Introduction
Italics in these comments denote quotations from your
papers, except where it is clearly not the case. First author firstnames
in brackets are references to my comments on your papers, in approximate
order of decreasing relevance here of my comments -- and not of the relevance
of your papers! Numerals refer to my answers to the questions in my faq.
"[paper(innocent questions)]" refers to the section in my paper
headed "Some innocent questions for the OMG...". "[3(firstly)]"
refers to the "Firstly" item of my answer to faq question 3.
Hopefully the hectic start of the
new year has passed for the academics among you, and all of you are hopefully
still in the reinvigorated surge of the post-vacation and post-OOPSLA season?
If so, good. It seems to me from far that this Business Object Workshop
still has much unfinished business.
Thank you Jeff for your continued support, despite all: I am still hiding the details of MACK (leaving that apparent technical "black hole" - see under "trade secrets" below), so there is still no easily convincing evidence of my proposed alternative to the current favourites such as Java, the Web, ActiveX, CORBA etc. That weakens my present criticisms of them (and I content myself with Medawar’s Dictum: "Theories are not displaced by facts, they are replaced by better theories." [paper(different perspective)]) So, for an idiosyncratic intervention such as mine to be given such support, the observed yet unmet needs of our industry - not to mention those of our customers - must indeed be enormous! (And that is of course nothing new to Jeff: see [Jeff(need for tools)].)
To repeat myself from my introduction to the earlier mailing-listed version: my comments are not intended to be balanced assessments of your papers in the context of conventional software approaches. I apologize if I sometimes rather rip your words out of your context and put them in my own! My bias is obvious. It is part of the plan as set out in this document.
However, I do also find excellent growth-points in every single one of your papers! So I point out what in your papers seems so worth-while, even in my so different scheme of things. Thank you for those insights, and I look forward to your growing them even faster with Metaset in the reasonably near future!
Maybe also - see under "Strategy" and "Looking ahead" below - you are in a position to help expedite that convergence?
Objectives of these comments
I have a very big scene to convey, and - much though I seek criticism now - I am still choosing to pitch my story at a level of abstraction that does not facilitate either concrete criticism or full comprehension, as explained under "trade secrets" below. (I know that the fuller comprehension will come more easily - with Metaset’s help - in due course, and everyone will see MACK as obvious and even Mainstream! [paper, 3])
I thank you all, therefore, for the opportunities your papers have given me to start with perspectives that are familiar to you and, bit by bit, build up a hopefully more helpful picture of what Metaset/MACK is about, why it is so different, and why I think my story should at this stage and in this form nonetheless have considerable credibility. That might then help lead to the short-term impact sought, namely the right kind of joint project to make it happen sooner and better (See [11] and under "Strategy" below).
My comparing and contrasting between your papers and the MACK approach has thus provided numerous points of contact with your own preoccupations (Thank you also for helping to clarify my marketing stance!), and an opportunity for me to fix Metaset/MACK in the IT universe without revealing the black hole which I am still preserving in the centre of my public story. Hence the following rather dense summary leading in a top-down way - via my comments on them - into your papers, and into my faq:
The uniqueness of MACK
"MACK is applied epistemology. It was designed to mirror the way we live our knowledge in a world of information system penetration, integration and wide-spread distribution, set against the background of the complexities of real life whose potential should uplift, not mire down." ([paper(Homeric keel)], italics in the original)
MACK grasps complexity head-on. Explicit interconnectedness, which is our best objectively sharable equivalent of complexity, is an opportunity not a problem. Appropriate conceptual simplification allows indefinitely complex models and more mechanically-reflective yet usefully-creative pursuit of coherence and consistency [paper(Homeric keel)]. That makes our models more easily sharable through simple views relative to their various users’ needs, interests and activities. That enables widespread and constructive contact between abstract model and empirical reality. That way we make reliable and really-livable knowledge from data, building on complexity rather than hindered by it [Ralph, Jeff, Wolfgang, paper(synthesis)].
It will, I suspect, bring us far closer than do conventional architectures to the way our minds usually work anyway. A comparable reflectivity-enabled flipping of complexity-as-a-problem into complexity-as-an-opportunity could help explain why homo sapiens has so outstripped its natural competitors: quite suddenly during our biological evolution, the human brain grew and with it our ability to survive and adapt entered the steeply exponential phase of a sigmoid growth curve. That would then further presage the dramatic outstripping of its competitors by the MACK-enabled Profoundly-Congenial PDA as a really-effective extension of the brain, as I have already said it presages the anticipated sudden flowering of the self-booting Metaset program itself. That was in this last paragraph from [paper(Homeric keel)]:
"But is all that plausible? Well, that’s the way Metaset is already booting itself. It is presently still below the elbow of the growth curve where an unconstrained self-booting process suddenly reflects the exponential, critical mass or take-off stage. But that delightful phase is taking ever clearer form, as are some of the constraints that will in due course give the curve its more natural sigmoid shape."
The simplicity of the architecture goes hand-in-hand with its epistemological appropriateness. It is justifiable to dare to build on transparent state as well as opaque behaviour [Stefan, Wolfgang]. That allows the clean incorporation of the simple logical principles behind OO, including the multiple inheritance that is essential to user-relativity and associated complexity-hiding, in contrast to Java and Smalltalk, for example [Ralph, Jeff, Mark].
The problems of designing, building and running IT systems are better addressed using a general-purpose and uniform hence easily-learnt approach uncluttered by non-fundamental analytic categories [Paul, Fred, Martin] and undistorted by conventional architectural divisions of IT function [Jeff, Wolfgang, Ralph].
Resource contention, errors, failures and inconsistencies are part of complexity. The architecture fully recognizes them and provides for their integrated and resilient management [Wolfgang, Stefan, 6, 3(firstly), 10], even in a fully enduser-driven networked world with non-minimal clients [Jeff].
Describable enduser problems are generally best solved cooperatively, taking advantage of the open market. Analysts and designers will formulate their diverse insights directly in people-sharable and computer-interpretable form, in terms of MACK’s typologies with their well-defined coherences [7, Ralph, Paul, Haim].
MACK fully enables the infrastructure needed for the market-based cultivation of the total variety of product required to address the degrees of complexity that people can deal with [Paul, paper(Homeric mast)]. Metaset will take first advantage of that.
Plug-and-play component-based market-supported reconfigurability will be pervasive, thanks to the completely semantics-driven, apparently-intelligent and visually-controlled application specification approach. A fully repository-based A&D environment is indispensable here for actively-promoted flexibility with distributed autonomy [3(firstly)].
Thus a complete new MACK-compliant operating environment - or "Universal Front End" (see Mark Baker’s paper) - will be created collaboratively. "Profound congeniality" will supersede mere user-friendliness in the user interface [Jeff, William, 3].
Together, in a difficult yet supported and managed way [Ken, Paul, William], people will be creating their own new, appropriate, flexible, controllable and challenging yet comfortable reality. They will be further stimulated to observe patterns [Ralph, Martin, Haim] and enabled in their casting them in immediately useful forms.
All the above is leagues ahead of conventional architectures [Andrew, Mark, Haim, 2], and is feasible because of its epistemological appropriateness and hence simplicity [Ralph, 12, 3].
Strategy
I agree with at least the global requirements that the OMG and all of you have so well expressed. However, the detailed reality in the IT industry trails far behind [Jeff, Fred, Mark, Tom]. My goal is for MACK to move into that vacuum.
My public strategy right now is to locate partners for a major joint project. I’ll call it "Team Metaset" here.
If that turns out infeasible at this stage, my own programming of Metaset - which this Workshop involvement has pushed into the background - will be brought into my foreground very soon. The aim then would be to go straight for a more demonstrable Metaset product and thereby make Team Metaset more easily sellable [11]. As far as I can see right now, some relatively minor technical contributions from outside will be required at a later stage [9]. I have colleagues locally who are well placed for that role, and we would together raise our public profile as soon as it became feasible.
Sooner or later Team Metaset will get going, with partners who could set up the administrative, legal and commercial infrastructure for launching the Metaset-based market and managing eventual MUCK-cultivation [8].
Within such a framework it should also be possible to organize a safe and quickly effective technical boost as well, "to make it happen sooner and better" (as I said above).
I would also like to start that growth soon in order to prepare for my own phasing out from technical and operational responsibilities (my own long-term objective being merely to be a user of ACK-based products [11, Haim]).
Such partners could well need substantial incentives. Hence also my concern about trade secrets. Hence my demand [11] for substantial commitment before they gain access to them (which would also be my insurance policy against non-performance by partners). My position [3] is that the right partners should be able to see enough of my picture at its already published - though high - level of abstraction. If I am unduly optimistic in that belief, well, then my solo programming will continue until Metaset itself can be more convincing -- as it will certainly be to an ever-growing degree. (And in due course, with relatively little help from the market that will be the case at every level of user capability or receptivity [3].)
Team Metaset would also in due course liaise with standards bodies (I presently have the OMG in mind [1]) to promote MACK as a standard and help manage its evolution into an ACK [8] (or fully standard "Architecture for Common Knowledge")
On "trade secrets"
To risk rather belabouring the point, let me expand on why I am not setting out more details of MACK and Metaset just yet. If I do then - who knows? - maybe you will be able to reason me into a position that would be more open and friendly as well as more efficient!
Looking ahead
Notwithstanding all the help from many colleagues [3], I have been very much alone in this project So there have been many avoidable obstacles. (Hence what inefficient use of my own time and resources, and at what an opportunity-cost for us all! That is most presumptuous, but I have to presume the considerable value of the project...). That has given me much cause for reflection on how - in general - to promote major innovation in a complex world.
The result is an ever-stronger determination to try to steer the MACK-compliant market in an evolutionary direction that will more readily spark and feed major quantum-leap or paradigm-shift growth directions. (Another reason for "incentives" lurks here...)
That should be feasible, as the profound congeniality in the MACK-compliant PDA is intended to promote the marriage of challenge with comfort in a natural way [Jeff, Mark, 3(firstly), 5].
Of course such ambitions land one squarely in the difficult arenas of psychology, sociology and politics [3]. (If one doesn’t see that and embrace such complexity then one should be elsewhere...)
But I have had enough experience as an IT-empowered change agent - and of socio-political change in general - to have become persuaded that the clichéed "resistance to change" is an unhelpful characterization of our familiar problem. Complexity, confusion and overcomplication - the various aspects of Charybdis - more constructively explain it [2, etc]. In my experience (and I am sure in the experience of BPR in general too [Ken, Haim]) people readily welcome change when they are part of it and/or can be led to see it as a real simplification of their own situations.
So, looking further into the future, I take it as both highly appropriate and quite feasible to try to foster - with the help of the MACK-based market medium - the congenial integration of complex, confusing and confused democratic demand with simplifying and hence dynamic yet attentive supply.
My references to the "learning organization" [3(firstly)] and education in general [3(thirdly), William] are also relevant here.
All that applies to radical change in general. But my observations on change also apply in the short term, and specifically to the current situation with Metaset/MACK:
If at this stage I have merely confused you with all my verbiage, have patience and Metaset will make it easier in many different ways.
On the other hand, I am sure - as this is The Mainstream, despite all appearances - that there are people out there who can nonetheless take this degree of abstraction and complexity, who have enough of a sense of challenge, and who are in positions from which they can help make it happen sooner and better!
Is that you just described? (Or a close colleague?)
My 1986 book, Beyond Apartheid [2, 3], opened with a quote from Karl Popper (p.4 of Vol. 1 of The Open Society and its Enemies, Routledge and Kegan Paul, 1943):
"We may become the makers of our fate when we have ceased to pose as its prophets."
My book was wrong enough for the experience to have taught me much, but on the whole right enough still to be highly relevant. We in IT increasingly have the opportunity - and the privilege - of being major shapers of our common future, and we will surely do it much better than we have up till now if - armed with an appropriate toolset and medium [Jeff(need for tools)] - we face the problems fully and turn them into opportunities, as set out under "The uniqueness of MACK" above.
Credibility
Such claims call for yet another set of comments (further to [12, 3, Andrew, Ralph]) on my own credibility!
Let’s recall some of the difficult technical fields encountered or raised by the Workshop papers and my comments on them: shared persistent object stores with federated control, accessibility through redundancy while preserving consistency, hot backups and recovery, cooperative editing, workflow, advanced transaction management, both undo’s backwards and what-if’s forwards, query facilities, temporal database, componentization, dynamic binding, version management, distributed link maintenance, garbage collection, easier error and fault management.
Anyone working in any one of those fields tends to have some pretty good ideas how many of its difficulties might be addressed within his or her own scheme of things, assuming interfaces and/or architectural tiers to look after all the others.
But I have argued throughout these comments that that assumption is no longer tenable. What about the time-management or control-design problems I have raised? And the evidence that has been mounting over the past five years that conventional OO cannot deliver on its early promise?
On the other hand, that is mere contradiction, and ignores Medawar’s Dictum as always! "Theories are not displaced by facts," so where is the "better theory" that will "replace" the earlier ones?
Metaset is coming, I insist, and will quickly lead to MACK as a strong and flexible basis for workable resolutions to all the above problems. But in its absence the question nags: Can there possibly be anyone anywhere who is fool enough to think that they can all be addressed with one simple architecture? Surely that would have to be the most monumental case of wishful thinking, total self-delusion and pure folly!
Well, the continuity in the story of this "fool" has not been a supply-drivenness by any one piece of technology (other than the very basics of universally-acceptable scientific method, tantamount to an inevitably-assumed "common sense" as a fundamental component of Common Knowledge in The Mainstream). Technology-derived "solutions" or mindsets are indeed often beguiling and most likely oversimplifying.
No, I have been unusually demand-driven. And it helps in such developments to be one’s own highly-demanding user, especially if that also soon leads to a democratic orientation as a fundamental component of a fuller stance towards maximal recognition of complexity. It also helps when, as tends to be true in a country such as South Africa, there other such people to work with (I am now thinking particularly of Guy Bullen-Smith and Stephen Davies [3(firstly)], and there are doubtless many others like them here who still have to be suitably empowered, as Metaset will do for us all).
I have been sharply spurred by enormous socio-political needs for at least 36 years (Guy and Stephen being somewhat younger), and guided for over 30 years by the methodological needs so well encapsulated by the allegory of Scylla and Charybdis [2,3]. That formulation of needs soon led me consciously into Information Technology as a surely significant tool to assist the cooperative resolution of social needs.
Accepting that "the price of freedom is eternal vigilance", it is reasonable to expect that such awareness of the Scylla syndrome might help protect one from dictatorial oversimplification by unsuitable conceptual structures.
Awareness of the Charybdis syndrome helps one escape both "analysis paralysis" and the mental grip by vested interests, whose partial mastery of complexity so easily blinds us to overcomplication by their merely artificial structures, and browbeats us into acquiescence within them.
Thus directed, emboldened and stabilized for several decades, and being continually forced by that Homeric discipline - so Mainstream after all! - to try to question and test all assumptions, including these very perspectives, we have for 25 years gradually assembled the technical bits and pieces that are increasingly rapidly confirming their own good foundation.