Ready.
This page tracks the plagiarism, and errors (not necessarily exhaustively), at the start of the second section of copycat83, the paper associated with copycat82-the-"PhD" which plagiarizes. The rest of the section, where it presents "itself" is in fact, E-nets, with a vague/camouflage presentation. Another page proves that copycat82 plagiarizes discusses those. A full list of errors, is yet to be published - through the page ...
Here, we track the paragraphs, in the section "Modified Petri Nets and Design Representation." This is the second section, in copycat83, after the introduction, and it is the third section/chapter, in copycat82.
The first paragraph (page 734) is about Petri nets. It ends by referring to "an example" in Fig.1 (drawn on page 735). A few observations about that (faulty) example:
It is not proper. It is not even conservative. i.e: It preserves neither any portion of the net, after execution, nor it preserves the tokens-amount. In, at most, a few turns, the whole net would be absolutely halted, especially with the 1-boundedness (safety) restriction.
By the very criteria, that copycat83 itself republishes, unsafety is taken as an "inconsistent or faulty design (on page 736; it is only an abridged version of the lines on page 188 of VD78). i.e: We ought to label that Fig.1 as a bad example.
That name is not defined anywhere in its text, and it appears in no other figure. But if, that example were not a bad example, it would be a VD78 subnet. There, only a name-difference stands-out.
To visualize, you may pick the VD78 paper. Look at a Petri net figure in it, and either rename "idle place" to "full place." Or otherwise, even renaming may not be necessary, because "idle place" is one of those places that keeps its token, after a subnet finishes (properly, as it is). i.e: In the latter case, you neglect any difference between the "idle place" (which represents the outside-world to the subnet), and the internal control-places.
Every figure in copycat83 is faulty - "except" Fig.4 that only imitates the shapes of UCLA/SARA i/o control logic. i.e: It is not a statement, any way. That figure shows how the signs are drawn.
See Pe77, for a more thorough literature overview (or, Pe81). To include a literature overview subsection is O.K., but it is not anything worth mentioning here - especially as the Pe77 paper already had existed.
Again, read Pe77, to learn about, a Petri net graph, with, versus without any tokens (i.e: whether a marking is associated).
The fourth paragraph of the section. See: VD78 intro. pages 180-181, and again, Pe77. In Pe77, especially the last paragraph of section 3 (on page 237).
The VD78 is the paper that studies top-down design, thoroughly. Although copycat82/83 is published later, it has only committed a lot of errors, when cutting and pasting ideas from both Macro E-nets and VD78. We discuss these mismatch problems thoroughly, at this site - with a conclusion that, although utilizing both Macro E-nets, and VD78, is rather straightforward, and would not require a PhD, at all, copycat82 is massively faulty even with such a task.
Both E-nets and VD78 were strong players in employing data, with Petri nets. They are both quite, if not exactly, perfect, in what they studied. We are now tracking, the fifth and sixth paragraphs of copycat83 (the first two paragraphs on page 735).
The point is that, copycat82/83 fails awfully about data-relevant issues - as I precisely show, on several pages, at this site. But in that sixth-paragraph, it writes, that E-nets "attempted to represent data flow by associating attributes with tokens." Such a comment is absurd, very especially as copycat82, in its very first paragraph, in introduction (page 733) expresses that it is for a "message-based" system. In other words, in most, if not every, figure, i.e., when it is not using "shared memory," all it does as its-absurd-"improvement"-attempt beyond E-nets is that, it draws n parallel paths between the sending and receiving nodes, and those paths each embed one of the names of the n token-attributes, associated with that token. Is that progress?
Even worse, it proposes doing the same with data-types. But next, it never does it, even itself - except in a single example figure, where a dummy type-name of "T" exists. In fact, even that is confusing, because there is an inbound data-item "X," and also an outbound data-item "Y." Which one is with a type of "T?" Both? The cause of increased confusion is that, because most types are not shown, at all, in any example (except this one, and sample-representation/slicing figures - not examples, all those types are "T," "T1," "T2" any way, and they are more probably sets-of-types, not types, everything being trivial, and still faulty.) Therefore, we cannot infer that both variables may have that type. One may have an unlisted type. Another source of confusion is that, although, data-item arcs are directional (paralleling the control-token), the data-type arcs are not. So, let alone sorting out among the inbound data-items (to data-types), even in- versus out-bound data-items must be sorted out.
Yet other problems, semantics anomalies exist, especially when copycat82/83 breaches the expectations about, even what a simple "message-based" system would do. Thereby, you have to infer, although copycat82/83 itself does not discuss anything about such, new semantics such as "similarity of data," or "virtual(ly shared) data," etc. Even such, after some point, gets into problems, because, after all, it has not kept with any particular semantics.
In short, I conclude that, Nutt has been successful in his "attempt," whereas it is not obvious whether copycat82/83 even attempted to do anything new and sound. As a result, I insist we should call E-nets (Nutt's PhD) an achievement, as we refer to successful "attempts."
The seventh paragraph, starts with telling its name changes. When it is from Petri nets, as you may find in Pe77, it refers to them as "corresponds" but when it is from E-nets, or VD78, we find copycat82/83 ignore prior art. e.g: What it refers-to, as a "data-transfer specification" is exactly the transition-procedure in E-nets (associated with a T-transition), and it is data-operator in VD78. (E-nets, and VD78 differ from each other, with what/how they study the net. Otherwise, the formal structures of both, are not so difficult, if at all, to map each other.) Neither one is cited as a prior art, although copycat82/83 presents its structure and behavior, as virtually identical to E-nets (if not exactly a subset, with no improvement), and it proposes analyzing it similar to the level-1 of VD78. i.e: Ought to cite, at least one of them, I think.
I give a few examples of errors/quirks here. The
question, to kep in mind, is that "What page, if any
at all, have the paper-reviewers found worth
publishing?" Every single page is impossibly full with
low-quality (let alone the plagiarism).
The major offense of it revolves around, collapsing
three very different (and well-studied, in the 1970s)
concepts into a single, and unmanaged/undealt-with,
vague, impossible-existence (without discussing
anything):
- macro-nets: e.g: NoNu73, and SARA SL
- proper reductions: VaDi78, (and the SARA
strong-reduction does this transparently)
- interpreted/mechanized/evaluated transitions/nodes:
e.g: E-net resolution/transition procedures,
SARA interpretation domain.
The collapse results in macros that are assumed as if
reductions, and/or as if mappable to ordinary Petri net
transitions. Otherwise, where are the list of rules
(except for the gate-macros - themselves with their
own errors/quirks).
Example 0:
----------
Fig.3 (on page 737), involves a loop around itself,
independent of the other inbound/outgoing arcs.
How is this supposed to be a Petri net transition,
if not expanded as a macro? This is a usual pattern
with it.
Example 1:
----------
Fig.1 "An example of Petri net" is a loop, but
it is not proper, not safe, not conservative, and
dead/deadlocked/dumped with either of the
interpretations of the vague 1-boundedness restriction
the paper/PhD itself declares/imposes.
i.e: Itself admits/accepts such not 1-bounded nets as
"design errors", but keeps committing.
Example 2:
----------
The "or-in" gate-macro on page 743, is not proper, and
unless tokens are deposited to both places at strict
sequence, it is unbounded, after blocked. In other
words, (almost) whenever this standard gate is used,
that subnet would be subject to get dumped by the
1-boundedness restriction, as a "design error."
The "priority-in" macros (potentially) loses priority,
after the first "priority"-side pass. What sort of
priority is that?
Example 3:
---------
Fig.2 "An example of modified Petri nets"
- The places p6 and p9 are totally useless.
- Except the Da80 X-transitions at the exit,
it is ordinary Petri nets, any way. In fact,
could have been E-nets X-transition, directly,
if those useless places were thrown away.
i.e: A (relatively subtle quirk because of)
Da80'ism, although neither the "me-too"-paper83,
nor the PhD cites Da80, at all.
- It is a loop. Yes. But C2 and C4 write-only, and
C5 and C6 read-only. i.e: No feedback. This does
entail that, unless totally at random, C2 and C4,
must preserve internal state. This, in fact, is the
pattern Da80 discusses,
In fact tell me any page, and/or figure (except fig.4,
which duplicates SARA shapes of i/o control logic),
and let me list the source and/or the
errors/quirks/absurdity.
An Empire, Unless You/We Notice
---------------------------------------------
.......
I have also designed protocols for discovering such
mediumic-attacks. i.e: To resolve/prove, who is the
victim, among the claimants.