Vitaly V. dr.vitall@gmail.com wrote:
Hmm? No, that would be silly. The only language that should get that priority in any kind of disambiguation scheme is the language of the wiki.
What? There is no "disambiguation scheme" in this proposal. Look at example http://uk.wiktionary.org/wiki/November provided by Анатолій Гончаров in original post.
...which is a page which apparently (for whatever reason) considers the English article primary, and has perfectly ordinary disambiguation links to [[November/la]] and [[November/de]]--no different than what you might find at the top of a wikipedia page--only displayed in a fancy simplified format instead of with the usual text which would be along the lines of "This page is about the English word 'November'. For the Latin word, see [[November/la]]. For the German word, see [[November/de]]."
This is entirely different subject. I feel against creating millions of disambiguation page, but this is something for different discussion.
Against millions of disambiguation pages when there are millions of things to disambiguate against? Could you sell that viewpoint on any wiki? Why this one?
As Wiktionary grows, there will be *fewer* like that.
How come? Have you ever paid attention how many words defined in English or French Wiktionary? Less then a million! And that is best Wiktionaries out there. In Russian language alone there are roughly one and a half million words. And there is absolutely nothing special about Russian language. Estimations for Ukrainian or Belorussian would be similar. With very low overlap - even alphabets are different. And there is thousands of languages out there.
Overlap between languages doesn't matter for the particular point that remark was about. There will be millions of articles, certainly. But the remark was about seeing all the information at once after searching for the word, which is irrelevant to the number of languages on it: as Wiktionary grows, most pages will need to have all their separate sections for homographs, pronunciations, alternative spellings, etymologies, parts of speech, declension or conjugation tables, associated terms of various kinds, derived terms, etc., depending on the format of the wiktionary. As I mentioned with the examples, some of those had *tables of contents* more than a screen high even with only one or two languages on the page, and this is supposed to be normal: when the pages are no longer stubs, you will not have the situation where "information is shown directly on [your] screen after search".
they are by definition not interested in anything else we feel like showing them.
I don't know who define what I'm interested in, as a user. And I'm a user indeed. And I rather have an easy access to information about word that interests me, other then have it hidden behind interface.
Exactly the point. If the user wants easy access to the word of interest to them, that pretty much excludes having several other words *not* of interest to them presented at the same time just because by some historical accident they happen to be spelled with the same letters. The irrelevant data, namely the words the user is not interested in, should not be part of the interface hiding the word that does interest them, but so long as the words are all on the same page, that is the state of affairs.
And if a user happens to be interested in words that have the same spelling, they can click through to find them just as people click through to reach words with the same rhyme scheme in pronunciation sections, or as they click through to reach words with the same meaning in translation sections, or as they click through to reach words with the same root in etymology or derivation sections. Is there a particular reason why words with the same spelling should be accorded special preference?
*Muke!