Elian:
I have nine browsers on my computer (needed for webdesign and some other purposes). Some don't even have a localize-option and do you think I bother to configure the rest? So I get english all the time.
I would think that you could at least bother to configure your main browser, if you have one. If you don't, you're very far from the typical user.
Also:
My mother doesn't even know that her browser has a localize-option and she doesn't speak english and - yes, she thinks also that webadresses always start with www. However, she would be a great contributor to the German wikipedia.
Her browser should respect the operating system's locale settings and not require her to configure anything. I don't use IE/Windows so I can't check, but if it doesn't do that, it's broken.
Besides, in both cases, I really don't see what the big deal is. You can still click the language link which is *clearly visible* on the wikipedia frontpage and bookmark your language Wikipedia. I think a state of things where Wikipedia immediately works for those who have configured their browsers properly and requires them to click a link for those who haven't is completely acceptable. Your start page, on the other hand, requires *every* user to click an entrypage link. And someone who so insists on the laziness of users should know that many users will immediately close the Wikipedia browser window the moment they notice they still have to click another time to get somewhere.
Esperanto is not available as language option in all of my browsers.
I have just checked, it's available in both Mozilla and Opera. It's not available in IE by default, but you can add it by simply adding "eo" to your language list. Besides, if you speak Espearanto, I think you would be intellectually independent enough to use a non-MS browser.
De facto, by depending only on browser settings, we would deprive the Esperanto wikipedia of many possible contributors
Those people who have not set their browser to prefer Esperanto, which they can with all major browsers, can still easily click the Esperanto link. Your extra page adds absolutely nothing to the user experience.
Next, I want to see the progress of all wikipedias at one central place and not somewhere hidden at an obscure statistics page somewhere at the english (Main?) wikipedia where nobody of the non-english contributors bothers to look for regularly.
Then create a translation of this page for your Wikipedia of choice. You'll have to do that anyway, or do you want the "central place" information to be English? I agree that the link (which would read "About the non-English Wikipedias" on the English page, "..non-German.." on the German page etc.) could be placed more prominently, perhaps right below the language bar.
Lastly, I want to have a simple, not crowded page to search wikipedia where I am not disturbed by masses of text and I have to look very sharply to discover the small "search" field in one corner.
The search field is both in the upper right and lower left, which is good design from a usability perspective because it makes it likely that the eye notices it. I can't remember having ever searched for the search because I intuitively looked in that place. Since we already have lots of links in the sidebar, I wouldn't recommend adding another one there, but if you insist, that would be the best place to put it. This link could lead to a more formal search page.
Besides, I can think of very very few instances where I would want to search language-neutrally, and many many cases where this would be highly annoying. So the multi-language search, if we need it at all, should not be the default.
I think a wikipedia, each language on a separate URL without a centralized
- and a really centralized - main page is a bundle of balkanized
projects.
By the same argument, if I translate my website into many different languages and redirect the reader to his language by default, I create a balkanized website. That doesn't make any sense. Balkanization comes from lack of communication between the different language maintainers and contributors, lack of consistent policies, use of different software etc. We're making good progress in that area.
Regards, Erik