In article <61esb5-igq.ln1@xxxxxxxxxxxxxxxxxxxxxx>,
Mark Kent <mark.kent@xxxxxxxxxxx> writes:
> Craig Gullixson <craig@xxxxxxx> espoused:
>> In article <up9pb5-l3s.ln1@xxxxxxxxxxxxxxxxxxxxxx>,
>> Mark Kent <mark.kent@xxxxxxxxxxx> writes:
>>> Roy Schestowitz <newsgroups@xxxxxxxxxxxxxxx> espoused:
>>>> ____/ [H]omer on Tuesday 25 March 2008 22:31 : \____
>>>>
>>>>> Roy Schestowitz wrote:
>>>>>
>>>>> [quote Martin Bryan]
>>>>>> The disparity of rules for PAS, Fast-Track and ISO committee
>>>>>> generated standards is fast making ISO a laughing stock in IT
>>>>>> circles. The days of open standards development are fast
>>>>>> disappearing. Instead we are getting ?standardization by corporation?
>>>>> [/quote]
>>>>>
>>>>> Or ?standardisation by corruption?
>>>>>
>>>>> One thing to bear in mind, is that the demise of ISO, and other
>>>>> standards bodies, would suit Microsoft's purposes perfectly, indeed it's
>>>>> likely that they /devised/ their downfall to begin with.
>>>>>
>>>>> If Microsoft succeeds in corrupting standards bodies then they win
>>>>> standards ratification (even if those "standards" are unworthy), and if
>>>>> (despite that corruption) they fail in their efforts, then they still
>>>>> "win" by destroying standards bodies (in reputation at least, which for
>>>>> a standards body is /everything/), thus making impartial standardisation
>>>>> irrelevant. All that will remain is *de facto* standards, which
>>>>> Microsoft maintains by protecting its monopoly (and vice versa).
>>>>>
>>>>> It's a win/win deal for the Vole, either way.
>>>>>
>>>>> IOW this is just Microsoft's typical MO of "assimilate or destroy".
>>>>>
>>>>> I see exactly the same thing happening with Mono, which will either
>>>>> assimilate FOSS into Microsoft's portfolio of Intellectual Monopoly, or
>>>>> destroy it with distrust and division. I'm sure that's the plan, at least.
>>>>>
>>>>> The only hope to stop and reverse this corruption, is to stop Microsoft
>>>>> ... permanently. The only hope we have of doing that, is with the law.
>>>>> But the law is in the pocket of corrupt politicians and corporations, so
>>>>> that just leaves a handful of crusaders like Lessig, or on the other
>>>>> side of the pond - Kroes. And people like us who expose that corruption.
>>>>
>>>> Kroes sort of fell for the RAND scam. Hopefully she'll realise this soon. The
>>>> Europa site has recently issued several press releases which favour FOSS. It
>>>> did so very quietly (low profile), some say because it does not want to give
>>>> the impression that it /actively/ fights the abusive monopolist.
>>>>
>>>> As you say, dead standards bodies and division serve Microsoft. The Novell deal
>>>> is another example and all have crossed my mind before. Money corrupts.
>>>>
>>>> In a land/state of chaos, the villains rule.
>>>>
>>>
>>> The death of standards serves the rise of foss well, though. As
>>> standards become more and more blantantly corrupt, like OOXML, then the
>>> alternatives, such as ODF, which are essentially defined by their source
>>> code, will become important.
>>>
>>> The need for traditional standards is really a hangover from the days of
>>> 100% proprietary equipment and code, such that the only way of getting
>>> any kind of interoperability was through a standards definition, whereas
>>> now that source code is available and we have COTS hardware, then the
>>> need for standards has waned significantly.
>>>
>>> There will always be a need for standard hardware platforms and
>>> reference designs, but the source suffices for software, file formats,
>>> protocols and so on.
>>>
>>
>> I beg to differ. As a developer, all I really care about is the
>> published standards.
>
> You should get some experience writing standards, too. It would help
> you understand the issues a little better. Standards for file formats
> require mentally imaging code which hasn't been written, for example.
> This is, at best, a rather silly thing to try to do.
>
I am not completely unfamiliar with standards. While standards
documents tend to be long, and often contain code snippets as examples,
they do tell you what you need to do to implement the standard. For
example, after a quick look at the ODF standard, I'm fairly confident
that I could build my own ODF document creator tool if I wanted (and
had the time) to do so. I don't have to reverse engineer code, not
necessarily written in a language that I'm completely familiar with.
>> Source is nice to have as a reference
>> implementation, but parsing such source to determine file formats,
>> protocols, etc., tends to be both difficult and propogates bugs in
>> the reference implementation.
>
> Err? If the source *is* the reference implementation, then as and when
> bugs are found, they can be fixed. The speed of open-source development
> is well in excess of anything every achieved by the standards process.
> Subversion, sourceforge etc., ensure that this can be quickly and easily
> achieved.
>
Between the time when the reference implementation is released and the
time a bug or logic error is discovered, people have been using the
reference implemtation as a bible in your scenario. One can try to
propogate fixes throughout ciberspace, but effectively, the bug or
logic error is now a part of the standard (much as the leap year bug
Microsoft Excel).
Yes, the speed of open-source development can be quite rapid. This
is a good thing as there can be a rapid response to bug reports, feature
requests, etc. However, it should be recognized that there is a down
side to a rapid pace of evolution such as new introduced bugs, interface
instability, keeping things that depend on a given package current with
the release of the package being used (i.e. the configuration management
thing).
And yes, standards development is a slow and ponderous process, often
taking years between updates. This is also a good thing as changes
can be well thought out, and a bad thing as it generally takes a long
time for new features to propogate throughout the system.
I actually think that the two approaches are complimentary. Useful
and popular features out of the rapid development cycle can become
a part of the standards, and the standards can serve to reign in
the chaos created by millions of developers doing their own thing.
>> Standards promote competition and
>> prevent vendor lock-in.
>
> Quite incorrect, standards are actively abused on a regular basis in
> order to promote lock-in. Open-source and open platforms, on the other
> hand, promote competition and prevent vendor lock-in.
>
Lock-in due to standards abuse, usally in the form of additional
"features", only happens if you fall for their tricks.
Don't get me wrong, I think that open source development is great
and I have been using open source packages for *decades*. However,
I also write code in a production environment where it is expected
to run with minimal maintenance for years. Durning those years,
I expect that hardware will be upgraded or replaced, there will be
operating system patches and upgrades, and so on. In order to deal
with this, the code is written with portability in mind, which, in
turn, heavily relies on the stability provided by standards.
---Craig
________________________________________________________________________
Craig A. Gullixson
Instrument Engineer INTERNET: cgullixson@xxxxxxx
National Solar Observatory/Sac. Peak PHONE: (505) 434-7065
Sunspot, NM 88349 USA FAX: (505) 434-7029
|
|
