Wiktionary talk:Votes/2008-01/Homophones section

Further discussion

Assuming that the archived discussion is "dead", I hope this is the right place for further discussion. Please consider the value of getting as much information as possible to the user on the first screen seen. Making homophones a header "costs" about two (2) lines more than the current WT:ELE approach if there is only one line for the homophones and one additional line for each homophone if the generic approach for derived terms, related terms, and see also is used. That's a lot of prime real estate. To use it without much knowledge of the effect on average WT users in the service of conceptual consistency seems unwise. DCDuring TALK 22:56, 27 January 2008 (UTC)

Exactly. Having it as a header will make it show up in the TOC, which it currently does not do. Homophones are of great importance to English learners, so the header will draw attention to the fact that homophones exist, which the current format does not do. And the header is already in widespread use. The vote is making practice official. Related terms and Derived terms may also have only one line below them, but Wiktionary is not paper. In any case, fail to see how conceptual consistency will negatively affect the average user. If anything, it should help. --EncycloPetey 23:11, 27 January 2008 (UTC)

Consider the anonymous comment posted earlier today at Talk:discrete. The user wondered why discrete wasn't defined as "lay low or unnoticable". That is the definition of (deprecated template usage) discreet, which is listed on the page as a Homophone, but in the discreet fashion currently advocated by ELE. If there had been a separate Homophones section, the other word would have been more obvious and the user would not have been confused as he was. --EncycloPetey 23:39, 27 January 2008 (UTC)

You mean, if there had been a discrete Homophones section? :-) —Ruakh_TALK 00:00, 28 January 2008 (UTC)

Yes, (deprecated template usage) discrete should have had a discrete Homophones section instead of a discreet one. --EncycloPetey 00:01, 28 January 2008 (UTC)

I got that, really. As a new person here I don't have a good feel for the frequency of various kinds of user interactions, who are likely to be loyal users, how easily users are discouraged by lack of visibility, what information users are looking for, conversion rate of first-time users to repeat users, what kind of users we want and can get, etc. That is a reason why I wish that there were some numbers or tested guidelines that we could use to make our design decisions. The considerations that you mention seem valid and I must defer to the judgment of the more experienced in the absence of any facts and figures. Looking at the other pronunciation/real estate issue: the extra verticle space taken by the Rhymes template, I see that the value of initial screen real estate is not a consideration taken as very important. If that reflects the experience of the seniors here with what makes a difference to users, I defer to it. If not, I beseech you to economize on the portion of first-screen real estate devoted to pronunciation. DCDuring TALK 00:08, 28 January 2008 (UTC)

We do economize sometimes. For example, we list the enPR, IPA, and SAMPA for a given "accent" on a single line, rather than sequentially. If there are multiple audio files for the same region, they are placed on a single line. However, there are some situations where economy must give way to clarity, such as the monstrous pronunciation section for (deprecated template usage) hello, where the pronunciation affects the meaning, or for (deprecated template usage) the, which has both stressed and unstressed forms. --EncycloPetey 00:15, 28 January 2008 (UTC)

I can't tell how important the space consideration really is. It depends on the type of entry. For both short and long entries my consideration is of no or little importance. Short entries can be defined as those with not enough content to overfill the initial screen. There are vast numbers of them, even restricting ourselves just to lemmas. There are also many entries that are so long that not even the first inflection line appears on the first page. Only some of the remainder really benefit much from the space-economizing considerations. A Pronunciation section with the current rhyme layout and a homophone section could use 2 inches/50mm of the total available vertical space of 5 1/2 in/140mm - more than a third. It is a much larger portion of the space after the table of contents. To reduce that by 1/2 inch has a noticeable (not huge) effect on how much else can appear on the first screen. Do we have any estimates of how many of each there are (and will be) and how many non-editing views each type gets? (Unfair rhetorical question.) I don't want the discussion to bog down on this consideration or to waste anyone's time trying to get facts that are not readily available. DCDuring TALK 00:58, 28 January 2008 (UTC)

Consider that most words with homophones will be short words. These either have a short entry with one or two meanings (like (deprecated template usage) rite), or have longer entries with multiple meanings ((deprecated template usage) right). Further, we have no control over the font size, screen size, or type of browser used by our clientelle. I don't think a general statement can be made. --EncycloPetey 01:01, 28 January 2008 (UTC)

This strikes me as a better reason to consider moving the pronunciation section down than to try cramming it into a small space. (By the way, I don't see why you assume that the more experienced of us know better how newcomers perceive things; if anything, I'd have expected the opposite!) —Ruakh_TALK 01:08, 28 January 2008 (UTC)

The placement of a whole header for a user group I don't understand (those who need the pronunciation) is a big issue to tackle. I hope for a technical solution analogous to the rel and trans bars, which would moot the whole discussion I suppose. I keep hoping that someone once saw some good user statistics. Or that someone has been able to learn enough from a large number of anecdotal contacts with users to grok a pattern. I know am not a good model of most kinds of users: non-native speaking users; casual users; young users. At best I can use my initial experience as a model of a want-to-be contributor whose errors attracted the attention of WT's protectors. DCDuring TALK 01:42, 28 January 2008 (UTC)