Rule definition terminology

[bubblegum]

I think that article should continue to refer to B2a as a transition.

Why? Do you think that's particularly intuitive or readable, from the viewpoint of a LifeWiki reader / beginner?

I recognise that your position has your own logic behind it which is intuitive to you. However, I personally reject the implication that all or most beginning readers will follow this same logic and develop this same intuition - I didn't.

In short, I persist in my belief that there is no need to systematically rewrite LifeWiki to fit your logic.

I'm aware there are others who share this belief but have not spoken up for it, and I hope this post will inspire them to chip in two cents of their own. I wouldn't want LifeWiki discussion to degrade into a game of wearing out patience.

[confocaloid]

Linking a previous post: viewtopic.php?p=187227#p187227

It is misrepresentation of my position to write "your own logic", "intuitive to you", etc.
I don't mean "my intuition" or "my own logic". I did write "from the viewpoint of a LifeWiki reader / beginner", because I did mean common-sense intuitive understanding.

In multiple Life/CA-related (and other) sources that discuss topics where these ideas are involved, the same words and phrases are commonly used in an intuitively understandable sense. There are several possible states for a cell. There are several state-to-state transitions that can happen. There are rules that prescribe conditions that must be met for a transition to happen.
Same intuition doesn't imply same technical meaning. The ideas remain the same, even when written definitions vary between different sources.

In short, it is my belief that
(a) There is a common intuition (neither "mine" nor "your") behind the words and phrases 'transition rule', 'transition', 'condition', etc., etc.
(b) The local habit of misusing the word 'transition' to refer to either a condition or an "if-then"-style rule goes against that common intuition, and shouldn't be transferred to LifeWiki, because LifeWiki aims to collect existing Life/CA-related knowledge in a consistent readable self-explanatory way.

(edited to add several links)

[dvgrn]

Here are a couple of related LifeWiki edits that I made just now, having to do with transition tables in Golly rule definitions. One is a change to the "Transition (rule definition)" page, and the other is a related change to H. H. P. M. P. Cole's new "condition" page.

The term "transition table", referring to the part of a Golly rule table that lists transitions, has a sparse but long history of people using it with no apparent controversy or confusion. Sometimes "transition table" is also used to refer to the part of a programmable metacell that stores equivalent information about transitions. Here are some past uses from the forums:

Paul Tooke (2011),
me (2013) and 2014),
calcyman (2016 and 2017, among others),
AforAmpere (2017),
shouldsee (2018),
Naszvadi (2019).

Also, in answering this script question yesterday, I was looking through ReadRuleTable.py. There are several references in the comments to "transitions" in the rule-definition sense, referring to lines of a Golly rule file's transition table:

Code: Select all

            # assume line is a transition
...
            # retrieve the variables that repeat within the transition, these are 'bound'

There are a lot of similar uses of "transition" in Golly's native code -- including in error messages that users will see, and will need to be able to understand to fix their rule files:

"Error reading " << full_filename << ": one or more of n_states, neighborhood or symmetries missing\nbefore first transition"

This all adds up to another reason why it seems important for the LifeWiki to acknowledge this common, unambiguous, non-jargon technical meaning of the term "transition"-- along with the older LIFELINE definition, which is also a perfectly valid, unambiguous, non-jargon technical meaning. It's just that in practice, in the time since Golly transition tables were invented, "transition" no longer gets used very often in the old LIFELINE sense to mean just oldstate -> newstate.

It's much more likely that if someone sees "transition" somewhere, it will refer to a line of a transition table, or to something like the 'B3' or 'S4c' mentioned in the Condition article -- a specification of the current conditions of a cell and its neighbors, and the future state of the cell.

[confocaloid]

Here are a couple of related LifeWiki edits that I made just now, having to do with transition tables in Golly rule definitions. [...]

[...] The term "transition table", [...]

A transition table lists transition rules. This isn't specific to Golly, and this isn't specific to RuleLoader.

The part "may have originated partly" is misleading. Transition tables are irrelevant to the issue of distinguishing between transitions and rules. When the word 'transition' gets used in some local conversation to refer to a combination of the condition of a cell and the future state of that cell, the way to disambiguate that is to say clearly that it is a transition rule that prescribes a certain transition under a certain condition.

Your changes in affected wiki entries aren't currently clear or consistent on whether or not a transition rule involving variables counts as a single transition rule or as multiple transition rules. (I don't see why one would need to prohibit variables in a transition rule.)

Previously I already explained that the local habit of using the word 'transition' to mean 'transition rule' doesn't make this sense of the word into a term. Even if someone advertises it as so-called "terminology", it remains nonintuitive local jargon. Terminology used for explaining concepts to LifeWiki readers should be intuitive enough and self-explanatory enough, to be helpful to a reader who didn't yet learn the definitions.

[...] or to something like the 'B3' or 'S4c' mentioned in the Condition article -- a specification of the current conditions of a cell and its neighbors, and the future state of the cell.

A condition doesn't tell anything about the future state of the cell.

The entry "condition" aims to explain the idea of a condition. Conditions tell something about the current states of some cells.
The B3 condition tells "the cell is currently dead and has exactly three alive neighbours".
The S3 condition tells "the cell is currently alive and has exactly three alive neighbours".

For comparison, the A3 condition is completely identical to the B3 condition. Likewise, the D3 condition is completely identical to the S3 condition.

However, the B3 transition rule is different from the A3 transition rule: the former tells that the cell will be born after one tick if the B3/A3 condition is currently met for that cell, while the latter tells that the cell will remain dead if the A3/B3 condition is currently met for that cell.

[dvgrn]

However, the B3 transition rule is different from the A3 transition rule: the former tells that the cell will be born after one tick if the B3/A3 condition is currently met for that cell, while the latter tells that the cell will remain dead if the A3/B3 condition is currently met for that cell.

Looking at the bigger picture, do you really think that it's a good idea to suddenly start using this suggested term "transition rule", to mean the exact same concept that plain "transition" has been routinely used for, for the last decade and change?

I tried really hard to make use of "transition rule", at your suggestion, a while back -- was going to put my "green writeup" in under that name.

But then I read the attempts that H. H. P. M. P. Cole made to change various LifeWiki articles to actually use that term in practice. In my opinion it made those articles quite a bit harder to read and understand -- which is to say, more confusing. Some private feedback came in about those same changes, including

"can you please tell H. H. P. M. P. Cole to revert the creation of this page and all accompanying changes of 'transition' to 'condition' or 'transition rule'

it is like the worst name they could possibly have chosen"

-- Which is just one person's opinion, of course, and there's quite a lot of hyperbole there! But that's the kind of response that this entire terminology-replacement idea has been getting from quite a number of people -- which exactly why I made the executive decision on this topic in the first place:

Common usage calls them transitions. There's no need to change this.

You're trying to prevent confusion on the LifeWiki. I understand that. But the particular method that you're choosing to try to prevent confusion looks to me like it's inevitably going to create a whole lot more confusion than it solves. You're trying to draw a bright line between "transition rule" and "transition", even though "transition rule" has only been used to mean atomic transition table / rulestring elements about half the time.

For the most recent out of many examples: when a new Golly user sees the error message

"Error reading " << full_filename << ": one or more of n_states, neighborhood or symmetries missing\nbefore first transition"

and they go and look up "transition" on the LifeWiki to find out what it means, what do you think they should find there?

Right at the moment, what they're going to find is a "transition" page that redirects to a "transition (disambiguation)" page that links to "transition rule" ... and when the "transition rule" page gets around to mentioning "transition" in the sense that that Golly error message uses, it actually links confusingly back to "transition (disambiguation)", rather than linking to the relevant definition under "transition (rule definition)".

It's pretty clear to me that the current state of things on the LifeWiki is still quite messy and inconsistent in this area, though H. H. P. M. P. Cole certainly made a good start by adding the three new pages.

So I'm going to go ahead and patch things up so that the various uses of "transition" on the LifeWiki link to the intended technical definitions. For example, the first use of "transition" in the Hensel-notation article will link to "transition (rule definition)".

You keep insisting on calling this usage "jargon", but that doesn't make it so. There are a lot more proponents of that usage than there are people who object to it. It's a valid technical definition, and it's so enormously widely used on the LifeWiki and elsewhere that -- quite simply -- it absolutely doesn't work to try to remove or downplay that definition.

It would be great if you could see your way to either simply dropping this topic, or finding a way to make use of LW:DR dispute resolution rules to get another moderator or admin to overturn my existing executive decision. Just repeating your claims about "jargon" again won't do any good.

[confocaloid]

[...]
Looking at the bigger picture, do you really think that it's a good idea to suddenly start using this suggested term "transition rule", to mean the exact same concept that plain "transition" has been routinely used for, for the last decade and change?
[...]

Looking at the big issue at hand (which is how to explain the ideas to LifeWiki readers), I believe it would be much more productive to word the wiki entries to clearly say 'transition rule' whenever the intended meaning is 'transition rule', to clearly say 'condition' when the intended meaning is 'condition', and otherwise to clearly explain the ideas, without requiring unreasonable and unnecessary efforts from the readers.

Adding links doesn't solve that. Creating other pages doesn't solve that. A problem with counterintuitive or otherwise poor wording of a wiki entry can only be solved by rewording that wiki entry.

Your labelling this as "suggested term" is incorrect (as I already explained). The same terminology can be found in earlier CA-related sources and discussions. It is already in use.

[...]

[...] I just don't see how that particular term can work as a replacement for the existing uses of "transition" on the LifeWiki.

In that case, I suggest to consider the possibility that someone else can see something you don't see.

Self-explanatory intuitive terminology that was in use for several decades is better than some relatively recent vague local forum jargon.
The term 'transition rule' correctly tells to the reader that the underlying idea is a rule that prescribes something. It definitely can work on LifeWiki, and it can be helpful there.

In contrast, your preferred choice fails that criterion. Following the local habit of (mis)using the word 'transition' to refer to a transition rule is counterintuitive and unhelpful.

[dvgrn]

Looking at the big issue at hand (which is how to explain the ideas to LifeWiki readers), I believe it would be much more productive to word the wiki entries to clearly say 'transition rule' whenever the intended meaning is 'transition rule', to clearly say 'condition' when the intended meaning is 'condition', and otherwise to clearly explain the ideas, without requiring unreasonable and unnecessary efforts from the readers.

The "clearly" part of this sounds very good in theory, but it's hard to implement in practice.

Maybe it would be easier for people to decide on their own opinions on this, if we had actual side-by-side examples.

Can you maybe provide a version of the Hensel-notation article section -- substituting whatever terms you want to substitute for "transition", and/or whatever else you think needs to be done to make that article clearly explain those ideas?

Then someone could, for example, set up a poll and see how many people vote for the new wording, and how many people vote for the existing wording.

[confocaloid]

Looking at the big issue at hand (which is how to explain the ideas to LifeWiki readers), I believe it would be much more productive to word the wiki entries to clearly say 'transition rule' whenever the intended meaning is 'transition rule', to clearly say 'condition' when the intended meaning is 'condition', and otherwise to clearly explain the ideas, without requiring unreasonable and unnecessary efforts from the readers.

The "clearly" part of this sounds very good in theory, but it's hard to implement in practice.

Maybe it would be easier for people to decide on their own opinions on this, if we had actual side-by-side examples.

Can you maybe provide a version of the Hensel-notation article section -- substituting whatever terms you want to substitute for "transition", and/or whatever else you think needs to be done to make that article clearly explain those ideas? [...]

I meant the general preference for clear, self-explanatory, intuitively understandable wording. Unnecessarily perplexing wording shouldn't go to LifeWiki, no matter how common it is in conversations between currently active members.

If I had to fix the problems that are visible in the section Hensel notation, then I would likely do a major rewrite, removing and/or rewriting and moving into another place parts of content added in the edits https://conwaylife.com/w/index.php?diff ... did=132028
I do not provide any "version" or "draft" here. Such a rewrite would be a separate issue, offtopic in this thread.

[...] Then someone could, for example, set up a poll [...]

A poll cannot solve this issue. (Same for any other issue that requires discussion.) Instead of waiting and letting actual discussion happen, considering arguments of each side and finding some way to take them into account and solve existing issues, you are basically suggesting to avoid dealing with the arguments of each side, and to leave the issues unresolved forever.

Polls cannot replace the discussion where people explain their positions/opinions.
Polls can misrepresent the issues and positions of sides.
Polls can undermine possibilities for further constructive discussion.

[...]
Polls, unless they're handled carefully, definitely have a tendency to reduce an issue to a supermajority vote, at best -- or just a regular majority vote, more likely -- or some kind of unpleasant popularity contest, at worst.

Generally none of those options is really anybody's favorite way to resolve LifeWiki issues. A gradually formed consensus is better -- in the rare cases where we can reliably figure out what the consensus is, given that usually very few people want to bother speaking up and say anything about any given issue!
[...]
If there turns out to be a general consensus that

• a poll should be created, and
• the article should then get named whatever the poll says to name it

-- then I'll be very happy that a non-controversial method has been found to resolve the issue.

However, I'm thinking if even just a couple of of people object to the creation of the poll, it's probably not worth bothering with it.

I think "usually very few people want to bother speaking up and say anything about any given issue", because the few active people usually quickly notice that some given issue became unnecessarily controversial (not because of some tricky details of the issue itself, but instead because of some active people not really wanting to discuss the issue with other active people), and because people who could get involved and provide some helpful feedback usually don't want to get involved in such an unnecessarily controversial issue.

[confocaloid]

How many 3D INT/Outer-totalistic transitions are there in a 3D range-1 cube neighbourhood?

How many 3D INT/Outer-totalistic transitions are there in a 3D range-1 cube neighbourhood?

What exactly you are trying to count?

How many 3D INT/Outer-totalistic transitions are there in a 3D range-1 cube neighbourhood?

confocaloid's previous summary gives some good information:

27 options for 3D range-1 outer-totalistic neighbor counts (0 through 26 neighbors)
1,426,144 distinct configurations for 3D range-1 INT, according to this complicated-looking program.
[...]

One obvious problem here is that they didn't really ask for "outer-totalistic neighbor counts", or for "distinct configurations for 3D range-1 INT".

The asked question apparently asks about the number of "3D INT/Outer-totalistic transitions [...] in a 3D range-1 cube neighbourhood", without bothering to explain the intended meaning of that.

The part "transitions in a neighbourhood" is meaningless gobbledygook. A transition specifies the future state of a certain cell (and the current state of the same cell). However, the current neighbourhood condition is just [some, possibly incomplete, information about] current states of some cells.

It is unclear whether the part "INT/Outer-totalistic" is supposed to mean two separate questions combined into one, or some way of combining two different ideas into one.

As a consequence, the asked question reads like meaningless gobbledygook, to someone who is reasonably familiar with relevant unnecessary counterintuitive local forums jargon.

There are further issues that could be probably discussed once the above issues are resolved / disambiguated, before it would be possible to answer the resulting question(s).

[dvgrn]

One obvious problem here is that they didn't really ask for "outer-totalistic neighbor counts", or for "distinct configurations for 3D range-1 INT".

The asked question apparently asks about the number of "3D INT/Outer-totalistic transitions [...] in a 3D range-1 cube neighbourhood", without bothering to explain the intended meaning of that.

There are a large number of people who read these forums, for whom Haycat2009's question is a perfectly clear question with a well-defined answer. I gave that answer in the Basic Questions thread.

@confocaloid, I've moved your response here -- because there don't seem to be many people who read these forums, who want to see the "transition" argument get duplicated in every thread where someone uses the word in a way that you don't agree with.

[confocaloid]

[...] I gave that answer in the Basic Questions thread. [...]

There are several different possible answers (different numbers), depending on the interpretation.
There are also several other possible answers that are not single numbers.

It's a bad idea to do anything that would amount to encouraging ambiguity / obscure wording. The asked question provides an example of obscure ambiguous wording.

[hotdogPi]

"INT/Outer-totalistic" is still unclear whether Haycat2009 is looking for two separate numbers or not.

[dvgrn]

"INT/Outer-totalistic" is still unclear whether Haycat2009 is looking for two separate numbers or not.

I gave two separate numbers, since that seemed to me like the obvious interpretation of the question.

If someone can think of a reasonable interpretation of Haycat2009's question that has an answer that is a single number, please go ahead and give that answer (and explain what it is).

[confocaloid]

"INT/Outer-totalistic" is still unclear whether Haycat2009 is looking for two separate numbers or not.

I gave two separate numbers, since that seemed to me like the obvious interpretation of the question.

If someone can think of a reasonable interpretation of Haycat2009's question that has an answer that is a single number, please go ahead and give that answer (and explain what it is).

There are multiple possible interpretations.

Is the intended meaning "conditions" (just some information about current states of some cells) or "rules" (including the future state of the middle cell of the 3-by-3-by-3 cube)?

Should the current state of the middle cell be taken into account in some way, or ignored?

What exactly "3D INT/Outer-totalistic" means? Is it two different questions abbreviated into one? Or should the answers be added, or combined in some other way? Or maybe that means some other rulespace that's neither full R1 3D isotropic nor R1 3D outer-totalistic?

As I already wrote, it feels like a particularly bad idea to do anything that would encourage that kind of obscure ambiguous questions. If the asker didn't care enough to write a clear readable question, that means they didn't do their homework.

[confocaloid]

[...] I add (yes: add) B5y, [...]

"Add" (or "remove") implies viewing the definition of a CA as a set of birth/survival conditions, with understanding that the omitted conditions aren't "living conditions".

If one is talking about "if-then"-type rules instead, then neither "add" nor "remove" gives the correct intuitive picture. There is always some rule that is followed when a certain condition is met. It is impossible to add or remove a rule, but only to replace one rule with another rule that is followed in the same case.
It is impossible to "add" or "remove" the rule B5y; it is possible to replace the rule A5y by B5y or vice-versa.

This is one reason why the quoted part below is incorrect (the other being incorrect use of the term 'transition').

[...] adding transitions [...]

[dvgrn]

[...] I add (yes: add) B5y, [...]

"Add" (or "remove") implies viewing the definition of a CA as a set of birth/survival conditions, with understanding that the omitted conditions aren't "living conditions".

If one is talking about "if-then"-type rules instead, then neither "add" nor "remove" gives the correct intuitive picture. There is always some rule that is followed when a certain condition is met. It is impossible to add or remove a rule, but only to replace one rule with another rule that is followed in the same case.
It is impossible to "add" or "remove" the rule B5y; it is possible to replace the rule A5y by B5y or vice-versa.

This is one reason why the quoted part below is incorrect (the other being incorrect use of the term 'transition').

[...] adding transitions [...]

@confocaloid, I've moved this post over here, since it seems likely to be unwelcome in the "pass the rule" thread.

You've continued to try to correct people for using "transition" in the way that it's commonly used on the forums -- and you have also continued to refer to things like "B5y" as a "rule" (where "rule" is a short form of "transition rule") instead of a "transition".

Nobody else is currently using that terminology on the forums. Your posts on this subject are continuing to create confusion and arguments where none would otherwise exist.

Saying just "rule" when you mean "transition rule" is really, really confusing. When everyone else says "rule" here on the forums, they generally mean a complete CA rule specifiable by a rulestring, like "B2i3-ckr4acew5ac6ce/S23-a4cqrz6i" -- not just one isotropic bit like "B5y".

Your "correct intuitive picture" is one that is shared by very few other people. In other words, the intuitive picture that you've been insisting on is not actually intuitive to the people you're trying to convince. Your attempts to offer helpful corrections are not helping anyone; no one is speaking up to agree with your corrections, and many people by this point have told you that they don't agree with you on these points.

Nobody else on the forums expects to see the word "rule", as a short form for "transition rule", referring to something like "B5y" -- it's simply not a workable substitution. Please stop trying to use "rule" in this way in future forum posts.

[confocaloid]

[...] Please stop trying to use "rule" in this way in future forum posts.

Of course no, I'm not going to stop using terminology that is used in existing Life/CA-related sources where the author actually aims to provide a clear, coherent, readable explanation of the topic. I already posted links and quotes from those sources.

@dvgrn
Please stop harassing me for my attempts to improve the situation.

As far as I can tell, your experience with Life Lexicon makes it practically impossible for you to understand that Life/CA is much more about understanding than about collecting common words and phrases.

[Sokwe]

@dvgrn
Please stop harassing me for my attempts to improve the situation.

As far as I can tell, your experience with Life Lexicon makes it practically impossible for you to understand that Life/CA is much more about understanding than about collecting common words and phrases.

Dave is not harassing you. He is telling you, as a moderator, to stop making certain comments that are leading to unnecessary conflict. I will second Dave in this case. Do not correct people on terminology in the sandbox, even if you think they're objectively wrong. It's not worth the conflict that seems to arise. This is me speaking as a moderator, seconding another moderator. Continuing to engage in such discussions could result in a temporary ban without further warning.

Furthermore, the comments quoted above are violations of rule 1. It should be clear that saying that it's "practically impossible for [someone] to understand" something is disrespectful and unkind. Do not make further comments like this, nor any other violations of rule 1. Such violations may result in a temporary ban without further warning.