (b) lots of people must have known, (c) closed-source, non-free obfuscation was used to hide it and (d) it was used to deceive government(s) and now they all look stupid.
In the law, people convicted for conspiracies are usually given bigger punishments than people convicted of doing something bad on their own.
I just feel this is an opportunity too good to miss for free software advocates.
There is a revolving door between industry and the regulatory agencies and they won't ever fix themselves without outside pressure.
With consumer organizations making lawsuits, it could also be a good opportunity for FSFE to forge alliances with other groups and share the free software message more widely.
http://danielpocock.com/the-only-way-to-ensure-the-vw-scandal-never-happens-... _______________________________________________ Discussion mailing list Discussion@fsfeurope.org https://mail.fsfeurope.org/mailman/listinfo/discussion
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256
On 09/10/15 08:57, Nico Rikken wrote:
Making the point that an open source development model would have prevented this issue seems a bit overblown to me. Yes it would be easier
http://www.computerworld.com/article/2986682/telematics/how-the-dmca-may-hav...
Don't get me wrong, this is an environmental crime which is serious and a through investigations and regulatory adjustments are needed. But to me this issue has more to do with regulations, requirements, company ethics, and less to do with the licensing or development model of the source code.
It is all about hiding things from the public, keeping the public in the dark and exploiting the gullible
It isn't just about emissions, they used the same approach with the fuel efficiency statistics. This is why many consumer organizations are now making law suits. Consumers were duped into believing their cars use less fuel than they really do.
I'm more concerned about the security implications of hidden source code (VW cars apparently had an unresolved security flaw in their lock system [1]) and the limited possibility of executing maintenance or modifications due to software restrictions (John Deere). I'm not sure how this should be regulated, but having decent laws in place to provide freedom for car drivers and car owners would be a great step forward.
The regulatory situation can be improved in various ways:
- - Governments (and the military, police, etc) own and lease many cars, they could make open source a procurement policy
- - the DMCA (and anything like it that is hidden in trade partnerships, such as the recently signed TPP or the proposed TTIP) should be scrapped/opposed
- - the law could even go the opposite direction, giving consumers a right to have software inspected and modified. Many countries already have laws giving consumers choice in other areas (e.g. the choice to have their car serviced at any garage and not just with the franchise where they bought it) so why shouldn't they have the same right of choice with software?
- the law could even go the opposite direction, giving consumers a
right to have software inspected and modified. Many countries already have laws giving consumers choice in other areas (e.g. the choice to have their car serviced at any garage and not just with the franchise where they bought it) so why shouldn't they have the same right of choice with software?
VW's fraudulence puts the public in a potent position to demand serious changes such as publishing complete corresponding source code. Strongly copylefted free software (such as AGPL v3 or later) will help users against future fraud and possibly make the affected vehicles trustworthy and saleable again. It would be good if VW drivers, for instance, could get complete corresponding source code for their entire car licensed under the AGPL v3 or later so they could take that code to someone they trust and know that whatever work was done to that code would also respect their software freedom. Non-copylefted free software under some pushover license could be extended with nonfree code and thus stop being valuable in the foreseeable short-term future, recapitulating the very problem that led to this fraud in the first place. A proper fix will mandate that all distributed code must be free and ensure that freedom travels with the code.
Hi Fellows,
I'd like to add another point. If you're not very rich you pay taxes. If your country does not meet emission criteria there will be very big fines. Which the tax payer will pay.
I am not willing to end up as a debt-slave, through government, for corrupt companies who cheat software.
Best regards,