- From: Leif Halvard Silli <xn--mlform-iua@xn--mlform-iua.no>
- Date: Thu, 10 Jan 2013 13:39:58 +0100
- To: Asmus Freytag <asmusf@ix.netcom.com>
- Cc: John Cowan <cowan@mercury.ccil.org>, Gunnar Bittersmann <gunnar@bittersmann.de>, www-international@w3.org
Asmus Freytag, Thu, 10 Jan 2013 03:51:11 -0800: > On 1/10/2013 3:00 AM, Leif Halvard Silli wrote: >> Asmus Freytag, Thu, 10 Jan 2013 02:13:34 -0800: >>> On 1/9/2013 4:19 AM, Leif Halvard Silli wrote: >>> And, it is riddled with side effects. Just because I am fluent in >>> some other language doesn't mean that I would like to be shuffled off >>> to the localized pages (and local subsite) of some global website by >>> default. I might want to occasionally visit such pages, but only when >>> I'm interested in some service (like media in that language) that I >>> can't access from the US site. >> I tried to guess if it was Swedish of German that was your mother >> tongue - or whatever. ;-) > > A guess from my name? Think so. (Did the guessing last year.) > I'm a native German speaker but fluent reader of Swedish (with a > passionate aversion against Svenglish). So no 'websajt' and 'mejla' on you then. ;-) >> But, sorry: have you tried to a) define what your preferred language >> is, and b) live with that preference to see how browsers behave - >> for a while? > > I have done that in the past. I had to give it up, or rather, I was > so annoyed with it that when I switched systems, I never bothered to > make a language seleciton again So the annoyance was linked to search engines? > Now, when I search for German or Swedish terms, I get pointed to > English documents for them. Before, I would get Swedish or German > documents when looking for English terms. You mean, when you use Bing or Google? Or when you click on the documents they provide? I wonder if you talk about something else - or side use, of the language preference tags. What I have in mind is the situation when you click on URL-A and get language y instead of the default language x. >> I.e. do you describe a real world problem? Take for instance >> Google.no, which I mentioned (and Google.com >> should work the same): The language negotiation effect can be >> overridden via cookies. So if you prefer the US English version of >> Google, you get that version, even if your browser prefers German, >> as long as you have made that choice in the Google settings/cookies. > > Or, for some vendors, you can actually register on both a local and > the global (US) site. If you do that, you can access both, no matter > where you are, if you don't, they force you to the local site (even > without language preference set, but as user of a US version of > OS/Browser). (The latter example is a recent observation). (I must travel more, to check this out …) >>> Location based defaults work equally poorly, because I'm more likely >>> to want to conduct business via my "home base" than locally, >>> especially when traveling. That's why I think the expectation of >>> being able to get automatic configuration to work for any >>> multilingual or multi-location users is a pipe dream - and sites that >>> don't supply easy overrides of their automatic choices are a pain to >>> use. >> It is not clear to me what consequences you draw from this - you are >> opposed to language negotiation? > > I would love language negotiation, if it were easy to correctly > configure my language preferences. In my case, they are not "static" > or "absolute", but highly context dependent. I suspect that is the > case for many people who are fully or nearly bi- or tri- lingual. > > In contrast to people who have a clear first and decidedly weaker > second language, perhaps followed by a "fall-back" language. I guess my expectations regarding what language based content negotiation can achieve are technically more modest. But (and sorry for mentioning FF again) Firefox allows you to, very easily, use several user profiles. You could create one user profile for each context. But otherwise it sounds like a good thing if vendors could allow us to define profiles of our own user profiles … >> In principle, language negotiation could help you operate in your >> "home base" language. E.g. I can go to the US and still get Google >> in Nynorsk. > > As US English is a common default or fallback language it seems to > not be as "marked" as some other choices. That's just a hunch, but > it would explain why it doesn't seem to have the same 'override' > effect (at least as far as I remember from the times when I bothered > with setting preferences). > > What I meant is the fact that the location-based service tends to > route people to the local site, no matter what. This happens with > nearly all services that have world-wide (and local) presence, so > it's not enough to set a google cookie. > > Cookie based sticky choice are in principle useful, but I've never > successfully had any of them survive a system upgrade. That's why browsers’ language settings are useful, I believe. They don't go away when the cookie is deleted. >> I believe language negotiation is *very* useful for providing defaults. > > It is very helpful for the situation where someone has a strong > preference for a single primary language, and perhaps also where > someone has a secondary language that's not English but which is > common enough that some services offer it as translation. Let me see … My "primary" language is Norwegian Nynorsk. My secondary language is Norwegian Bokmål - and e.g. Google Translate offers translation to Bokmål (but not Nynorsk). And, thirdly, I have never had trouble getting stuff in English — that happens "by itself". So may be I fall into that category - I don't know. (But it is a bit hard to follow the scenarios you describe.) Linking back to Henri’s Wiki article: Finnish is is native tongue, I believe. But Finland also have native Swedish speakers as a minority. I suppose that those users - if the know how it works, would be be more keen on language negotiation as this allows them to get things in Swedish whenever it is available. Thus, they avoid having to click a button to get the information in Swedish. -- leif halvard silli
Received on Thursday, 10 January 2013 12:40:25 UTC