Talk:Entropy (disambiguation)

From Wikipedia, the free encyclopedia

WikiProject Disambiguation This page is part of WikiProject Disambiguation, an attempt to structure and organize all disambiguation pages on Wikipedia. If you wish to help, you can edit the page attached to this talk page, or visit the project page, where you can join the project and/or contribute to the discussion.
WikiProject Systems
Systems science Portal
This article is within the scope of WikiProject Systems, which collaborates on articles related to Systems science.
Systems rating: Disambig Class High importance Field unassessed.
Please update this rating as the article progresses, or if the rating is inaccurate. Please also add comments to suggest improvements to the article.

Contents

[edit] Removed some items

I removed some items. The general idea is that an item should only be on a disambiguation page if a user is likely to expect it on looking up "entropy". Because of the length of this page, I applied these criteria for removing items, to ensure the page was no longer than it needed to be:

  • Remove if there is no reference to entropy in the destination page, or
  • Remove if it is a small part of a topic that it is reasonable to expect a user to begin or continue his search with the wider topic (for example, a character in a video game), or
  • Remove if there is no associated article on Wikipedia, or
  • Remove if it is not generally referred to as simply "entropy".

There is some judgement involved, but if necessary, the restoration of items should be done with care -- entries should be short and only there if there is substantial risk of confusion. See Wikipedia:Manual of Style (disambiguation pages) for further information. Neonumbers 09:35, 4 October 2006 (UTC)

Neonumbers' comment referred to this edit (Oct 2006) -- Jheald 19:44, 15 October 2007 (UTC)
Seconded. I'm willing to remove more entries from the current list, actually: this is a disambiguation page, not an exhaustive discussion of uses of the word. Chris Cunningham 14:46, 15 October 2007 (UTC)
I think the recent changes by Chris Cunningham ("Thumperward") are steps in the wrong direction, and I would prefer this revision. I have put up requests for comment at WikiProject Physics and WikiProject Maths. Jheald 19:40, 15 October 2007 (UTC)
Replying over there, though I don't feel that putting links up on WikiProject talk pages (which are necessarily biased towards their own domains) is a good way of resolving copy edit issues. Chris Cunningham 20:54, 15 October 2007 (UTC)

[edit] Discussion on WP:WPM

The following discussion has been copied from WP:WPM. --Salix alba (talk) 21:39, 15 October 2007 (UTC)

I'd be grateful if people could have a look at the Entropy (disambiguation) page, where I'm in a dispute.

It seems to me that that this previous version was a lot more helpful than the current edited one for users trying to find their way to the article they need.

In particular the older one

  • Helpfully grouped articles around the main two scientific meanings -- information entropy and thermodynamic entropy
  • Helpfully also contained links to the different Entropy categories

The latest edit -- removing Introduction to entropy from the page completely -- seems to me particularly user-unhelpful, verging on WP:POINT.

But I'd be grateful for third party input on this, as maybe I'm too close to what's been edited before. Jheald 16:23, 15 October 2007 (UTC)

Disambiguation pages exist to help people out if a link isn't precise enough, or if they search for too broad a term. They do not exist to provide overviews of topics, nor are they intended to be site maps for keywords (as the list of categories the previous revision included would seem to assume).
Introduction to entropy has its own problems. I removed it for the time being because if it's deserving of its own article at all it should be a sub-article of Entropy, as as such doesn't need its own entry on disambig (compare to, say, the number of links on the Macintosh presented on Mac). Chris Cunningham 21:01, 15 October 2007 (UTC)

Personally I can't see the problem with a link to Introduction to entropy, a reader who has say followed the disambig link from Information entropy may well want to go straight to an introductory article.

Whether the thermodynamic Entropy articles should be separated from information entropy ones, is a different question. I think its a little clearer if they are seperated.

Overall, both versions are fine, differences for the end user are a matter of one click or two. In any case a search for Entropy will take the user to Entropy and few pages link here. Not really worth an edit war. --Salix alba (talk) 21:56, 15 October 2007 (UTC)

Comments from WikiProject Physics talkpage:


I also like the old one better. I think Thumperward may be applying some of the conventions for simplicity in disambiguation pages too rigidly, and his argument that "'Useful' is not a criterion that dambig pages are judged" is not persuasive. Usefulness (for navigation) is the only reason for the existence of disambiguation pages! MOS:DP clearly says: "For every style suggestion above, there's some page which has a good reason to do something else. These guidelines are intended for consistency, but usefulness to the reader is the principal goal. So ignore these guidelines if you have a good reason." (Emphasis added.) However, as a compromise, since Thumperward doesn't seem to like the indented sublists for thermodynamics and information science, perhaps you could turn those two subareas into top-level lists, by replacing "In science and technology" as a heading with "In physical science" and "In information science", or something to that effect. --Itub 17:26, 15 October 2007 (UTC)
From the quick lookover I gave it, I agree with Itub. I'd also like to bring up WP:IGNORE. --Falcorian (talk) 19:19, 15 October 2007 (UTC)
WP:IGNORE is for cases where policy prevents good editing. In this case, I think policy is perfectly sensible, and using the page for something it isn't makes Wikipedia more difficult to use. Chris Cunningham 11:54, 18 October 2007 (UTC)
I don't appreciate the WP:POINT allegation; what "point" am I trying to put across? That disambiguation pages should not randomly deviate from convention? And the previous version ignored disambiguation page convention to such an extend that it was arguably trying to be a WP:SUMMARY rather than a disambig. Those pages do not exist to "be useful" in any form other than to point users to the page they were actually looking for in a more helpful form than a context-free list of search results. Especially notable were the category links; Wikipedia is not dmoz, and one does not search the article namespace to get category results. Chris Cunningham 20:56, 15 October 2007 (UTC)
Chris, the way you've edited this, by lumping together the articles variously related to thermodynamic entropy with those variously related to information entropy, removing the structured grouping that was there, actually makes it harder for people to find the page they were looking for. I fail to see how this is useful.
A page on entropy (disambiguation) ought to disambiguate some of the different ways, and specific contexts, in which the word entropy is used. That's how you help people find what WP has called the article on the subject they are looking for. If you look at those cats, you will find articles on a number of yet further particular entropies -- eg Conformational entropy, Loop entropy, Entropy of mixing, Tsallis entropy, etc -- that it is entirely appropriate for the dab to reference the categories to help readers to find. Jheald 21:31, 15 October 2007 (UTC)
No, this isn't the point of Wikipedia's disambiguation pages. They aren't meant to discuss the dictionary definition of the word; they're not meant to give indight into its usage; they exist solely as signposts for people who know where they're going but not how to get there.
I'm happy for this particular property to be improved. I'm not happy for this article to buck the trend (and ignore policy) for no better reason than the significant number of scientific concepts presented therein. Chris Cunningham 11:52, 18 October 2007 (UTC)
They exist as signposts for people who know where they're going but not how to get there. Exactly. Now, will you stop messing around with the page, making it harder for people to find the articles they want. Jheald 12:53, 18 October 2007 (UTC)
No, that's not how it works, Jheald; if I thought your version made it easier to do this, I wouldn't be arguing against it. Please continue to discuss this. Chris Cunningham 13:11, 18 October 2007 (UTC)
I found Jheald's interpretation of the usefulness of some additional struction to distinguish between thermodynamics and information entropy articles quickly and easily very persuasive after looking the two different pages. Further, I think the disambiguation pages must also serve the people who do not know where they are going but still wish to get there. Breaking out the sub categories thermodynamics and information is very useful to someone like me(with an engineering education rarely used professionally) to find material I am looking for related to thermodynamics ... I have encountered Shannon information theory version of entropy before and it was useless to me in the engineering context in which I was working. I would not want to find myself there while attempting to use Wikipedia as a quick reference to brush up on some dimly remembered concepts. The helpful sub structure makes it much less likely that I will be wasting my time while using Wikipedia as a quick reference on engineering concepts of entropy. Lazyquasar 07:00, 26 October 2007 (UTC)
This was settled amicably a week ago. The subcategory stucture has been kept, and indeed reinforced by sectioning. Chris Cunningham 07:39, 26 October 2007 (UTC)

[edit] Another view

In my view, somewhat ironically, removing Introduction to entropy from the dab page was actually the most sensible of the changes made! Introduction to entropy is linked, from Entropy, as an introductory article to the subject, and there is no place for such an article on a dab page. On the other hand, combining articles on thermodynamic entropy and information entropy into a single list is utterly confusing. The two meanings should be separated. Geometry guy 22:19, 15 October 2007 (UTC)

[edit] MOSDAB

After the recent revert of the "primary meaning" thing, I'm assuming that people need to go and re-read WP:MOSDAB before engaging in too much more editing here. There are guidelines for a reason, they shouldn't be WP:IGNOREd out of hand over aesthetic concerns. Chris Cunningham 13:14, 18 October 2007 (UTC)

It might help if you actually knew something about the subject. To state, point blank, that
Entropy, in science, refers to the thermodynamic homogeneity of a system, limiting the ability to extract mechanical work.
is not correct. Entropy in science refers both to thermodynamic entropy, and information entropy. Arguably, according to some, information entropy is the more fundamental concept, and thermodynamic entropy is an application. Others see more separation between the two.
The point remains, that both are referred to as entropy in science; it is useful to group the more specific articles around these two headings.
As for "stop reverting whole edits over single-line issues" -- I reverted your edit because I thought that none of the changes you were making were helpful.
  • That includes removal of the categories, to which users should be pointed, to find what name WP has given to further articles on particularly specialist kinds of entropies;
  • it includes the "Introduction to entropy" article;
  • and, particularly, it includes the parallel grouping of articles under "thermodynamic entropy" and "information entropy", which makes it easier for users to get to what they want.
There hasn't been a single respondent on this talk page who has supported what you are doing. Doesn't that suggest there might be other more useful things for you to be working on? Jheald 13:30, 18 October 2007 (UTC)
No, and I don't appreciate being browbeaten over this (or, of course, be be told I'm an ignoramus, but I'm aware that calling one's opponent a buffoon who knows nothing of the subject matter is the very first thing most editors do in content disputes). Entropy's status as the "primary meaning" is conferred upon it by being located at the root article. This is pretty clearly documented in the Manual of Style, which you haven't yet shown any intention of following. The category links are, as per my edit summaries, irrelevant to a disambiguation page, insofar as they are completely unrelated to many of the possible meanings of the term.
As for not "a single respondent" being supportive of these edits, the one person who did contribute an opinion on the introduction to entropy article was supportive of its removal. So watch what you're claiming, please.
Wikipedia has style guidelines for good reason. They should not be ignored without one. Chris Cunningham 13:43, 18 October 2007 (UTC)
The good reason is: "Will it help people find what they want?"
  • Clearly structuring the entries broadly under thermodynamic entropy and information entropy does help to help people find what they want.
  • Advising them that further, more specialist entropies may be found in the categories does help to help people find what they want.
  • As for Introduction to Entropy. Well, I think that may help to help people find what they want. But I respect that G-Guy thinks it's not absolutely necessary.
As for "primary meaning" being "conferred by being located at the root article" -- MOS presumes that an article will only be made root article if it is the overwhelming primary meaning. (Otherwise, if there are two sdtrong meanings, the dab page would normally be root). But in this case the presumption is incorrect. Thermodynamic entropy is the root page, but it is not the overwhelming primary meaning. Hence, why the structuring on this page is useful.
Guidelines are guidelines. But per WP policy, they are subsidiary to particular circumstances and common sense. WP is not a blind bureaucracy. Jheald 14:05, 18 October 2007 (UTC)
And I'm not suggesting they be followed blindly. But they should at least be followed consistently in individual cases. If Entropy has been chosen to point to thermodynamic entropy, the matching disambig page should reflect this. And when the MOS specifically says not to do something, as it does for WP:MOSDAB#Categories, we should pay attention to why this is the case and follow it.
So: points of contention right now are the use of a primary article, the inclusion of introduction to entropy, and the category links. I'm still in favour of the first and opposed to the other two. I can be swayed in favour of removing the first, in favour of linking to it indirectly via thermodynamic entropy, if that's going to end this dispute. I don't think either of the other two points are negotiable, as they provide negligible value for their MOS violation. Chris Cunningham 14:22, 18 October 2007 (UTC)
You're misreading WP:MOSDAB. It says don't categorise dab pages (because we don't want dab pages coming up in the category lists). It doesn't anything about pointing to categories, when appropriate. In this case, it is appropriate.
Introduction to Entropy is signposted fairly heavily at the top of the thermodynamic entropy article, so if you really insist, I'd be prepared not to push it. Though I still think it's useful on this page.
As for the organisation under thermodynamic entropy and information entropy, in an even handed way, I think that's transparently useful, and I'm glad if we can lay that point to rest. Jheald 14:44, 18 October 2007 (UTC)
Right, I've reorganised this again with sections. I've kept the cat links (though I still disagree with them) and made the appropriate edits to the other two. Comments? Chris Cunningham 14:59, 18 October 2007 (UTC)
Okay. One minor change: I've restored some indents. If we accept that its useful to group articles which broadly fall under Information Entropy, it's also useful to group the articles which broadly fall under thermodynamic entropy. Jheald 15:08, 18 October 2007 (UTC)
Cool, yeah; I should have caught that. Anyway, (mostly) best of both worlds now, so that was another happy and productive edit war. Chris Cunningham 15:13, 18 October 2007 (UTC)

[edit] Revision 9 Feb 2008, plus some questions

I've merged several sections today, for example, mathematics, computing and medicine all sound to me like "science and technology", and film, music and television, as far as I can see, are similar enough to be one. Too many sections and it takes too long to find the right one so sections become pointless; too few, too many articles per section and sections become pointless; I think this is a better balance. I would have used TOCright, but there are enough right-aligned templates on the page as it is, and a TOCleft would only just take up space.

I was going to remove several entries on the page, but I see this page has been the subject of a certain edit war which was ultimately amicably resolved. Therefore I'd like instead to pose some questions, lest I remove entries that were there for less obvious but nonetheless important reasons. The entries I refer to are those grouped beneath the thermodynamic entropy and information entropy links (but not those two links themselves).

The disambiguation guideline, as we know, discourages the inclusion of articles merely for the appearance of "entropy" in its name—the main criterion is risk of confusion. Can a reader, looking for the article (e.g.) Gibbs entropy, reasonably expect to get there directly by entering "entropy" and pressing Go? Is (e.g.) the Boltzmann entropy formula often referred to simply as, "entropy"? It is not the job of this page to educate about the different types/measures of entropy. It is not enough for someone to want to know the name given to a family of diversity measures generalising Shannon entropy; they must be able to expect to get straight to an in-depth article on the topic, by entering "entropy" as the page name.

Something to ponder over—I'm deliberately not editing the page (for this aspect of it) because I want to hear what people think: are all those entries things that, to us, are typically just "entropy"? Neonumbers (talk) 11:42, 8 February 2008 (UTC)

The question is: are there contexts in which the unqualified term "entropy" is used to mean one of the entries of this disambiguation page. If yes, those entries need to be kept, IMHO. I would not be surprised to find many introductions to statistical physics that give Gibbs entropy as "the" definition of entropy without further qualification of "entropy". Nor would I be surprised to find many (introductory) books on thermodynamics or gas dynamics that give the Boltzmann entropy formula as "the" definition of entropy without further qualification. The same again for statistics books and introductions to communications-electronics and Shannon entropy. I would prefer to keep at least these entries on the disambiguation page. — Tobias Bergemann (talk) 12:44, 8 February 2008 (UTC)
Okay—I'm no expert on thermodynamics or statistics, which is why I asked—your question seems to match mine so we're thinking on the same terms. If you don't mind, Tobias, could I ask for your opinion on the remaining five such articles? Thanks in advance. Neonumbers (talk) 03:05, 9 February 2008 (UTC)
The next question I want to ask, as a question of fact whose answer I don't know, is, would someone looking for (e.g.) the Boltzmann entropy formula (as opposed to a discussion on thermodynamic entropy in general) probably and reasonably just type in "entropy" expecting to find exactly that? I realise it's sort of directly consequential from your question, so I apologise if it seems like the same question twice (to me they're subtly different). Neonumbers (talk) 03:13, 9 February 2008 (UTC)
I don't have a strong opinion wrt the Boltzmann entropy formula which when I learned about it was simply refered to as "Boltzmannsches Prinzip" ("Boltzmann's principle"), "die Boltzmann-Gleichung" ("the Boltzmann equation", not to be confused with Boltzmann equation) or as "Boltzmann-Entropy" ("Boltzmann entropy").
The term "entropy" is heavily overloaded in physics and statistics, mostly for historical reasons I guess, and the somewhat artificial naming scheme article naming scheme reflects on this. For example, before reading about it on wikipedia I had never heard of Shannon entropy being refered to as information entropy. After some thinking I believe we should keep only those entries on the disambiguation page for which we can find references which really use the unqualified term in a meaning matching a wikipedia entry. Then again, I have never seen references on a disambiguation page. mdash;Tobias Bergemann (talk) 13:55, 10 February 2008 (UTC)

In their own context, each of the "science and technology" entries could be described in the body text of a paper in the relevant discipline simply as "the entropy". As there are substantial differences, it is appropriate to include them here, to make sure that users are getting to where they want to get to.

Note also in passing, that unlike the majority of dab pages, users do not come to this page directly on typing in "Entropy", but rather, on following the hatnote on the Entropy page, which says for other uses of the term Entropy see this page. To make good on that claim, arguably this page should tend to a more complete, not a less complete, list of things which get called Entropies. Jheald (talk) 21:43, 10 February 2008 (UTC)

Thanks for your comments, both of you, I appreciate them. I concur with your line of thinking, Tobias.
You've raised a good point, Jheald. The fact that Entropy is a discussion of thermodynamic entropy means that all the user will see many of the thermodynamic definitions on that page, making their inclusion on this page unnecessary. Anyone after any of the thermodynamic entropy topics would most likely stop on Entropy and look through that to find what they want.
The "other uses" template refers to things that are in entirely different contexts. I believe most people would interpret it that way, as opposed to "uses that are just sub-classes of this topic".
I ask to clarify if a formula (as opposed to a concept) can really be referred to as "entropy", as in, "Entropy says that S = k ln W". Before anyone interprets this as a challenge, please remember that I'm not an expert on thermodynamics or statistics (though I have studied really basic thermodynamics) so I would appreciate enlightenment rather than a reprimand.
That said, inclusion of entries should really be discussed from a layman's view. I don't think using the unqualified term in an extremely specialised paper counts. If by any chance there's a note on the paper that implies "all references to 'entropy' in this paper are to the Gibbs entropy", that definitely doesn't count. I'm not saying anything falls into this category, I simply don't know if that's the case, I'm just clarifying what I meant.
I made an edit today and said there was an important assumption I would explain here—I assumed that there is no usage of the word "entropy" that is more common than any other. I have a feeling my assumption is incorrect, and I want to explain why this is so pivotal. As a principle (going by the manual of style), more likely targets (common usage) should be first on the page; if they are that common (I'm thinking the main thermodynamic and information articles) then they should be before the first section header. If this is the case, then my edit needs to be rethought and the rest of the page would likely fall into line much easier. Neonumbers (talk) 10:02, 12 February 2008 (UTC)
Undone. It makes sense to group thermodynamic entropy and information entropy together, because there are close (though sometimes controversial) links between the two. As opposed to just the advertising name for a bit of medical kit. Entropy really does divide into these two deeply related meaning-groups, which are the overwhelmigly important subject areas, versus the rest. (btw: apologies for the broken edit summary on the main page, I mishit return before it was complete).
You say these aren't subjects you know a lot about. Don't you think then you might give a little more trust to the judgement of those who maybe do? IMO, the current list is the most useful selection of articles to offer people that don't want a basic introduction to thermodynamic entropy.
Are there particular usability/usefulness concerns you have with the page, ie things that you think make it *specifically* less usable or less useful than it could be? Or ask yourself, are you just busybodying ? Jheald (talk) 10:49, 12 February 2008 (UTC)
I'm going to ignore that (not so) subtle attack on my integrity and good faith and just proceed to answer your question.
Several particular concerns:
  • Having too many sections, as I've already explained, defeats the purpose of having them. I explained that the week before last, in the subsection "Section mergers" below. Sections with only one link in them are wasting vertical space, and in a disambiguation page, vertical space is precious.
  • Including unnecessary entries was (past tense) a concern. Normally when I deal with disambiguation pages, I just go ahead and read the article, and then remove anything I don't think needs to be there. I might not be an expert on entropy, but I've worked with disambiguation pages a lot and I know how dab pages work, and I was hoping to work collaboratively. I also know that, before the disambiguation wikiproject started, entries tended to be grossly overincluded (and they often still are, due to common misunderstanding of the criterion for inclusion). This wasn't a distrust on your judgement, and I'm sorry if that's how it seemed; rather, I was asking for your judgement, because I thought it would be more trustworthy than making inferences based on the articles.
  • It seems counter-intuitive to me to have a "science and technology" section, a "mathematics" section, a "medicine" section and a "computing" section, given that mathematics, medicine and computing are all subfields of science and technology. To this effect, and in acknowledging that information and thermodynamic entropy should be placed close to each other (as you said), and that they are probably (correct me if I'm wrong) jointly the most common intended meanings, I would like to move the thermodynamic and information entropy entries before the table of contents, and then move on to make minor rearrangements, including merging maths and med into "science". (If they're not the most common intended or "primary" meanings, then this won't work.)
No, please don't make this change. The whole reason the ToC was introduced was to let people who knew they didn't want a thermodynamic or information theory meaning jump past these. That goes for the hip-hop record. It also goes for medical meaning. Please leave this out of the sections relating to the core thermodynamic / information theory meanings. Jheald (talk) 13:14, 17 February 2008 (UTC)
The dab page should be aimed primarily at those who typed in "entropy", not expecting to arrive at Entropy (as opposed to any exploration purpose, something specified in this section of the MoS). Neonumbers (talk) 09:52, 17 February 2008 (UTC)
Per WP:IAR, what matters is less what a rule says, but more why it says it - and how the overall balance of usefulness and usability works out here. Jheald (talk) 13:14, 17 February 2008 (UTC)
Then rename the "science and technology" section "thermodynamic and information entropy"? The science and tech, med, maths and computing sections all co-existing is still counter-intuitive, even if my proposed solution was a bad idea. Rather than just shooting my proposals down, I'd appreciate it if you'd try to understand why I believe the page could be improved.
I'm very aware of the flexibility surrounding manuals of style, and I don't see why this should be an exception to the principle—not rule, but principle—that disambiguation pages are for navigation not exploration. Anyway, in its current form, I don't see this as an issue for this page.
I would, btw, like an apology for the good-faith attack; any other user could accuse you of trying to "own" this page, but I know that's not true, I can see your contributions have good faith behind them. Neonumbers (talk) 06:41, 21 February 2008 (UTC)
Sorry if my style was brusque. I never meant to impugn your good faith; I just thought some of your suggestions were misguided. I can see your point that the label "science and technology" could be construed as an umbrella that might usually include medicine, mathematics and computing. A alternative way forward might to to move medicine, mathematics and computing to subheadings under "Science and technology". I'd quite like to keep "science and technology", because it's everyday language. Jheald (talk) 12:27, 21 February 2008 (UTC)

[edit] Section mergers

Leaving your question aside, but talking about the specific issue of the section mergers, I would like to back out your changes.

These sections were specifically introduced by Thumperward in order that there should be a Table of Contents at the top of the article that would make it easier to jump to meanings of entropy other than those directly associated with thermodynamic entropy. Specifically, the ToC at the top of the article made it easier to find Entropy (computing), which he had recently edited, and which he felt was 'buried' in a list the reader would have to plough through line by line. Hence the careful sectioning. Similarly, for somebody wanting to jump straight to (say) the Hip Hop album.

Unfortunately your change has had the effect of removing the ToC altogether, making this no longer possible; and the merged sections are less identificatory, making them in any case less useful.

I therefore think your change has made this page less effective; so, unless people weigh in to the contrary, I would propose to back your changes out. Jheald (talk) 20:00, 8 February 2008 (UTC)

If you wish to force a ToC (I believe there are several techniques for doing this), I will have no objection—in fact, I think that is a valid point; I will do that now. ToCright is generally preferred on dab pages but in this case, there are so many right-aligned templates that ToCleft might work better. Also, if you wish to split the science and technology section into two, so there are four sections (I concede it's a bit long), I will have no objection. However, I must protest to having eight sections. Too many sections (yes, even five) means a long ToC to sift through, and if it's counterbalanced by there only being one (or two) articles per section, isn't it just more efficient to have one list? The correct section of a group of articles must be identifiable at a glance (not a read). Neonumbers (talk) 03:00, 9 February 2008 (UTC)

[edit] Cleanup

There seems to be some issue as to whether "Entropy is a measure of the unavailability of a system to do work." is the proper introduction to this disambiguation page (not article, just dab page). That should be addressed by one of two ways:

  1. Change the intro on the Entropy page if the definition is wrong.
  2. Move the Entropy page to another title (like Thermodynamic entropy or whatever) and move the dab page to the base name if there is no primary topic.

Barring either of those (which will probably require reaching consensus at Talk:Entropy first, or at least following WP:RM), the dab page should note the current state of the Wikipedia article space. -- JHunterJ (talk) 19:02, 22 March 2008 (UTC)

I'll have to agree with J on this. The way it was before did not deduce to WP:MOSDAB format. Lord Sesshomaru (talk • edits) 19:07, 22 March 2008 (UTC)
Per WP:IAR, the WP:MOSDAB format has a vote, but not a veto. What matters is how to help people get to the article they want most easily. Per extensive disussion on this page, structured organisation grouping together closely-related senses is much more helpful than simply lumping everything together in an alphabetical list.
This page should reflect reality, which is that both Information Entropy and Thermodynamic Entropy are primary meanings of the word Entropy. If people think the article Entropy should be moved, I suggest they raise it at Talk:Entropy. Jheald (talk) 23:08, 23 March 2008 (UTC)
Good point; since you think that Entropy doesn't reflect reality as you just described, that would be the path to take. Please raise the issue there. -- JHunterJ (talk) 23:12, 23 March 2008 (UTC)
You want the change, you do the legwork. Myself, I'm quite happy with things as they were. This page should reflect the reality, which is that *both* are primary meanings. Whether or not the Thermodynamic Entropy article is called that, or just Entropy, is entirely immaterial. I am quite happy for it to be called Entropy. But this page, which navigates our articles, should reflect the two fundamental primarily important groupings that there are. Jheald (talk) 23:34, 23 March 2008 (UTC)
I don't want the change -- the primary usage at Entropy and its definition suit me just fine. You disagree with them, but are trying to circumvent the legwork by changing them here rather than on their pages. -- JHunterJ (talk) 21:34, 24 March 2008 (UTC)

I came here to give an independent opinion, but I don't know exactly what the disagreement is. I see two things: the introductory text, which I don't think MOSDAB encourages, and the section titles, which are fine per MOSDAB. Could someone (everyone is OK) give a one-sentence summary of the situation? — Carl (CBM · talk) 23:42, 23 March 2008 (UTC)

Carl, My preferred version would be this one, which groups together the meanings which most closely relate to the two primary meanings of the term (Entropy in thermodynamics; and Entropy in information theory); and places these groups before other meanings.
JHunterJ and Abtact prefer this version, where entries relating to these different main groupings are largely muddled together; and still further even less related meanings, like Entropy (anesthesiology), Entropy (anonymous data store), and Entropy (package manager) are chucked into the mix to make it even harder to navigate.
I submit that the original structure makes it *much* easier for people who know what they want to quickly find the article they are looking for; and that this is therefore a case where WP:IAR should be applied in preference to WP:MOSDAB. -- Jheald (talk) 00:00, 24 March 2008 (UTC)
JHunterJ and Abtact - what are the MOSDAB issues you are concerned about? — Carl (CBM · talk) 00:08, 24 March 2008 (UTC)
My point is simply that the version suggested by Jheald does not conform to mos:dab. First, it relegates the primary meaning entropy to just one possibility (see JHunterJ above for why the dab page is not the place to argue this point). Second, DAB pages are for help navigating and not for imparting information about how different articles relate to each other. Third, there are several redlinks which should be removed. For the purpose of disambiguating, ENTROPY should be considered as just a word, not a concept. I tried to be brief, JHunterJ would do better. Abtract (talk) 09:47, 24 March 2008 (UTC)
The intro paragraph about the primary topic (that is, the subject of the article at the base name Entropy), the formatting of titles. Overlong descriptions and the inclusion of an article category on a non-article page are less of an issue, but still issues. If the problem is that information theory entropy is the second-most common usage (behind thermodynamic entropy), then that entry could be moved to the top of the list (still after the intro paragraph). If the problem is that thermodynamic entropy should not be in the intro paragraph, then current Entropy needs to be moved so this dab page can be moved to the base name. -- JHunterJ (talk) 11:27, 24 March 2008 (UTC)
Abtract and JHunterJ, now I see more what you're concerned about, and I hope it will make sense to everyone else as well.
It's certainly correct that the redlinks should be removed, article categories should be removed, and the descriptions are meant to be short (although, given that many of the things here are very closely related, the descriptions should be clear enough to let the reader know what each article will discuss).
Abtract: I agree that MOSDAB does request a link back to Entropy at the top of this page.
My impression is that, as a reader, it would be much less useful for me for this dab page to group together the physics, probability, dynamical systems, and software articles into one long section on "science and technology". As a reader, I scan a dab page for a section that relates to the meaning I am looking for. It seems to me that splitting the "science and technology" section into several shorter sections (science, math, programming) should make navigating this large collection of articles easier. — Carl (CBM · talk) 12:18, 24 March 2008 (UTC)
mosdab suggests only that, "The list may be broken up by subject area" and "Section headers may be used on longer lists". Personally, I find that dab pages with several sections with only one, two or even three items per section tend to be more confusing than helpful unless I know what I am looking for. This is because I spend more time wondering which section I should be looking in than I would if there were fewer, more generalised, sections. IMHO sections are fine if they genuinely assist the average reader, but should not be included to impart information by categorising the entries. Again personally, I never (never say never) use categories with fewer than 3 items and I prefer 4 or more. Abtract (talk) 12:47, 24 March 2008 (UTC)
My impression of the idea behind disambiguation pages is that they are intended to help readers who do know what they are looking for, but didn't find it when they typed that term into the search box. So for example I type "Entropy" into the box, and I am taken to an article on physics that I wasn't looking for. I already know what I am lookig for, because that's why I typed entropy into the search box.
The restriction that there need to be 3 or 4 items per group seems to be arbitrary to me. The grouping should be done naturally, keeping in mind that other articles might be added later. Personally, I find it surprising that articles about computer software are mixed in with articles about physics. Certainly someone who arrives here knows already which of those they were looking for. — Carl (CBM · talk) 12:57, 24 March 2008 (UTC)
That's all fine provided readers think in boxes ... and their boxes are the same as those we design for this page. It is just because we all don't think in boxes all the time and, when we do, my boxes may well differ from yours, that I am dubious about too many small sections. Leaving space for future additions is, as you will have realised by now, a non-argument as sections can be changed at a later date as necessary. Having said all that, the number of sections was one of the least important problems with the page whan I first started cleaning it and it isn't one I would go to the barricades for.Abtract (talk) 13:19, 24 March 2008 (UTC)
Considering that this version is what existed before, cleanup was overdue, and it's good you decided to do it. I think that now it's more the fine nuances that concern some people - the phrasing of definitions, the sectioning. But I'm sure everyone can find some agreeable solution. — Carl (CBM · talk) 14:43, 24 March 2008 (UTC)
Thanks, Carl; I completely agree. Reverting the cleanup efforts rather than continuing to refine them serves no one. -- JHunterJ (talk) 21:34, 24 March 2008 (UTC)

[edit] Lists

I have removed several redlinked items and instances of evanscent meaning (songs called "entropy" which do not have dedicated articles at WP), in the spirit of "Wikipedia is not an indiscriminate collection of information". Arcfrk (talk) 03:09, 24 March 2008 (UTC)

As long as the entries include a blue link which covers the "Entropy" portion, they should be disambiguated. Things that have no blue links or only blue links to articles that do not cover "Entropy" should be removed. That is the discriminating criterion. -- JHunterJ (talk) 11:22, 24 March 2008 (UTC)
Removing the red links is a good start. Every title in this version seems to contain the word "entropy" and is blue. Are there titles in that version that anyone disagrees with including here? — Carl (CBM · talk) 12:26, 24 March 2008 (UTC)
Looks good to me ... I suggest the "sister projects" tag is removed and any relevant article put into "See also". Abtract (talk) 12:33, 24 March 2008 (UTC)

[edit] Arguments, please

The grouping structure imposed on the list is far too coarse to be useful and inferior to a more refined grouping structure, as is commonly employed on other dab pages, such as Arrow, Bell, Constant, and so on, up to and including Zero. The ordering is also such that the most common meanings are buried between meanings that are rarely encountered. The warriors for a coarse structure apparently either do not care for correctness, or understand the topic insufficiently to ensure correctness. The very first sentence, "Entropy is a measure of the unavailability of a system to do work", is wrong. It is not the unavailability of the system that is the issue (like a computer system that is down), but the unavailability of the system's energy for being put to work. It is also really awkward to just give the word you are disambiguating as an entry like that, with a characterization that is not only wrong but also not giving a context for actually disambiguating the term. Topological entropy is not "the number of distinguishable orbits an iterated map can have" (which would always be an integer). And what happened to Measure-theoretic entropy?

Can we please see some arguments on this talk page from the proponents of this coarse structure explaining why it is supposed to be superior? The latest revert to the coarse structure had an edit summary Should be more like Zero (disambiguation), a swell example of a good dab page. However, that page is much more similar in structure to this dab page before that revert, having no fewer than six sections. What gives?  --Lambiam 12:46, 24 March 2008 (UTC)

I think that one issue is that there were other MOSDAB issues that people were concerned about - so perhaps one person was reverting based on issue A and another person was reverting based on issue B.
I'm sure that everyone agrees that the definitions need to be correct, and that we can include measure-theoretic entropy. Some people were concerned that the definitions might need to be reworded to be shorter (but still correct).
The division into sections does seem to be an area where more discussion is needed. — Carl (CBM · talk) 12:52, 24 March 2008 (UTC)
The current version does have a problem in that it says that Entropy (classical thermodynamics) ("a measure of how close a system is to equilibrium") is a different meaning than the primary meaning Entropy ("a measure of the unavailability of a system to do work"), while in fact Entropy (classical thermodynamics) is more like a subtopic of Entropy spun off according to Wikipedia:Summary style. I don't think this situation is treated in WP:MOSDAB.
So I'd like to see something like I wrote in #Proposal by Jitse below. I don't like the "in particular, we have" phrase, but I couldn't come up with anything better. I copied the descriptions from the current version, but it looks like other versions have descriptions that are better thought out. I think that after grouping the pages relating to thermodynamic and information-theoretic entropy together, it will be easy for the reader to pick the intended meaning out of the "science and technology" section, so it won't be necessary to split it further. -- Jitse Niesen (talk) 13:23, 24 March 2008 (UTC) (via several edit conflicts)

[edit] Proposal by Jitse

Entropy is a measure of the unavailability of a system to do work. In particular, we have:

Entropy may also refer to:

In science and technology:

In modern culture:

  • …
How about, "Entropy is a measure of the energy in a system that is unavailable for performing work" or something like that? I like the sublists, but MOSDAB seems to be silent about them. — Carl (CBM · talk) 14:45, 24 March 2008 (UTC)
There is no point in suggesting new definitions here ... Talk:Entropy is the place for that. Abtract (talk) 18:53, 24 March 2008 (UTC)
Huh? We should be able to word the definition here independently of that article. But looking at Entropy will show the issue: both the wording I propose and the wording there emphasize the issue that it's the system's energy being unavailable, not the system itself being unavailable, that's important. I think this is an issue that some posts above were concerned about. — Carl (CBM · talk) 19:05, 24 March 2008 (UTC)
The word "energy" has clearly been missed in error, I suggest adding it in. The reason that any bigger redefinition should be discussed at Entropy is that dab pages simply reflect the content of the lead in target articles, they should not invent different definitions (however valid they may be). Abtract (talk) 19:52, 24 March 2008 (UTC)
Certainly they should reflect the lede, but there's no requirement that the text has to be identical. Original prose is just fine. — Carl (CBM · talk) 19:56, 24 March 2008 (UTC)
Looks good to me. I'm in favour of sub-lists as it makes it easier for the reader to find what they are looking for per WP:IAR. --Salix alba (talk) 18:43, 24 March 2008 (UTC)
Sublists are allowed by mos:dab#Longer lists. Abtract (talk) 18:53, 24 March 2008 (UTC)
I didn't think that that really discussed sublists (** lists inside of * lists). — Carl (CBM · talk) 19:07, 24 March 2008 (UTC)
See mos:dab#Longer lists. Abtract (talk) 19:52, 24 March 2008 (UTC)
I just read it again. Could you point out where it discusses sublists? Here's the text:
"The list may be broken up by subject area: ...
"Section headers may be used on longer lists instead of, or in addition to, bold subject area headings, but using more than a single level, as on Aurora (disambiguation), is rarely necessary. Section headers should not include links. See Wikipedia:Writing better articles#Headings for more.
"On longer lists, {{TOCright}} may be used to move the table of contents to the right hand side of the page. This reduces the amount of white space and may improve the readability of the page. (For more information, see Help:Section#Floating the TOC.)"
— Carl (CBM · talk) 19:56, 24 March 2008 (UTC)
I have bolded the words to watch for, thus: "section headers may be used in addition to bold subject area headings" ... and ... "more than a single level is rarely necessary". Abtract (talk) 20:09, 24 March 2008 (UTC)
I'm not talking about headers - I'm talking about:
  • This is a list
    • This is a sublist
Certainly subsections are permitted; the language is clear. — Carl (CBM · talk) 20:15, 24 March 2008 (UTC)
If the problem isn't sublists, then I don't see what's wrong with the version of Arcfrk (talk · contribs), here [1].
But I'd like to respond to some of the comments made by JHunterJ and Abtract last night.
  • Primary meaning. The issue is this: the meaning of Entropy in thermodynamics was the first meaning historically, and is also the first meaning most users are likely to be exposed to, in high-school chemistry classes. So it is not inappropriate for Entropy to direct to an article on Thermodynamic Entropy. However, arguably the more fundamental, certainly the more broad and more general meaning of Entropy is Entropy in Information Theory, even though Information Theory only appears at university level. So, for the purposes of disambiguation, both should be considered primary meanings. Jitse's proposal would IMO be improved if the group of thermodynamic definitions appeared under science and technology.
When Abtract says this would "relegate the primary meaning entropy to just one possibility", this is appropriate. Thermodynamic entropy is just one of the two main possibilities.
  • Descriptions. The descriptions in Arcfrk's version seem, without exception, to be more accurate than what is being proposed (eg in Jitse's version above) to replace them. There seems to have been no detailed case made for change. Is a specific layout problem being asserted? I'm not seeing it on my screen.
Abtract writes: DAB pages are for help navigating and not for imparting information about how different articles relate to each other. I think that misses part of the picture. Sketching in the relations with (usually very few) more words can make the structure much more easy to assimilate and to navigate. It helps people to understand why things have been grouped together. And it helps reassure them that they have properly "chunked" the whole list, and that they either don't need to look at an article that's definitely not what they are looking for, nor do they need to worry that they may have missed an article which is what they are looking for.
So for example, Entropy (ecology), use of Shannon entropy as a measure of biodiversity in the study of ecology is more helpful, strictly for the purpose of disambiguation, than just a measure of biodiversity.
Or Entropy encoding, data compression strategies that set out to achieve message lengths close to the Shannon entropy is much more definitive than just a lossless data compression scheme; and by clueing why the word Entropy is there, it makes the page as a whole easier to chunk and assimilate.
The changes made have been made without any detailed explanation, discussion or review; and I am not entirely convinced that they were made by people who fully know what they are writing about. I believe there should, at the very least, be line-by-line discussion of what was perceived to be wrong with the previous summaries (cf Arcfrk's version [2]), summarised by people who did actually know what they were writing about.
MOSDAB is pretty clear that categories aren't used. We might as well follow it. The descriptions can always be tweaked later once the general format is in place, so I'm not too worried about those at the moment. — Carl (CBM · talk) 20:17, 24 March 2008 (UTC)
In this case there is a good argument that these category links should be included. If the only argument being put up as to why they should not be is just that MOSDAB says so, then that seems an open-and-shut case for WP:IAR. Jheald (talk) 20:33, 24 March 2008 (UTC)
It's a category for articles. Dab pages aren't articles. That seems an open-and-shut case for its omission, ignoring all rules. -- JHunterJ (talk) 21:38, 24 March 2008 (UTC)
I'm not proposing putting the article into the category. I'm proposing linking to the category from the article. You do, I hope see the difference, and therefore understand that what you have written is completely off the point. Jheald (talk) 22:18, 24 March 2008 (UTC)
I can be forgiven for the mistake, I think, since your knee-jerk reversions had the effect of restoring such a category, and since the edit you were reverting from also included links to the categories without placing the article into it. -- JHunterJ (talk) 02:46, 25 March 2008 (UTC)
As to descriptions, my point is one of principle: it is helpful to include detail, when this helps users more easily assimilate the whole page, and therefore more quickly and more confidently navigate it. Jheald (talk) 20:37, 24 March 2008 (UTC)
I find this related to Wikipedia:Disambiguation#Partial_title_matches, which says that dab pages aren't meant to include every article with a partial name match. In this case, we should link to the "main" articles on the various types of entropy, but not to all their descendant articles. — Carl (CBM · talk) 20:47, 24 March 2008 (UTC)
Besides, as far as I can see MOSDAB says only that dab pages should not be categorised (other than as DAB pages). It doesn't say anything one way or another about links to categories, particularly when these are sequestered to the side in a useful side-box.
And note above all the last words of MOSDAB: These guidelines are intended for consistency, but usefulness to the reader is the principal goal. So ignore these guidelines if you have a good reason. Jheald (talk) 21:09, 24 March 2008 (UTC)

Re: "for the purposes of disambiguation, both should be considered primary meanings." No, the primary meaning is defined as the one at the base name. If both are the primary meaning for purposes of disambiguation, then the disambiguation page goes to the base name. Simple. -- JHunterJ (talk) 21:38, 24 March 2008 (UTC)

Except that in this case, both are the primary meaning for purposes of disambiguation, but there are understandable reasons for leaving Entropy to redirect as it currently does.
Why is that such a problem for you? Why do you think it makes the page less useful to readers? Jheald (talk) 22:09, 24 March 2008 (UTC)
I think that part of the issue is that ordinarily there would be no "primary meaning" at all. But there is a special exception: a page entitled "Foo (disambiguation)" starts with a sentence linking back to Foo, This is purely stylistic; otherwise the link to Foo would be in the list of disambiguated topics instead of the top. So instead of "primary meaning" it's batter to think of it as "page whose name matches this disambiguation page".
There are many shades of utility, and I don't currently see how this particular issue (whether to put a sentence at the top or in the disambig list) makes a significant difference. On the other hand, I do agree with you that breaking apart long lists into shorter lists by topic is beneficial. — Carl (CBM · talk) 22:13, 24 March 2008 (UTC)
We're not talking about just a sentence. We're talking (per Jitse's proposal above) about separating off a group of half a dozen uses. Stylistically, that seems a lot inferior to merely opening "Entropy can have several meanings:", with no distinguished "primary meaning" -- the way things used to be. [3] Jheald (talk) 22:26, 24 March 2008 (UTC)
"Entropy" has a primary meaning (or, as Carl has explained, a page whose name matches the base name of the disambiguation page). Enter "entropy" in the search box and click on "Go" and you will see what it is currently defined as in Wikipedia. That's the way things used to be, but this page mistakenly was written as if there were no primary meaning. The ways to fix that are ... ah, never mind, you've already said you don't want to do the legwork for changes there. -- JHunterJ (talk) 00:12, 25 March 2008 (UTC)
Those of you who keep applying WP:IAR here, please, read what it means. Jitse's proposal is not helping one bit and if there is anything disagreed with in WP:DISAMBIG or WP:MOSDAB bring it up there. Lord Sesshomaru (talk • edits) 00:19, 25 March 2008 (UTC)
Sesshomaru, please read the reason behind my proposal. Let me repeat it for you: the current version says that Entropy (classical thermodynamics) ("a measure of how close a system is to equilibrium") is a different meaning than the primary meaning Entropy ("a measure of the unavailability of a system to do work"), while in fact Entropy (classical thermodynamics) is more like a subtopic of Entropy spun off according to Wikipedia:Summary style. Do you disagree with this, or do you disagree that this is a problem, or do you have a better solution?
Jheald, the only link that should go to this page is at the top of Entropy. Thus, the readers that this page should target are those that went to Entropy (probably by typing "entropy" in the search box), saw that that page did not talk about the entropy that they had in mind, and came here. That's why the meaning at entropy (the so-called "primary meaning") should get special treatment. -- Jitse Niesen (talk) 01:43, 25 March 2008 (UTC)
That's indeed what happened when I was editing an article on "Topological entropy", where a link at the very beginning pointed to the "Entropy" page. The intended primary meaning in that case was totally different (either measure-theoretic or information theory entropy, as far as I could tell), and had nothing or little to do with thermodynamics, which is the context of "Entropy". I am puzzled as to why, having come to a dab page via the wrong article, does a reader need to see that wrong (for his purposes) meaning as the first option again? Arcfrk (talk) 03:15, 25 March 2008 (UTC)
Thought some were still fighting for this revision. If this is no longer the case, then what was up with the WP:IAR argument? Lord Sesshomaru (talk • edits) 02:10, 25 March 2008 (UTC)
No one is fighting for this revision. As far as I could tell, everyone agreed that it was factually more accurate and also removed clutter (red links, irrelevant links). There is still some disagreement about whether the leading sentence must conform to "primary meaning" scheme or not, and the categories. Other than that, Jheald explicitly stated that he liked it and asked for specific line-by-line comments of what was perceived to be wrong with it, and no objections were raised. Except, that is, for one editor with itchy fingers who stated that "Zero (disambiguation)" is a "swell example of a good dab page", and proceeded to replace it with something that looks a lot less like his "swell example" – the point made by Lambiam above (can you say Bad faith edit fast enough? or is it just "Undo first. Think later" phenomenon?) Arcfrk (talk) 03:15, 25 March 2008 (UTC)
Wow, you're still stuck on that? You want a good reason, here it is. Lord Sesshomaru (talk • edits) 03:26, 25 March 2008 (UTC)
I am quite familiar with the revert you linked to in your reaction, and I don't see how or why it is a "good reason" for your revert being discussed by Arcfrk. Could you elaborate?  --Lambiam 20:25, 25 March 2008 (UTC)
Excuse me? Lord Sesshomaru (talk • edits) 20:28, 25 March 2008 (UTC)
You wrote that a certain edit is a good reason. I don't understand why or how that edit is a good reason. Could you explain why that edit is a good reason?  --Lambiam 21:25, 25 March 2008 (UTC)

I diagree with Jitse Niesen's proposal. Reproducing my comments from below. I don't see a reason to use Jitse Niesen's proposal without changing the headers to make clear that those familiar with the arts do not privilege one sense over the other. i.e. change "Entropy is a measure of the unavailability of a system to do work. In particular, we have:" to "Entropy has two senses in the sciences / Thermodynamic Entropy is a measure of the unavailability of a system to do work:" and "Entropy may also refer to: / In science and technology:" to "Information Entropy is a measure of uncertainty:" Pdbailey (talk) 03:55, 26 March 2008 (UTC)

[edit] List?

I have a suggestion that may solve the problem. Set up a list called (something like) Entropy in science and technology into which should go all the articles previously discusssed and any other terms (even without articles). lists are more relaxed than dab pages in particular lists are to impart information as well as assisting with navigation. If we use the list, then this dab page only needs mention the primary usage Entropy and the list plus of course all the modern culture items and maybe we would all be satisfied. The list could be used by experts in the field to be as informative as they think suitable. Imagine having no dab specialists breathing down your neck ... Abtract (talk) 21:18, 24 March 2008 (UTC)

Except that (1) Thermodynamic Entropy isn't the primary usage. It's one of the primary usages.
(2) Why bother? Why not just leave things as they were? What scintilla of usefulness is all this officiousness supposed to be achieving? Jheald (talk) 22:13, 24 March 2008 (UTC)
I don't think I can support moving most of the disambig page to a list. Lists have their own issues (scope and sourcing, in particular) that aren't as much of an issue with disambig pages.
I think it might help if we all avoid the term "primary meaning", which can be a loaded term. — Carl (CBM · talk) 22:15, 24 March 2008 (UTC)
MOSDAB uses the term "primary topic" (see MOSDAB#Linking to a primary topic). If the term Title has a "well-known meaning", for which it is unlikely that users are looking when they type in the search term "Title (disambiguation)", the guideline recommends to place the link back to that "primary topic" at the top, and to not "mix it in" with the other links. So we may try to avoid the loaded term "primary meaning", but it is hard to ignore or proscribe the term "primary topic".
If the article Entropy had been named Entropy (thermodynamics), we would not be having this discussion. On the applicability of the – by itself quite reasonable – MOSDAB rule to this dab page I want to say the following. Applied to the title Entropy, the assumption of the rule is that a user will only reach the dab page by entering "Entropy (disambiguation)" in the search box. I submit that it is actually rather unlikely that a user types that in; if they reach this page it is much more likely they were following a link such as "entropy" on the pages Complexity and Kipple, or the link in the hatnote of Information entropy. Thus, in my opinion, the rationale for the MOSDAB rule does not apply here.  --Lambiam 21:09, 25 March 2008 (UTC)

[edit] NB

Informatics and thermodynamics aren't really talking about different things when they speak of entropy; the physics concept is pretty much just what is meant in the information context. However, randomness is a troublesome concept: the mathematical defintion a random variable is a measurable function" doesn't convey the flavor.

When you heat a physical system, the molecules bounce around, and are more random then when you cool the system and the molecules settle into a uniform crystal. The physics defintion is about the heat content of the system, but it pertains to randomness. When you buy a deck of cards, they come in fixed order from the factory, from the ace of spades to the deuce of hearts (something like that) and contain no information (the information content is zero, the entropy is zero); when you shuffle the deck, the entropy goes up. A random deck is capable of conveying information: to see that, imagine coding a message with numbers; the sequence will look like shuffled numbers. The entropy of the deck is a measure, like hamming distance, of how far away it is from a default ordered state: that defintion doesn't convey the flavor of information content but it's computationally tractable.

I don't know a great definition for entropy in the abstract, but I'd start with the defintion from physics, and hope for unification. Pete St.John (talk) 19:07, 25 March 2008 (UTC)

Some people are arguing that, since we have an article with the name Entropy (which happens to be about thermodynamical entropy, with some other notions thrown in very late in the article as a mini dab page), MOS:DAB prescribes that this dab page should start with a separate paragraph along the lines of "Entropy is <thermodynamical meaning>. Entropy may also refer to:". How does, what you are writing above, relate to that? Do you disagree with them? If so, is that because you think that MOS:DAB does not apply (it is being interpreted the wrong way), or is it that you think WP:IAR should prevail here? Or do you mean yet something else?
For an intro paragraph (unacceptable to some because of the MOS:DAB issue), what about: "In general, entropy is a measure of the disorder in a system. Originally a notion developed in thermodynamics, a branch of physics, the term is also used and has different definitions in various other areas."
 --Lambiam 20:07, 25 March 2008 (UTC)
I like that paragraph ("In general...") fine myself. I would word it, "...has different technical definitions..." because I, myself, want to emphasize that there is a unifying (if not entirely enunciated) concept. Perhaps:
Entropy is <thermo meaning>. The term is used for related concepts in other fields: <list>
would something like that be acceptable? I just want to advocate inclusiveness, because I don't believe computer scientists and physicists are talking about entirely different things, just using different technical defintions for different contexts, and that together these give us a more tractable concept of randomness. Pete St.John (talk) 20:54, 25 March 2008 (UTC)

[edit] There is a solution

I thought at first that creating a list for the science related articles would, of itself, solve the problem due to the more relaxed rules concerning lists. Had it not been for the heated debate on the primary meaning this would have worked well. JHunterJ has pointed out the way to solve the primary meaning problem as follows - simply move the content of Entropy to Thermodynamic entropy and use the Entropy page for disambiguation. This would remove its primary meaning status and the dab page could be laid out more flexibly. Obviously there would need to be a consensus of editors who actually know something about the subject to agree the move but, listening to the above debate, I don't think that would be a problem. I still feel that splitting the dab page into a list showing all article on and related to (science) entropy (which could utilise the page currently occupied by Entropy and a dab page showing all the non-science uses and one ref to the list (which could then stay where it is). Whatever you think of the list idea, I urge you to consider carefully JHunterJ's suggestion to eliminate the primary meaning by moving Entropy to Thermodynamic entropy - it would solve the problem at a stroke. Abtract (talk) 21:15, 25 March 2008 (UTC)

It would be more appropriate to move it to Entropy (thermodynamics), which is also more in line with the page titles Entropy (classical thermodynamics) and Entropy (statistical thermodynamics). Although "thermodynamic entropy" would be generally understood, it is in that form not a very common term. I'd estimate that of the uses of 'entropy" in the sense of thermodynamics, less than one in fifty would use "thermodynamic entropy".
However, an even simpler solution (in my opinion) is to agree that the MOSDAB rule does not apply here, as I've argued above.  --Lambiam 21:38, 25 March 2008 (UTC)
No problem, call it what you will ... but sadly mos:dab rules do apply. Abtract (talk) 21:45, 25 March 2008 (UTC)
I see no reason why one is privileged over the other, both are fundamental concepts in science and critical to their fields. However, Wikipedia has a readership and I would guess that those looking for "Entropy" without further knowledge of these two definitions are looking for the Boltzmann sense of the word, and not the Shannon sense of the word--if they know they are at the wrong page, they can click on the disambig link at the top. Now, that said, I don't see a reason to use Jitse Niesen's proposal without changing the headers to make clear that those farmiliar with the arts do not privilege one sense over the other. i.e. change "Entropy is a measure of the unavailability of a system to do work. In particular, we have:" to "Entropy has two senses in the sciences / Thermodynamic Entropy is a measure of the unavailability of a system to do work:" and "Entropy may also refer to: / In science and technology:" to "Information Entropy is a measure of uncertainty:" Pdbailey (talk) 03:52, 26 March 2008 (UTC)
Keep the current version per arguments above, and WP:MOS-DAB (especially this). Lord Sesshomaru (talk • edits) 03:59, 26 March 2008 (UTC)
Sesshomaru, I would disagree with the assertion in the MOS-DAB that, "it is unlikely that this well-known meaning is what they are looking for." and that it increases the clarity of the article to link to it. The rule appears to be more of a way to help along DAB page clarity, but it does not improve clarity for this article (I have no idea what it should link to and have not clicked, so I don't know where it goes). Your proposal appears to violate the spirit, if the letter of the rule. Pdbailey (talk) 15:37, 26 March 2008 (UTC)
I would add that the current version is a good example of a clear page. Pdbailey (talk) 15:40, 26 March 2008 (UTC)

[edit] Attempt to move forwards

I've put up a new version [4], which I hope takes on board some of the DAB project people's concerns, while retaining some of the structure and more accurate definitions from earlier page versions. I hope this may provide a useful basis for further discussion, and if there are still issues with it for some people, that we can work through this talk page, with any concerns explained in detail here and then discussed until consensus can be reached, rather than knee-jerk reverts.

Finally, let's remember that WP:MOSDAB is intended as flexible guidance, not a rigid straitjacket. As MOSDAB itself concludes: These guidelines are intended for consistency, but usefulness to the reader is the principal goal. So ignore these guidelines if you have a good reason. Jheald (talk) 10:52, 26 March 2008 (UTC)

I continued on with that attempt; doing the same (I hope) grouping in a more-usual dab page style. I also shortened the descriptions, I hope not too much, since the dab should provide just enough information to get the user to the desired article. -- JHunterJ (talk) 11:39, 26 March 2008 (UTC)
But you have removed "Kolmogorov–Sinai entropy", which is the standard name. Also, it makes very little sense to bury two extremely common mathematics uses within general "scientific usage". Does anyone here object to using the standard section headers, as many other dab pages do, including the "swell example of a dab page" Zero (disambiguation)? Arcfrk (talk) 19:13, 26 March 2008 (UTC)
If Kolmogorov–Sinai entropy is the standard name, why isn't it referenced here? Lord Sesshomaru (talk • edits) 22:12, 26 March 2008 (UTC)
You may well ask why no one bothered to write a dedicated article about Kolmogorov–Sinai entropy since it is so important. The answer is that contributors are volunteers, they have other uses for their time, and not everyone has the same level of expertise or exercises the same standards in researching the topic they write on. As a result, there are many mistakes, omissions, misattributions and glaring gaps in Wikipedia coverage. Nevertheless, people will come to this page looking for Kolmogorov–Sinai entropy. Why do you have to interfere in the subjects which you do not understand? It is a huge waste of time to argue with you and to clean up your errors, and this time could have been much better spent, say, writing an article on Kolmogorov–Sinai entropy. By tying expert editors with your unconstructive "contributions", you prevent them from improving the content elsewhere. Arcfrk (talk) 23:10, 26 March 2008 (UTC)
Please stop trying to pick a fight. All I'm saying is that if you want to use the link, cite it in the article. Be bold in updating pages. Lord Sesshomaru (talk • edits) 23:36, 26 March 2008 (UTC)
Then you should perhaps have asked Arcfrk to update the article, instead of reverting aggressively. Anyway, it's in there now.
I think that Shannon / information entropy is clearly more important than either of the mathematical meanings, so I'm moving it up. I'm sure I'll hear it if people disagree. -- Jitse Niesen (talk) 00:00, 27 March 2008 (UTC)
Re: Kolmogorov: the correct thing to do was to avoid surprise for the reader. Reverting until the change could be made on the target page causes no harm. Now that it's been added (which it might not have been otherwise), that redirect can be used without surprising the reader. -- JHunterJ (talk) 00:36, 27 March 2008 (UTC)

Hey, I wrote the Kolmogorov bit, and I almost did create a new page for it, I just got lazy, because it would have needed a long intro to the notation, and I had worked plenty hard enough already. Hardy volunteers are welcome to split it off into its own page. FWIW, the entire field of dynamical systems is rather weakly represented at WP. linas (talk) 04:24, 27 March 2008 (UTC)

[edit] Rationale for MOSDAB rule does not apply here

I repeat something I wrote before that no one reacted to.

The essence of MOSDAB#Linking to a primary topic is as follows. If the term Title has a "well-known meaning", for which it is unlikely that users are looking when they type in the search term "Title (disambiguation)", the guideline recommends to place the link back to that "primary topic" at the top, and to not "mix it in" with the other links.

Applied to the title Entropy, the assumption of the rule is that a user will only reach the dab page by entering "Entropy (disambiguation)" in the search box. I submit that it is actually rather unlikely that a user types that in; if they reach this page it is much more likely they were following a link such as "entropy" on the pages Complexity and Kipple, or the link in the hatnote of Information entropy. Thus, in my opinion, the rationale for the MOSDAB rule does not apply here.  --Lambiam 22:04, 26 March 2008 (UTC)

Bearing in mind that there should be no piped links to a dab page, the main way a reader should arrive here is by typing "entropy", discovering that it was the wrong article and then clicking the link to the dab page in that article. Of course there will be some particularly savvy readers who might guess that there is a dab page and come straight here. I am amazed that it hasn't become clear that the primary topic problem can be made to go away only by the (simple) device of moving Entropy to Entropy (thermodynamics) and this page to Entropy ... there would then be no primary meaning! Abtract (talk) 23:51, 26 March 2008 (UTC)
Woof! I'm with Abtract here, the current article called entropy utterly fails to discuss the concept in the whole, and is very narrowly focused on thermodynamics. I put in a page-move request for this, I hope it doesn't get bogged down with goofiness. linas (talk) 04:21, 27 March 2008 (UTC)
Step one is taken here. Talk:Entropy#Survey. Lets get this out of the way, then we can copy the bulk of this page to Entropy, and be done with it. linas (talk) 04:40, 27 March 2008 (UTC)
Thanks for pointing out that option, Abtract, and thanks, Linas, for doing the legwork. I pointed out the option when this brouhaha started, but no one was willing to do the legwork then. -- JHunterJ (talk) 11:38, 27 March 2008 (UTC)

[edit] the general case

Hi,

In the above discussions, it became clear (or so I thought) that the majority of folks thought that the "lead" article on entropy was heavily biased towards thermodynamics, and utterly failed to make any mention whatsoever of entropy as it is used in information theory, dynamical systems, or other branches of science or mathematics. The obvious course of action was to rename the article "Entropy" to "Entropy (thermodynamics)", while creating a new article on the general concept of "Entropy". I see that the majority of ya'll voted for the page move, but the overall page move didn't happen because some, I'll call them "amateurs", voted against it. I think this is shameful, and I think WP should have a policy against "amateurs" with respect to science topics; but that's another matter.

What can be done: the next most immediate course of action is to create the article on entropy that everyone seems to want, an article that does discuss the concept in its general setting, and does actually review its uses in fields other than basic thermodynamics. Once that article is created, we could once-again try to replace the current "lead" article with the appropriate once. Towards that end, I created a stub at Entropy (general concept). Clearly, this stub needs a huge amount of work to turn into something reasonable. However, I cannot even get that off the ground, as I seem to be finding myself in yet more, ahem, pointless conversations. linas (talk) 19:43, 9 April 2008 (UTC)

I see a different if not consensus then at least majority view. (See many of the "Oppose" votes at Talk:Entropy). Namely that thermodynamic entropy, particularly classical thermodynamics, is the most common (and most commonly first met) usage. Information-theoretic entropy is substantially an independent usage. And the relations between the two are probably best treated as an advanced topic. Especially when there are chemists like Denbigh who flat-out decry the idea of identifying physical entropy (to them objective) with an information entropy (subjective). Okay, many people (including myself) may think that's a spurious objection; but the fact remains that you don't need to identify Gibbs entropy as a Shannon entropy to use it, and many don't.
Note that some people also have issues with describing entropy as "disorder", particularly in a chemistry context. See eg talk page archives here, here and here.
As I tried to indicate in my edit summaries, I think it's a mistake to move the science and technology links off this page. I really think one redirect (ie to here) is about the most we can make people go through to get to a disambig page to find their article title. I just don't think people will click through twice to get to the links a new page is offering; I think that it is much easier for users if those links remain here on a first dab page.
Finally, regarding a page for "general" entropy. We do already have Information entropy, which I think does a reasonable job communicating Shannon entropy. Do we really need another page, besides that one, and this one, on Shannon-like entropies? Is it not enough for the various pages linked from here to each explain their relation to Shannon entropy? Jheald (talk) 20:04, 9 April 2008 (UTC)