Tempest in the standards world

2006-06-08 08:37 | Permanent Link | Standards

Given that I seem to be a central figure in the drama that Howard Butler has written about and that Chris Tweedie responds to maybe I should say something. And given that my blog has been a bit, shall we say, dormant lately, this is opportune.

As I said in my posting to the list,

*I believe it’s in the best interests of keeping GeoRSS open and available to everyone for us to keep it under a Creative Commons license.*

But why? And why not let OGC take over development of the spec? If I'm willing to let people profit from the application of the spec, why would I not let an organization profit from the development of the spec?

Why is Creative Commons important?

The mere fact that the Creative Commons license was able to stop OGC from taking the contents of the GeoRSS.org site, dumping it into a document, and submitting it to its members is clearly a major plus. (Let me say at this point that Carl Reed of OGC is reworking the document and has been asking me for my opinion on how to rework it, so maybe the next version will be better). Let's assume for a minute that OGC's motives were benign. Great. But this would presumably have also worked with an organization whose motives were not benign. It might have taken a bigger stick, but I think the protection afforded by the license would prevail.

Why not let OGC take over development of the spec?

As much as the OGC staff will deny it, the OGC process is geared towards those who can afford to participate. Entrance fees might seem low to them, but they are high to the kind of people on the GeoRSS mailing list. Travel costs to get to meetings are even higher. The OGC Portal is password protected and only members can see the contents. I've tried to convince OGC staff and members to release working documents and have been rebuffed.

Why would I not let an organization profit from the development of the spec?

That one gave me pause, even though it's my own question. So now I wonder, why not? What if I could design the ideal spec development process? Would I care how the process is financed if the result is an open spec everyone can use?

The ideal spec development process should have the following characteristics:

Anyone can participate at no cost if they have access to email and the web and all email records, documents and other artifacts pertaining to the development of the spec will be available to everyone at all times (again at no cost).

All other characteristics would follow from that. Details would need to be worked out. So, if an organization comes along and says "we can support that", what then?

It turns out I do care about this. In the case of who benefits from the spec itself, I believe that the beneficial aspects of the spec's existence will outweigh the negative aspects. It's not posssible to control who uses it for good or for evil, as it were.

But when it comes down to which organization I support directly by my efforts, it matters a lot. Right now, the first compellingly neutral organization that jumps to mind is the Open Source Geospatial Foundation.

I wonder if the time is right to have the Open Source Geospatial Foundation become the sponsor/host/supporter of GeoRSS?

Comments

2006-06-08 18:52 | Posted by Chris Tweedie | http://chris.narx.net

The last point made me think for a moment Allan. Does OSGEO even want to get into the pseudo-standard space? If they do, it may cause some interesting conflicts in the future

2006-06-08 19:58 | Posted by Allan

Chris - I put in a trackback to your article but it seems to not have worked, Thanks for the link back.

I don’t know if OSGEO wants to do standards or even pseudo-standards. By conflicts do you mean internal or with other organizations?

Let’s assume you mean internal, that there might be two groups working on roughly the same thing.

I see no specific reason not to be working on more than one spec that covers the same topic, particularly if they are different enough in approach. I’d rather have two well thought out things to deal with than a hodge-podge. Also, the two camps might learn from each other and might realize they are not that far apart. Better to have that happen in an open environment than a closed one.

Also, you have to assume that when the time is right for an idea, more than one person or group will start working on it. Again, better to have some open process so people can notice each other.

Then let’s distinguish between development of the spec and ratifying the spec. I think people should get less hung up on premature ratification. It forces you do behave in silly ways. The IETF model was to only consider specs that had at least two independent implementations that are shown to interoperate. That means that the spec maturity was high enough to have gained some traction. I’m not sure if they do things exactly that way anymore, but it’s a good model.

So forget about the “standardization”, let’s develop some specs. The good ones will bubble up and win out. If they don’t it’s more than likely due to personal ego issues or business reasons usually dreamed up by the marketing departments, not the techies.

2006-06-12 21:35 | Posted by Jeff Harrison | http://www.thecarbonportal.net

One simple question always amazes me - Why do people feel it is necessary to have any organization or standards bodies involved to advance open specifications? Why can't you and I do it together if we so choose?

One thing I learned early on in my career with internatial standards is that in the world of the Internet and the Web there is nothing preventing people from publishing an API or specification, double checking for any IP issues, and making it royalty free under an appropriate license and there you go - open specification or API.

The issue gets more complicated when governments get involved and the definition of a "standard" is invoked for procurement purposes. But you'll survive.

Here's a hint - if you have an open specification for interoperability and want to share it, just collaborate with folks, obtain consensus, publish the thing under Creative Commons and (here's the important part) IMPLEMENT IT. Later, if it has "wings" you can choose an appropriate body for caretaking (or not).

Regards

Comments are closed for this entry.