Ready.


copycat82, the un-credible Ph.D., in one of its absurd advertisements/comments claims an ability to "postpone the data-dependent specifications to lower levels of design."

With the everything-inside-out graphs of copycat82, neither data-sharedness, nor data-use could be postponed to lower levels. That is the way the plagiarism of copycat82, has lead its "single"-graph to be, which is only a full-dump of every linked item, as NN73, VD78, and SARA had kept separate, but copycat82 collapses altogether. It is only a cut-and-paste monster, and next, the claim about postponeability is absurd, with such a collapsed graph.


data-sharedness cannot be postponed

The data-sharedness cannot be postponed because if two lower-level macros/subnets, L1 and L2 refer to a data-item, the upper levels, which contain them, must show the sharedness, too. This is not postponeable, if L1 is within U1, and L2 is within U2. Does copycat82 confuse itself with E-nets, or SARA SL? Both of them well-isolate the data, from the control-hierarchy. The control-net structure does not refer to any data. E-nets keep the data-set within token-attributes. No redraw needs, if a few new attributes were added, or deleted, renamed, reshuffled, etc. Quite similarly, SARA SL associates data with each module-interface socket.

The problem of copycat82 is especially severe, because it is supposed to be the single-graph to contain all such control-to-data links. (Such a collapsed-graph leads to other awkwardness, too. e.g: lopsided exhaustive searches.)


data-use cannot be postponed

What would be a context of use, if data-arcs were shown at some levels, but not at the others? Is it for reachability-test? The reachability-test, as copycat82 adopts, does not refer to data, at all? Does copycat82 confuse itself with Da80? It is Da80, where the X-transition of E-nets is introduced to Petri nets, to compress the nets, and wraps data within (internals/procedures of) X-transitions.

Is it in a context of data-ful simulations? The original E-nets is an example of such. SARA SL, too. But when the simulation is runnable, the data must have been specified, at all levels. Otherwise, what would it mean to calculate, if part of the numbers "at some levels" cannot be calculated? No preferences made at the upper levels? For a preference, some statistics are needed, or else, it is random. Would that mean a non-deterministic model/verifier at the upper levels?


what sort of a "distributed system" to postpone data?

Does copycat82 guess that the upper-levels may only use F- and J-transitions? (SARA/UCLA refers to them as "and" at both input and output, and copycat82 adopts SARA-style names.) This is the only possibility to avod absurdity, but if so, what is the sense of a claim to extend Petri nets, if the so-called distributed-system would not share any data, between separate installations? It is the sort of absurdity, in an application-side.


could "postpone" mean "macro-expansion" ?!?

At upper levels, presumably, the "component"s may not contain any program-fragments themselves, but only contain nets. That is, macro-nets. But such a case, does need macro-expansion before the interpreter runs it. This is what NN73 suggests for macros. Is copycat82, with, or without macros? Does copycat82 mean to suggest such a necessary macro-expansion, at all? The so-called "component"s of copycat82 turn out to be ultimately vague, themselves, any way.

As such, it would only postpone the problem, until we discuss those vague "component"s, and a variety of complexities, with the non-expanded components, as copycat82 does appear to attempt to do reachability-test with - in absence of any intermediate step to manage the expansions of copycat82's vague macros, in general.


vague. vague. vague. absurd, too

copycat82 neither explains what it means, nor cites any source (such as NN73, when/if copycat82 means macros). Therefore, we may only guess. If "postpone data," means "macro-expansion"s, it is trivial, any way - and if copycat82 claims anything out of it, then it is plagiarism, too.



The rest of this page is the same content, with different words.





The UnPhD is ignorant about its graphs. The UnPhD (p.53) is still talking about "postponing the data-dependent specifications to lower levels of design." But how? With such an explicit representation of data, within the same graph, along with the control graph (Petri net), how can we not show the data-sharing relationships among different parts? Keep in mind that, we are talking about distributed systems, here, and not about totally-separate systems. i.e: Data-sharing (messaging and/or shared-memory) exists, by definition. If the lower levels of two higher/container subnets share the data, or message among themselves, the higher entities must also show the data as being shared between them. Otherwise, the path/bridge is broken/untraversable through the vertical and horizontal links. And also keep in mind that, the UnPhD neither has a repository run by a software system (like SARA), nor it lists its system in full/formal/explicit predicates (like E-nets). The UnPhD has neither of these, and when the graph does not show it either (because of torn apartness at gross levels), the information is lost in its representation.

And even in the case of VD78, where data-graph is separate, the two transitions/subnets must still be non-overlapping in execution (unless race-condition is acceptable). Doing this, at the higher levels (with some subnets), leads to a granularity of parallelism at the coarsest level (as opposed to finding the exact-points of use, and mutex'ing only those specific points).

That sounds as if commenting on Da80, the "internal event" because of the ability of In a data-only graph, where operators/predicates


a self-contradiction: How would it use two separate representations, within a single hierarchy?

copycat82 is self-contradictory. Is it a simulation (with data), or is it an exhaustive verifier (without data)? Only Da80 mixes the two.

The UnPhD's/copycat82's comment (p.53) about "postponed data-dependence to lower levels" could not make sense, if it is a precise simulation.

And vice versa, the data-ful lower-levels would not be relevant, if it is a verifier. How would they relate? It is Da80, there, too?


vague, vague, vague

The case about the "postponeable data-dependency," is yet another case of copycat82 vagueness/ignorance. That is, copycat82 is vague, and it probably is ignorant about the vagueness. Otherwise, such a vagueness, and faultfullness, would not be published that way.


1. data-names optional?

What does "data-dependent" mean? It is not possibly about the data-name rectangles, because if lower levels of two upper-level macros share a data-item, the upper-levels must show that data-sharedness, too.

If copycat82 meant this, it is obviously faultfull about it. In such a case, the nature of the fault suggests the copycat82-plagiarism as its cause. Macro E-nets, and SARA SL, when they build their macro-hierarchies, they isolate the data-relevance within the entry/exit of a path. E-net formalism refers to it as token-attributes, SARA refers to it as module-interface-sockets. As such, the culprit is the kindergarten-style cut-and-paste of copycat82, when it turns such isolated/abstracted data inside out, by a full dump of data-names, as a cluttered-graph. It is not only tedious to draw, it also destroys such abstractions well-kept by the other/real researchers. Next, if the comment about "postponed data" means to suggest not-to-draw those data-names, at upper levels, it commits yet another grave fault.


2.a. Does it mean "macro-hierarchy vs. control-nodes?"

2.a. Upper-levels are only macro-containers, without any program-fragment shown in them, because they only represent a net that MUST be expanded. i.e: It is not a single-node, but it is the macro-subnet case.

Da80 isolates/hides programs at the internal-level, but it is not vague macro-nets. Every element of Da80 is strictly an E-net X-transition. It is copycat82's confusion of nodes with macros - unless it expands them as macros, which copycat82 never discusses, when and how to expand.

2.b. Da80 summarizes with data. (cf. NN73, p.724, too) But whereas, the data is kept by Da80, copycat82 does not use data with the Petri net verifier.




Forum: . . (Fair Menu . . . . . Fault Report? . . . . . Remedy for your case . . . . . Noticed Plagiarism?)

Referring#: 0.0.1
Last-Revised (text) on Sept. 7, 2004 . . . that was http://www.geocities.com/ferzenr/decalun.absurd_to_postpone_iff_not_macro.htm
(except: revised/rewrote the sentence "...with, or without macros ..." on Nov. 29, 2004)
Refreshed presentation/links, on Nov. 29, 2004
mirror to mid80.net, on June 16, 2009
Written by: Ahmed Ferzan/Ferzen R Midyat-Zila (or, Earth)
Copyright (c) 2003, 2004, 2009 Ferzan Midyat. All rights reserved.
mirror