What is Open? Ask IBM. Its a Four Letter Word. A Sutor-ble apology

While i was away on holiday in Greece for one week Stephen posted 24 blog entries… all of them solid. If only he paid so much attention to his day job…. 😉

Anyhow, Stephen evidently took a bit of stick (mostly from IBM) for calling Microsoft’s forthcoming Office XML format “open”. But you know what – over the course of the week he explained himself and took some heat out of the argument (oh for such perspective, and ability to avoid binary extremes…). For us, it felt like a seminal moment in open source analysis. The snowball rolled downhill, and we ended the week smarter than we started it. IBM employees were participants, and made great contributions.

Given that RedMonk was quoted on the OASIS OpenDocument press release thus i figure our position is the right side of wrong. But of course, that is just context.

Stephen did some great definitional work here. Unlike Stephen, I would probably have pointed out (pace Sean McGrath) that defining “open” is like defining obscenity – I know “open” when i see it…. But then again, Sean and I have both read Wittgenstein so we’re stuffed when it comes to eau de whatever (uisge beathe notwithstanding)… its all about family resemblances, not an essential Platonic essence of open.

Open is a four letter word.

Anyway the point of this post is to make an apology to Bob Sutor and IBM, which may also help to explain some of my thoughts on “open”. First some context.

I have known Bob for a few years now. He is neverendingly polite in the face of my sometimes aggressive questions and atitudes. He does have a bit of a Janus face, as must any deep thinker thinking about commercial concerns, balanced around standardization and commoditisation. Bob is as responsible as anyone of helping to create the WS-I and its welter of XML standards. Bob is particularly interesting in terms of his views on standards and their role because he entered the world of de jure standardization as an academic/research guy, but after his experiences in this field, he became substantially more hard-bitten and de facto oriented. After taking on a WebSphere marketing role for a while, he is back into standardization, leading all IBM efforts in the area. For those that don’t know IBM, this role makes him one of the most influential executives in the company. IBM’s business strategy for the next five years–i think its called On Demand or something-is predicated on partner clustering and standardization, of everything from server componentry (see BladeCenter-T) to XML web services and ITIL (service management), even vertical industry processes, enabling collaboration, and so joint innovation. Ask Irving.

Am I ever going to get to the darned apology? And what do i have to apologise for? And why not just apologise to Bob?

The nub of it is royalty free. I have long been a fan of ebXML because its royalty free and free of patent liabilities, and so allows for innovation in the field by smaller and bigger players, without fear of chilling effects. The great thing about ebXML is It originally came out of UN EDI work, designed to allow ecommerce in emerging geographies, with lots of other laudable goals. The problem with ebXML was of course that it… came out of UN, with its culture of consensus, which means it doesn’t always move so fast. And doesn’t always make the right decisions.

Politics- politics is just brutal. its ugly, often pointless, and involves constant equivocation.

It would be remiss of me to tell tales here (lets wait for Mr Sutor’s book for that, whenever he gets round to it) but suffice to say “working with” Oracle and Sun on ebXML was rather wearing. Kind of like trying to get France, the UK and the US to all agree on set of policy goals and instruments for achieving them. No problem there then…

In fact the ebXML experience was so brutal that IBM and Microsoft just decided to say screw it, and start again, with an approach combining de jure and de facto standardization. In effect, the approach became if we standardize it they will come. If we can build business process orchestration across IBM mainframes to Microsoft smart-phones then everyone else can just fall into line. IBM, having handed over tpaML (trading partner agreement markup language) as a core underlying component for ebXML, walked away from the technology, at least in terms of “pure Web Services”. Bob told me shortly afterwards that ebXML is dead.

He repeated this line to me a couple of time, on different occasions, which doesn’t make him right. I am still a contrarian on this core. Anyone that thinks ebXML is dead isn’t considering the healthcare market (PDF, sorry) and international disease control efforts [what is more life and death than disease control, Mr Champion?]… many active end points will be ebXML component based; healthcare initiatives mandating ebMS won’t be swapped out anytime soon.

Lucky IBM can support ebXML through WebSphere Business Integrator using connector structures and transformation with some of the platform’s slick new tooling.

Ironically, the de jure standard has shown itself more willing to embrace and extend WS-I bases, than the other way around. Want to use SOAP? Sure you do – that is why ebXML adopted it as a transport mechanism… But if you want to avoid patent encumbrances its probably best to avoid MS-developed standards. Like i say, i know open when i see it. Is Microsoft open? Sure it is. But its also more focused on IP protection than any vendor i know of – Macrovision is included here. Oh yeah maybe IBM Global Services (more fixated on IP protection that any other part of IBM, because its in the business of selling ideas rather than technology. The “Other IBM”, that is).

WS-I is resolutely RAND. But RAND is bullshit [just my opinion, but i think its a tremulous concept, “reasonable” terms are far harder to define than obscenity, open or “I”…

Reasonable to who? The EU? Microsoft? IBM?

Anyway, i am getting to that apology.

I basically figured that the pragmatic-business-focused,-win-advantage-through-things-like-RAND-combined-with-patent-and IP-protection, Sutor had come to the fore [remember i posited a Janus face]. I assumed that WS-I, and the Microsoft lockstep, was about advantage to IBM, rather than standards because they are a good in themselves.

Well as we now know, things havent been plain sailing for WS-I. Some people have said it is dead – i have argued it could be niched…

I know IBM does a ton of open stuff but you always have to ask what’s the benefit to the organization? IBM has been making shareholders happy for a 100 years and Mr Palmisano has no plans to stop now.

IBM then, not just Bob, has a Janus face. But don’t we all? Its perfectly possible, downright sensible, to argue for open standards if they are going to benefit your business. Is IBM “really” open. What do you think? Linux, XML Web Services, Eclipse, PHP, Gluecode.

For all that, i can’t help my skepticism. I believe in standards. really “open” standards. I know ’em when i see ’em. I like ’em. Oi, Listen right!

So when OASIS first announced it would allow for RAND standards in its governance structure, my first reaction was: “this has IBM’s finger prints on it.” I sctually said “Bob Sutor will be pleased”…The folks from OASIS were a little surprised. As it turns out I was not the only outsider to be rather queasy about OASIS’ decision. From a patent unencumbered, royalty free standards body to something rather different in one step.

But you know what? Bob had nothing to do with it. He had not been lobbying for OASIS to change its terms. Indeed Bob is a big supporter of the OASIS OpenDocument spec, and credits Sun accordingly.

Yes I know IBM recently announced 500 patent protections if methods were “used in open source” [why is it IBM can do some rights reserved while Sun gets slammed for it?]. But i figured some of these patents were somewhat orthogonal to many open source developments. I have no objections to some rights reserved – but some rights reserved means “less open”, right?

IBM’s recent work with OASIS has all been royalty free. Well done Bob. Well done IBM.

I apologize to Bob; he is evidently a friend of royalty free. Which is good.

Did i say how important standards and “open” are going to be to IBM? How come? Look out Microsoft and others. IBM isnt just thinking technology standards. Far from it. IBM isn’t just thinking about IT patents – no its thinking genomics and every other big business problem and opportunity on the horizon, automotive, industrial design, electronics, gaming and all sorts.

While Microsoft is likely to go along with Big Media, or big Pharma, or any other big when it comes to protecting and ring-fencing IP (that is, until the current senior management step aside, which could be a rather long time). Microsoft’s first thought it to protect IP rather than collaborate and open it up, Shared source and blogging notwithstanding. From the top down the mantra is control.

IBM isn’t bullshitting when it says it will compete on the basis of “open”. Has any other vendor got such brand recognition on that word? Lord knows IBM has invested enough in the last 10 years to associate itself with “open”. At IBM the multiplicity of voices is more open than not…

IBM has a clearly differentiated vision- Bob will be talking to retailers, manufactures, financial services companies and others about the benefits of open intellectual property. Telling them to open up. Trustworthy? That means a handshake, right, not 256-bit enryption? Now all Sam Palmisano needs to do is mandate IGS to start sharing in the way STG and SWG do.

From my perspective IBM is siding with the angels here. Science and innovation are stymied by IP protection. Indeed, corporate scumbags use “IP protection” as an excuse to, for example, hide negative research results about the drugs they sell. Shrub recently argued that the US is doing more research into global warming than anyone else – yes George but you’re also hiding and whitewashing more of this research than any one else.

I believe business needs to be more like science, not the other way around, if we are to solve the big problems of our time.

One of the points i can see from last week’s open discussion, and the fierce argument from IBM that to be truly open a standard must engender participation, is that this seems a clear architecture of participation argument. If we buy into architecture of participation, its clear that standards help them along. These thoughts are based on Bill Higgins’ “open” matrix.

Bob will be agitating in that direction. Open standards, open source, open minds, open opportunities. indeed.

So what does Bob believe in his own words, Janus notwithstanding?

To me, that document formats for “office” applications should be completely open, not hindered by patents, and not owned by a single vendor is just obvious. I wasn’t brought up to think otherwise, and so this whole business around why everyone should be rushing to implement the new OASIS Open Document format standard is a big “duh” (that is, slap-yourself-in-the-head-obvious).

Just one question Bob – why not apply this logic to business payloads too? Open:

is available on a royalty-free basis and has no restrictions that might limit its use for any reason in any software, be it customer-unique code, a vendor product or open source.

You’ll be reassessing ebXML next.. 😉

Any way – sorry Bob and thanks for all your good work.