Separation And Grouping Are Fundamental Concepts

In a very long page, SeparationAndGroupingAreArchaicConceptsDiscussion, an attempt is made as to spell out the Archaic nature of Separation and Grouping. I inspected the page and found the following words used, and I believe all of which can be said to contribute to separation and grouping:

 names variables
 hidden visible
 static dynamic
 namespaces scopes 
 files sections paragraphs sentences words
 segments parts 
 contraptions schemes methods threads
 local global 
 searching sorting grouping summing reporting inspecting
 interpreting processing 
 measures techniques priorities
 call-stacks databases sets

If something is useful, usable and used, as all of the above are, I fail to see in what way the concepts behind them (that of Separation and Grouping) are anything else but fundamental and irreplaceable. If a concept is archaic, at least as I understand the meaning of the word archaic, it is in the process of deprecation and replacement. It seems to me a vital thing to be able to separate (as in wheat from chaff) and to group as to (put apples in a basket). How this is done may change, and the schemes, mechanisms and processes involved may be modernized to take advantage of new hardware and software, but the concepts in themselves are fundamental. --BlueHat

I suspect what we have here is a vocabulary dispute, not so much a technical dispute. The concept I called "archaic" was "single dimensional" partitioning, which is the norm in the industry right now (but seems to be hitting walls). Multi-dimensional separation is usually NOT called "separation" in my experience, or at least is often not sufficient. But I will agree that perhaps English is not sufficient to describe modern cyberspace, so our 3D-world vocab is tripping us up.

I have been working on the concept "that perhaps English is not sufficient to describe modern cyberspace" in my quest to establish a scheme for separation and grouping based on what I once called HyperArtifacts. I am only beginning to get a handle on how this might be done, but fundamental to it is the use of single and multi-word combinations as atomic elements in a Hypertextual language. It will handle objects, processes, states and persistance and as many other things as might be required. --DonaldNoyes

Perhaps a different title should have been used, but until somebody finds a significantly better replacement, wiki tradition is that "good enough" titles stand if they have been there a while. It's not practical to keep flipping titles for incremental improvements. Further, titles should NOT be required to carry all potential scope limiters and disclaimers in my opinion. Other's have disagreed, although their arguments are weak and border on being childish in my opinion. That being said, you are welcome to propose new titles. Hopefully, we can reach a consensus before any change is made.

[You choose a big title and attack such a small aspect of it. You sow much confusion by doing so, and you create confrontations regarding the title that are not beneficial towards your purpose. Even responding to pages such as this one consumes your time, your energy, to no purpose other than once again explaining that your title is bigger than your claims. It would be wise to narrow the scope of the title to reflect the narrow scope of your claims (if you need help, terminology that describes what you so despise includes "DominantConcern?" and "Tyranny of the Dominant Concern"). Since you don't like flipping titles, you would have been wise to apply greater thought when choosing the initial title. As far as "Multi-dimensional separation" being not called separation, you'll be disappointed to know that MDSOC (multi-dimensional SeparationOfConcerns) is a well known term among those who follow progress in ProgrammingLanguageTheory and have read about HyperSpaces.]

And I do indeed agree that *some* kind of "partitioning" is needed, and thus "some kind" of partitioning is "fundamental". It is absolute or single-dimensioned partition that is "archaic". The main point of the original topic was that we need relative partitioning and the current tools do not provide it well. If there's a better way to say it, that's good. {I added the blue-hat moniker. I hope you don't mind. Lack of handles can cause difficulties.} --top

Reference Pointers Versus Block Markers

Strangely enough I am wearing a navy-blue hat right now from Juneau Alaska. I don't mind at all. I have long been an advocate for PositiveDialogue. It is far more productive and satisfying than confrontation and dispute. About partioning and dimensioning - What I see as needed is something that points rather than something that counts, and inserts in slots, positions or spaces. A sort of connection of concerns that is associative, so to speak. Thus when connecting something as complicated as a sentence or paragraph (made up of words and complex-words, the pointer(s) may be multiple from the same or several points in the sentence or paragraph. Not the simple single, dual or triple type found in linked lists and the like. --DonaldNoyes

[Not archaic yet, at least not as most English speakers understand the term. Wait until WarAgainstDominantConcern? is finished before you declare victory.] {misplaced/drifted context?}

That is an interesting open question: references versus "tagged chunks". It almost reminds me of closures-versus-start-stops in other debates. Or how to represent text formatting (nesting versus starters/stoppers). Chunks just seems cleaner for this use, partly because the scope of the category markers is cleaner and partly because functions, modules, and blocks already exist as possible pre-made frames of reference. If we have scope starters and enders, they may drift apart or get deleted out of sync. But it's likely a case of WaterbedTheory at play. Perhaps a tool could support both, at the price of added complexity. --top

Open Questions:


View edit of April 12, 2013 or FindPage with title or text search