[OSGeo-Standards] Re: [Board] Open Source and Open Standards White Paper

Thu Apr 14 17:29:28 PDT 2011

Arnulf,
Sorry for the delayed review, and rather messy use of email in applying 
my review comments:

====

I assume that the target audience for this white paper is:
* People who don't understand the difference between "Standards" and 
"Open Standards"
* People who don't understand the difference between "Proprietary 
Software", "Freeware", "Open Source"
* People who may understand the terms "Open Source" and "Open Standards" 
but don't appreciate the strengths (and weaknesses) of these models over 
the alternatives.

As a general comment, I feel that much of this paper uses terms which 
are only understood by seasoned developers who already understand what 
Open Source and Open Standards are, and as such I feel that the paper is 
"preaching to the converted". For instance, there are concepts which are 
introduced, like the Open Sourced development practices, which are left 
hanging without an explanation. A reader not familiar with Open Source 
will be left wondering "so why should I care about these practices". I 
suggest that if these concepts are introduced they need to be expanded 
upon and describe why they are important.

====

Further are some more specific comments:

* 1st Introduction paragraph:
This should be concise, and describe what the white paper is about. It 
is not the place to reference source material. It needs to include a 
sentence or two describing what the OGC and OSGeo organisations are.
Of note: I have been at many geospatial trade shows and almost always 
talk to a few people who don't know what the OGC is, or what OSGeo is.

* Under "Open Standards":
** I feel the first few paragraphs should define what an Open Standard 
is in the first paragraph before expanding.
** I feel that the section "OGC's position regarding Open Source 
Software" does a better job of describing importance of standards than 
this section.
** The list of criteria for a standard is good. I'd like to see each one 
of these points expanded on, and provide a description of why that 
criteria is important. In particular, I suggest that a counter example 
will usually benefit the description. Eg: Explain the limitations with a 
company publishing an API then calling it a standard?
** After reading this section, someone should understand why it is 
important to use Open Standards. I'd expect to see discussion of 
Inter-Operability, Reduced cost of data processing, Reduced long term 
maintenance cost, reduced risk due to obsolescence of products, reduced 
risk of vendor lock in.
** I've recently been refuting claims (at a National Government Policy 
level) that the ESRI REST API is an Open Standard because REST is an 
open standard. While I don't suggest picking on this as an example, it 
would be valuable if the reader can understand the difference between 
extensions to a base standard, and an overlying API.

* In "Proprietary Software", I suggest that you don't pull out "java" as 
a specific example. It is too "down in the the weeds" for this type of 
high level document, and will likely confuse rather than help readers. 
If the java issues are to be explained, then much more detail should be 
included about the specific java case, and I don't think that is 
appropriate here.

* In "Open Data", I think this section requires a stronger statement 
about what Open Data is. We should state what we believe "Open Data" to 
mean. The Creative Commons licence should be mentioned, and note how it 
is applicable to documents with just a few authors. We should note how 
Open Street Map is moving away from Creative Commons due to the 
practical limitations associated with crediting thousands or millions of 
authors, as is often the case with geospatial data sets.

* In "Similarities and differences":
** terms such as “benevolent dictator” and “rough consensus” should be 
explained if they are to be used.
** IPR accronym should be expanded.
** I feel the discussion of the OGC "Reference Architecture" is 
wishy-washy. We have some very clear guidelines about what is Open 
Source and what is not (as explained above). Either say that the 
Reference Architecture is Open Source, or don't mention it.

* OSGeo's position regarding Open Standards:
** Re: "... ideally simply [standards] because they are good." "good" 
needs to be expanded if this statement is to be made. Why are simple 
standards good? I think this statement should be dropped, or else it 
needs a paragraph or two discussing the pros of a comprehensive standard 
vs the pros of a simple standard. I think the answer to this question is 
more than "simple is the best version of a standard in all cases". I 
think it would be closer to "it is preferable that all standards, even 
complex standards, include a simple mandatory core version of the 
standard which is easy to implement which leads to widespread adoption 
of the standard."

* Memorandum of Understanding:
** I assume the 6 OGC memberships are for "OSGeo members"? If so, this 
should be mentioned.

* Please be consistent with capitalisation: Either ("open source" and 
"open standards") or ("Open Source" and "Open Standards").

On 15/04/2011 4:55 AM, Seven (aka Arnulf) wrote:
> Folks,
> please be so kind and give this paper a moment of your attention:
> http://wiki.osgeo.org/wiki/Talk:Open_Source_and_Open_Standards
>
> There is still a lot of confusion in people as to what an Open 
> Standard is and what Open Source is. Even within the educated 
> community of OSGeo there are very different positions wrt to standards 
> and what an Open Standards should be. It ranges from "standards suck 
> and prevent innovation" to "standards are the only way out of the 
> misery of the data silos".
>
> The same is true to what the educated part of the OGC community thinks 
> about Open Source ranging from... well you know. No need to repeat here.
>
> Therefore OGC and OSGeo are interested to find a common position on 
> Open standards and Open Source and if we don't disagree too badly this 
> will become a joint white paper. Please give it a pass and comment if 
> you think something is seriously wrong.
>
>
> On 04/13/2011 09:42 PM, Markus Neteler wrote:
>> On Wed, Apr 13, 2011 at 9:02 PM, Seven (aka Arnulf)<seven at arnulf.us> 
>> wrote:
>>> Folks,
>>> as discussed on IRC we have further developed the article on Open 
>>> Source and
>>> Open Standards:
>>> http://wiki.osgeo.org/wiki/Talk:Open_Source_and_Open_Standards
>>>
>>> It would be great if you could give it a pass
>>
>> I have done a few edits (see history).
>>
>> Markus
>
> Thanks Markus,
> anyone else from the board with comments? The page had 100 views in 
> the past two days and if no one else follows up with comments I 
> consider this to be generally accepted.
>
> Best regards,
> Arnulf.
>
> PS:
> I am sure that as soon as we publish it the usual die hards will start 
> to scream at the top of their lungs :-) but we can probably not do 
> much more than we did, can we?
>

-- 
Cameron Shorter
Geospatial Solutions Manager
Tel: +61 (0)2 8570 5050
Mob: +61 (0)419 142 254

Think Globally, Fix Locally
Geospatial Solutions enhanced with Open Standards and Open Source
http://www.lisasoft.com