Wiktionary:Votes/2010-02/ToC format

Definition from Wiktionary, the free dictionary
Jump to: navigation, search

ToC format[edit]

  • Voting on: Deciding which way to format the "Table of Contents" (ToC) boxes in entries.

  • Vote starts: 00:00, 25 February 2010 (UTC)
  • Vote ends: 24:00, 27 March 2010 (UTC)

Option 1[edit]

Using the current format, with the table of contents being above the entries, uncollapsed, on the top-left, with all headings and subheadings being shown along with the numbers.

Option 2[edit]

Having the table of contents boxes float to the left, showing all headings and subheading being shown along with the numbers.

Option 3[edit]

Having the table of contents boxes float to the left, showing only language headings, with the numbers being hidden.

Option 4[edit]

Having the table of contents boxes float to the right, showing all headings and subheading being shown along with the numbers. (This option can be tested by going to WT:PREFS and checking "Put the table of contents onto the right of entries.", or by turning on the "Right-hand side Table of Contents" gadget.)

Option 5[edit]

Having the table of contents boxes float to the right, showing only language headings, with the numbers being hidden.

Option 6[edit]

Having the table of contents boxes collapsed by default. (This option can be tested by going to any entry and clicking the "[hide]" button in the ToC.)

Option 7[edit]

Having the table of contents being above the entries, uncollapsed, on the top-left, showing only language headings, with the numbers being hidden.


This vote is using the Schulze method of voting. Please add your vote by adding the number of your most preferred option, followed by a ">", followed by your second-most preferred option, followed by ">", etc. Note that there needs to be clear consensus that the winner is preferable to option 1 (the current format) for the change to be implemented.

  • 3 > 5 > 6 > 7 > 4 > 2 > 1 --Yair rand 01:07, 25 February 2010 (UTC)
  • 5 > 4 > 3 > 7 > 2 > 6 > 1 --Bequwτ 05:49, 25 February 2010 (UTC)
  • 2 > 1 > 6 > 4 > 3 > 7 > 5 -- Prince Kassad 07:37, 25 February 2010 (UTC)
  • 1 > 7 > 4> 5 > 3 > 2Ƿidsiþ 13:52, 25 February 2010 (UTC)
  • 1 > 2 = 4 > 6 > 7 > 3 = 5: We have many, many entries with only one language but several POSes or the like, so listing only languages by default on all entries seems like a bad idea. (It will also look a little silly: "Table of Contents: English".)​—msh210 19:03, 26 February 2010 (UTC)
    Most of the single language entries don't have ToCs at all, but for those that do, I don't think it will look much sillier than the standard where the ToC is almost as large as the entry (compare option 1/option 3).--Yair rand 19:23, 26 February 2010 (UTC)
    Consider [[ישב]].​—msh210 19:32, 26 February 2010 (UTC)
    I've been using option 3 on my own css for the past three weeks so I can't tell exactly how ישב looks by default, but I can't understand how a long list of "Etymology 1, Pronunciation, Verb, Conjugation, Etymology 2, Pronunciation, Verb..." could be useful. --Yair rand 19:41, 26 February 2010 (UTC)
  • 7 > 5 > 4 > 6 > 1 > 3 > 2 --Vahagn Petrosyan 20:21, 25 February 2010 (UTC)
  • Symbol oppose vote.svg Oppose this vote; I oppose this multi-option vote, which uses the Schulze method. --Dan Polansky 09:41, 26 February 2010 (UTC)
    Why? --Bequwτ 12:50, 26 February 2010 (UTC)
    Because the vote is intransparent to me when structured in this way and using Schulze method. I think that votes that use Schulze method are harder to design well, and can be subject to various side effects much more than simple approval votes. Such votes may be more sensitive to the slicing of the choice situation into particular options. In particular, the vote "7 > 6 > 5 > 4 > 3 > 2 > 1" does not in any way indicate the strength of the distance between the options; it merely indicates their order. Neither does it indicate acceptability, merely the preference, so if I find 7, 6, 5 and 4 acceptable while 3, 2, 1 inacceptable, the vote does not capture that. By contrast, Wiktionary:Votes/2010-01/Setting ToCs to be on the right hand side by default gives a clear result. --Dan Polansky 23:09, 26 February 2010 (UTC)
    Of course no voting method is perfect but the Schulze method fairs very well against approval voting. One may have a personal bias towards certain criteria but there's no obvious reason approval voting is better. Almost no method guarantees "independence of irrelevant alternatives" so I don't know why you'd say Schulze is more sensitive to choice slicing or harder to design for. If you want to mimic approval voting just do 7=6=5=4>3=2=1 (of course the determination algorithm is different but this will capture the same amount of "information" from the voter). It may seem opaque (one can't determine the winner in <10 secs), but it is much better than simple majority voting for when there's >2 options (when there's 2 both give the same result). The other vote was "clear" only because it had 2 options and was therefore uninformative as it didn't include other options that people may care about. --Bequwτ 23:51, 27 February 2010 (UTC)
    The vote 7=6=5=4>3=2=1 does not capture the distinction between "I accept as good enough" and "I oppose": the vote 7=6=5=4>3=2=1 is wholly consistent with "I find 7, 6, 5, and 4 slightly preferable to 3, 2, and 1, but all are okay with me", which would lead a person to cast a support vote for each of the options, or to cast support to 7, 6, 5, and 4, while casting abstain to 3, 2, 1. --Dan Polansky 13:38, 1 March 2010 (UTC)
    Approval voting has no per-option abstentions since, being a single-winner voting method, abstain=disapproval. Obviously there are interpretational ambiguities when comparing votes across different systems, but the method I said has a "similar" (ie close but not exactly identical) marginal affect on the vote outcome, especially in this case where the "winner" is only valid if it "super-dominates" the status quo. --Bequwτ 18:06, 1 March 2010 (UTC)
  • Oppose this vote. It is the norm in Wiktionary votes that the status quo be the default, with a strong majority being required in order for any change to be made to it. Perhaps there are times when that norm should be set aside for a given vote; but if so, I think that must be discussed beforehand, rather than being tucked in as an implicit consequence of a complex voting scheme. The good news is, the Schulze method is easily converted to a Wiktionary-compatible form: step 1, determine the regular Schulze winner; step 2, if the status quo is not the regular Schulze winner, then perform an "instant runoff" to see whether the regular Schulze winner defeats the status quo with a strong majority. (Sorry for not speaking up before the vote began; I wasn't paying enough attention.) —RuakhTALK 20:01, 26 February 2010 (UTC)
    I was assuming when I started this vote that if the winner was not placed higher than option one in at least two thirds of the votes (or whatever the standard is), then it would be determined no consensus. The problem with the original ToC vote was that some people were voting against changing it because they had thought of a better version, even though they would prefer rhs toc over the current format. This fixes that, so that all options are available. If there isn't consensus to switch, then nothing happens, just like any regular vote. --Yair rand 20:18, 26 February 2010 (UTC)
    O.K., whew. Could you update the top part to make that explicit, then? :-)   —RuakhTALK 20:26, 26 February 2010 (UTC)
  • 4 > 6 > 5 > 1 > 7 > 3 > 2. I personally do not like left-alignment. It makes the list-item markers beside the ToC appear in the margin and misaligns the list below with the list beside, which I find atrocious. —Internoob (Disc.Cont.) 00:33, 27 February 2010 (UTC)
  • 1 > 3 > 7 and oppose all other options, which are horrid. --Nesksock kanetsv verify? 17:42, 1 March 2010 (UTC)
    • The way this vote is structured (and see above for others' comments thereon), a vote of "1 > 3 > 7 and oppose the rest" is the same as "1 > 3 > 7 > 2 = 4 = 5 = 6". Just fyi.​—msh210 17:45, 1 March 2010 (UTC)
      • Too much confusing when you start putting more symbols in it. The entire format is confusing, and I would prefer the usual way of our multioption voting. --Neskaya contribs talk? 22:23, 1 March 2010 (UTC)
  • 4 > 5 > 3 > 2 > 7 > 1 > 6  Content is king. It's important to show the reader an entry, and it's also important to point to their goal if they must scroll for it. Showing only navigation is poor. Michael Z. 2010-03-05 00:01 z
  • 5 DAVilla 20:26, 8 March 2010 (UTC)
  • 6 > 1 = 7 > 3 = 5 > 2 = 4 --Thrissel 22:25, 12 March 2010 (UTC)


  • To try out a ToC with only the L2 language headings, add the following to your skin CSS file
.ns-0 .toclevel-2,.ns-0 .toclevel-3,.ns-0 .toclevel-4,.ns-0 .toclevel-5,.ns-0 .tocnumber {display:none}

—This unsigned comment was added by Bequw (talkcontribs).

Please note that this vote is deciding what will be the default, so it would be best if the votes reflect what's best for unregistered users, not personal preferences. Any individual user can customize the ToC format to their preference by gadget, PREFS, or css file.

Also, to reiterate what Bequw said on the talk page, note that about 25% of views are of pages that, with high probability, will not have a definition on the initial screen. This number will only continue to rise as more words are added, severely limiting Wiktionary's ability to be useful as a dictionary.

One more thing: it is very likely that, should someone find a simple way to hide the toc when there's only one language for options 3 and 5, it would be viewed as an uncontroversial change, and could be added without another vote. --Yair rand 20:21, 1 March 2010 (UTC)