Re: [css3-ui] Issues 40, 72, 75, 80 ime-mode section shouldn't say "Implementations should drop support for it as soon as possible."

On Fri, Feb 27, 2015 at 4:41 AM, Florian Rivoal <florian@rivoal.net> wrote:
>
>> On 27 Feb 2015, at 15:11, Masayuki Nakano <masayuki@d-toybox.com> wrote:
>>
>> Hello,
>>
>> First, I don't disagree with dropping ime-mode from CSS3-UI. As I said in 2012, I still believe that ime-mode shouldn't be standardized due to serious UX problems and bad dependence with platforms and IMEs)
>> http://lists.w3.org/Archives/Public/www-style/2012Feb/0959.html

Thanks Masayuki.


>> However, CSS3-UI spec shouldn't say, "Implementations should drop support for it as soon as possible".

Yes the working group agrees with this request per consensus
resolution to Issue 72.


>> Most ime-mode users, especially enterprise system developers, must be afraid browsers to drop ime-mode by this. Because if ime-mode were dropped from browsers actually, their system would become inconvenience than today.
>>
>> I believe that ime-mode is still useful for internet applications in Japan. Therefore, browsers shouldn't drop such users without providing alternative API or CSS propery. If browsers would drop ime-mode under current situation, somebody couldn't believe web browsers as stable platform.
>>
>> So, I think that the spec should say "Implementation may keep supporting it *only* for backward compatibility until alternative feature is standardized".

After the changes that went into the latest public working draft of
CSS3-UI, there was additional discussion in the group, and a
resolution to use "should not" rather than "drop" which was captured
as Issue 72.

Thus the issue you raise about browsers shouldn't drop such users I
believe has already been addressed compatible with your request in the
resolution of issue 72.

https://www.w3.org/wiki/CSS3-UI#Issue_72


<snip> re: lang, inputmode, pattern, type HTML features

> There is a standing issue against css3-ui (https://wiki.csswg.org/spec/css3-ui#issue-75) to refer to these in the section about ime-mode.

I have also added added an informative note referencing these HTML features.


> I do agree though that some more work in this area is needed before we can consider that a fully functional replacement to ime-mode is available, but I think it is important to point people to these approaches, especially since they need more attention.

To put it another way, we know ime-mode is a bad feature, but we have
too little information to suggest a complete replacement for it.

If you have suggestions for a better replacement, please bring them to
the group.


> To keep the essence of your message, while at the same time avoiding suggesting that there is nothing in the web platform to address this need, I would suggest the following phrasing:
>
> "Implementation may keep supporting it *only* for backward compatibility until alternative features are sufficiently developed and supported"

I'm not sure we can make such demands on the future of hopeful
alternative features.

I believe the previously group resolved (per Issue 72) language that
says "User Agents should not support the ime-mode property" provides
sufficient leeway (per should not) for implementations that feel they
must continue supporting it (e.g. for intranet web compat) to do so,
while indicating the working group's technical consensus.

Thanks,

Tantek

Received on Wednesday, 4 March 2015 05:05:27 UTC