Ready.

Page 35. This page is in the page-by-page raw-dumps section.



((The copycat82 refers to E-nets, all-and-only here on p.35, at a literature-overview context. It never again cites it anywhere when publishing some feature, although copycat82 is almost, if not exactly, like Macro E-nets - save the errors in copycat82. And even within those few lines on p.35, it tells of E-nets, in a faulty way.))

The Origin of the Data Places: copycat82 tells of E-nets, as if both of the places/locations of E-nets, are corresponding to the Petri net places. But in fact, the second sort of place, the resolution location of E-nets, is for the E-net interpreter to manage token-flow, deterministically, by representing data-dependencies, too, as evaluated by resolution procedure.

Knowing need not mean taking: copycat82 suggests an firing of an E-net primitive (transition) "takes time." But it need not. It may take zero-time, either - if specified like that. In fact, how about Petri nets? Although the time is not managed deterministically/explicitly, after all, it is there. i.e: Petri nets, with a non-deterministic reachability-test, "take time," too, but we simply cannot tell the elapsed time, in precision. Only qualitatively, we may infer it; not quantitatively.

Not to mention that, if there is no time, how would you schedule events, in a (deterministic) simulation? Specifying data "to provide determinism" but avoiding time, suggests an assumption of one-to-one correspondence of the time-elapsed with the simulated code-fragments, to the time-elapsed with the real-life system (software, etc.) Otherwise, the time-taken in real life, must be captured/modeled as the "time-taken" setting for that transition - if time is relevant at all, and if not everything is managed with (J-transition) syncronizations only.

On other pages, we will discuss that, in fact, copycat82 does make assumptions about timings, and commits to self-contradictions. It proposes immediately-firing, and token-removing entrance-macros, but at another page, expresses an intention to use "time Petri nets," later. But the "time Petri nets" is different in start-timing. If you specify it to be "zero," what is left? The latter is yet another Da80'ism (although Da80 is never cited in copycat82), but Da80 is not self-contradictory, because it expressly leaves E-nets notion of time, in favor of time-Petri-nets.

On p.724 of NN73, also notice the option of representing at a gross level, the transition procedure involving statistical terms. This suggests the time-taken to be modeled in statistical terms, too, as it is all too common in simulation-modeling, any way. In basic Petri nets, by contrast, time is always non-deterministic. An enabled transition may or may not fire, and even if it fires, the time it takes before firing, is arbitrary. The copycat82, with assumptions spread around its text, makes itself more like the E-nets, but provides no proofs that such a necessarily-incomplete translation effort (i.e: assuming Petri nets, with E-nets behavior), would still be verifiable by a Petri net verifier.





Further Reading

(page-by-page raw-dumps section) . . . . . previous . . . . . next




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

Referring#: 0.0.1
Last-Revised (text) on June 10, 2004 . . . that was http://www.geocities.com/ferzenr/decalun.pg035.htm
Links-update: June 18, 2004
mirror to mid80.net, on June 18, 2009
Written by: Ahmed Ferzan/Ferzen R Midyat-Zila (or, Earth)
Copyright (c) 2004, 2009 Ferzan Midyat. All rights reserved.
mirror