W3C home > Mailing lists > Public > whatwg@whatwg.org > January 2005

[whatwg] For Japanese language users, Web Forms should have a way to active/deactive for IME

From: Henri Sivonen <hsivonen@iki.fi>
Date: Tue, 18 Jan 2005 23:47:55 +0200
Message-ID: <9C5799FE-699A-11D9-80E1-003065B8CF0E@iki.fi>
On Jan 18, 2005, at 09:41, Hallvord Reiar Michaelsen Steen wrote:

> Unlike us, Japanese users know what they are doing when using an IME
> :)

Yes, but asking questions (perhaps na?ve) from the outside may 
sometimes be helpful and may actually uncover something that is not 
thought about because the close stakeholders are taking things for 

I think it is appropriate to ask for use cases just like use cases are 
called for with other proposals. It may look politically incorrect, but 
I don't think "i18n issues" should be exempt of the scrutiny other 
proposals are subjected to.

An example of lack of critical thinking in i18n: ISO-8859-15 was 
created to address issues related to Finnish and French. I am a native 
Finnish speaker, but I still think ISO-8859-15 was a bad partial 
solution to the supposed problem and actually made the problem worse. 
Because of i18n hypersensitivities, many Linux distros took ISO-8859-15 
seriously and actually shipped with an ISO-8859-15 default for Finnish 
instead of doing the right thing (UTF-8 for every locale) right away.

And yes, I have been on the other side as well. I think the lack of an 
"en-dash" list-style-type in CSS forces foreign bulleted list 
typography on Finns. Still, I was unable to convince the editor of the 
relevant spec that such a list-style-type should be added.

To get back to this case, fantasai already pointed to a generalization 
that isn't limited to the Japanese-specific point of view.

> It is actually very annoying to, say, submit double-width characters
> to a server that expects ASCII.

Point taken. However, I do consider such a server non-robust. I also 
consider a server that can't deal with decomposed umlauts non-robust, 
although I recognize the need to cater for such inevitable lack of 
robustness by normalizing in the browser.

Henri Sivonen
hsivonen at iki.fi
Received on Tuesday, 18 January 2005 13:47:55 UTC

This archive was generated by hypermail 2.3.1 : Monday, 13 April 2015 23:08:20 UTC