Is standardization deemed to be against FS and how can it be tackled?

Tue Jun 2 08:37:44 UTC 2015

> On 1 Jun 2015, at 15:52, Tom Blecher <blecher.tom201645 at yandex.com> wrote:
> 
> Hi Carsten,
> what do you mean.
> Standard bodies might be allowed to charge a fee for verifying compliance.
> "charge a fee" to whom? For example "compliance" with what?

Typically where there is regulatory compliance needed for deploying software, the software needs to be checked for compliance to the standard by a registered testing agent. This usually incurs a fee, after which the software is certified compliant and thus suitable for deployment. (This does provide some challenges to FOSS where the culture is one of continuous releases.)

For example, software used for the management of clinical trial data in the UK is required to be tested against a set of standards before being allowed for use with hospital patient data. Some FOSS solutions such as OpenClinica pay for this certification.

In other industries, “badging” for compliance is an optional requirement and mostly FOSS doesn’t bother. There are also some SSOs that offer free self-certification using a self-test harness. When I was implementing a W3C spec for an ASF project I just published our raw test harness results on the web so users could judge the state of compliance for themselves (the tests were all implemented in JUnit); closed source vendors on the other hand could - and did - just issue a statement saying they passed everything! Which was funny as several of the tests had bugs and were impossible to pass unless your implementation was broken :D

> Ok:
> Let us take the rfcs. Who could charge a fee to who_? And for what. When? Each time reading them? A annual fee? How much? Let us say a thousand € per year and document, as it is indeed the case with ISO or DIN?

The business models for formal SSOs such as ISO, DIN, BSI etc are pretty old fashioned publishing models. They charge a fee for the document; there is no annual fee or any fee associated with the implementation of the standard. So only the developer interested in reading the standard would need to pay; in fact, to implement a standard you don’t need to even read it if its embedded in an existing library, or there is sufficient freely-available general guidance material to implement it.

> Thanks for some clarification in advance.
> 
> Ps:
> Hm, do I have the impression that "open standards" could grow a good candidate for: wait..
> http://www.gnu.org/philosophy/words-to-avoid.en.html <http://www.gnu.org/philosophy/words-to-avoid.en.html>

Well yes. “Open” can cover a multitude of sins and virtues in the standards world. Many of the more “open” and “agile” freely-available standards folks like to use on the web implement what I like to think of as the “golden rule” governance approach - you bring the gold, you make the rules! - which favours the large vendors and service organisations (Google, FB etc). Whereas the rather closed-looking formal BSI, CEN and DIN processes are ironically quite open to engagement by individuals and SMEs, but the resulting published standards cost money to buy.  YMMV.

- Scott

> ?
> Ah I see the word "open" is already in.. thunderclap.
> keep healthy.
> 
> 
> 
> 
> 
> 
> 
> 01.06.2015, 14:20, "Carsten Agger" <agger at modspil.dk>:
>> I personally believe that open standards should be free to share (not necessarily to modify) and to implement, and should be available for download on the Internet, and also free to share in hardcopy. Standard bodies might be allowed to charge a fee for verifying compliance.
>> 
>> Just like the Internet standards are available in the RFCs.
>> 
>> 
>> 
>> 
>> On 06/01/2015 02:07 PM, Tom Blecher wrote:
>>> 
>>> Hi,
>>> unfortunately I stick still on the linguist layer at determine what is the fsf position to standards such as
>>> din, iso and the like.
>>> http://en.wikipedia.org/wiki/Open_standard <http://en.wikipedia.org/wiki/Open_standard>
>>>     http://en.wikipedia.org/wiki/Reasonable_and_non-discriminatory_licensing <http://en.wikipedia.org/wiki/Reasonable_and_non-discriminatory_licensing>
>>>       but the modalities discriminate against a whole category of intangible goods such as
>>> free software <http://en.wikipedia.org/wiki/Free_software>[6] <http://en.wikipedia.org/wiki/Reasonable_and_non-discriminatory_licensing#cite_note-6>
>>>      The
>>> Free Software Foundation <http://en.wikipedia.org/wiki/Free_Software_Foundation> suggests the term "uniform fee only" (UFO) to reflect that such "(F)RAND" licences are inherently discriminatory.
>>> So if somebody can help me out?
>>> 
>>> 1. is Din or p_iso to be considered as ufo?
>>> 1.1 is Din or p_iso considered fsf's fight against limited to "patents". Cause DIN or such are not patents.
>>> 1.1.1 Anyway I find it rather simple reckoning how dirty DIn harms the same way? No? Even the questions come up if a software that integrates such drugs can be called GPL-compatible. Why? Are there any duties of paying fee to propaganda complexes whithin the GPL for a source code understanding user? No.
>>> 
>>> 2. Is this output from standardization propagandists any relevant in terms of "open standards", cause these cover explicitly only "formats" and "protocols", which is "Din" apparently neither nor.
>>> 
>>> 3. Is there maybe some middle wide recognition gap, where propaganda causes fs-harming ufo-standardization to be still unnoticed. to be overseen, to be even protected that way? Is it  that we are dealing with, actually? A forest for a tree problem?
>>> 3.1 It is that I am interpreting Nico's post: "interesting". So what could be interesting or new specially in this issue for you long timer?
>>> 3.2 is the document freedom day then claiming for non-ufo standards in broader sense, including for example "DIN or ISO Standards"?
>>> 
>>> Thank you for an answer? And thanks you for any comment, it would help me.
>>> 
>>> 
>>> 
>>> 
>>> 30.05.2015, 10:40, "Nico Rikken" <nico.rikken at fsfe.org> <mailto:nico.rikken at fsfe.org>:
>>>> Dear Tom,
>>>> 
>>>> This has crossed my mind as well. Although I wasn't aware about
>>>> standardization organizations offering these standards free of cost. In
>>>> the Netherlands one related aspect has been taken to court, namely that
>>>> some of the laws refer to standards which aren't available freely or
>>>> free of cost. It was ruled that this was not particular issue, as the
>>>> cost was justifiable for setting and maintaining the standards, and the
>>>> standard was available in a non-discriminatory fashion (if I remember
>>>> correctly).
>>>> 
>>>> The collection of standardization bodies are quite complex, with
>>>> national organizations, industry-specific organizations, and
>>>> international organizations (ISO, EN, IEC), often approving each other's
>>>> standards. Coming from a power systems background, standards defining
>>>> electromechanical systems like fuses, power cables and circuit breakers
>>>> is very industry-specific and is mainly of interest to manufacturers and
>>>> system engineers, which then again are mostly larger organizations.
>>>> Somewhat remarkable my university has stopped adopting standards because
>>>> the little use in academics didn't justify the cost of the license.
>>>> 
>>>> The main difference with software standards, and web-standards in
>>>> particular seems to be that even individuals have the ability to create
>>>> a working product, as no industrial manufacturing process is required.
>>>> Adhering to closed, costly standards would be much more significant,
>>>> unless maybe a reference implementation (library) would be available for
>>>> use, removing the need for the actual standard to be read. So the cost
>>>> of common software standards is therefore required to be approaching
>>>> zero.
>>>> 
>>>> Scott's writing on standard adoption explain the way in which project
>>>> can adopt standards and the many issues related to bringing about open
>>>> standards.
>>>> 
>>>> I was reluctant to read an article by Gijs Hillenius in the Dutch Linux
>>>> Magazine regarding the updated Open Source strategy of the European
>>>> Commission, in which he pointed out that the EC was explicitly
>>>> considering open standards in favor of other established standards. I
>>>> consider this to be the confirmation that not-open standards are non
>>>> preferable in relation to free software.
>>>> 
>>>> As society seems to become more decentralized and dynamic, the
>>>> conventional standardization model will be under ever more pressure to
>>>> lower the barriers of access, regarding cost, license of use, and
>>>> transparency of process.
>>>> 
>>>> Thanks for bringing up this interesting topic. I'm very interested to
>>>> hear the viewpoints and findings of others on this as well.
>>>> 
>>>> Kind regards,
>>>> Nico Rikken
>>>> 
>>>> ,
>>>> _______________________________________________
>>>> Discussion mailing list
>>>> Discussion at fsfeurope.org <mailto:Discussion at fsfeurope.org>
>>>> https://mail.fsfeurope.org/mailman/listinfo/discussion <https://mail.fsfeurope.org/mailman/listinfo/discussion>
>>> 
>>> _______________________________________________
>>> Discussion mailing list
>>> Discussion at fsfeurope.org <mailto:Discussion at fsfeurope.org>
>>> https://mail.fsfeurope.org/mailman/listinfo/discussion <https://mail.fsfeurope.org/mailman/listinfo/discussion>
>> ,
>> _______________________________________________
>> Discussion mailing list
>> Discussion at fsfeurope.org <mailto:Discussion at fsfeurope.org>
>> https://mail.fsfeurope.org/mailman/listinfo/discussion <https://mail.fsfeurope.org/mailman/listinfo/discussion>_______________________________________________
> Discussion mailing list
> Discussion at fsfeurope.org
> https://mail.fsfeurope.org/mailman/listinfo/discussion

OSS Watch - supporting open source in education and research
http://www.oss-watch.ac.uk

scott.bradley.wilson at gmail.com
scottbw at apache.org
http://scottbw.wordpress.com
@scottbw

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.fsfe.org/pipermail/discussion/attachments/20150602/148563a2/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 841 bytes
Desc: Message signed with OpenPGP using GPGMail
URL: <http://lists.fsfe.org/pipermail/discussion/attachments/20150602/148563a2/attachment.sig>