Emoji [And crash in the Web interface to the mailing list]

Philippe Verdy verdy_p at wanadoo.fr
Thu Apr 3 02:25:15 CDT 2014

But then why did I these emoticons (using "ASCII art" with additional CJK
characters) correctly from the same mailing list ?
I did not see any dot, or squares for PUA, but the stars, triangle and some

I don't think that the mailing list itself is broken (or may be Gmail
corrected things after).

You cannot always predict the encoding used by the effective sending
mailing agent (or the first reaying agent). Even Gmail will try to fit the
"best" (popular) legacy 8-bit encoding according to content and the
recipient (where it will try to geolocalize the target domain name, or use
its own knowledge of the languages and encodings most often used by senders
in that domain), instead of always sending with UTF-8, when the default
user settings are for using a "default" encoding if the user had not
specified that the mail would be forced to UTF-8.

2014-04-03 5:18 GMT+02:00 Doug Ewell <doug at ewellic.org>:

>  It's really quite simple: Sending e-mails in ISO-2022-JP to the Unicode
> mailing list causes problems.
> ��
> --
> Doug Ewell | Thornton, CO, USA
> http://ewellic.org | @DougEwell
>  ------------------------------
> From: Philippe Verdy <verdy_p at wanadoo.fr>
> Sent: ‎4/‎2/‎2014 20:51
> To: Doug Ewell <doug at ewellic.org>
> Cc: Ilya Zakharevich <nospam-abuse at ilyaz.org>; Unicode Mailing List<unicode at unicode.org>
> Subject: Re: Emoji [And crash in the Web interface to the mailing list]
> There was no such browser bug in my Chrome install GChrome (in Gmail)which
> rendered the full string correctly (no dots, all characters displayed
> properly).
> So this looks like a Firefox bug. There's a rendering problem in IE, but
> no such critical bug that breaks the rest of the page. It looks like a
> problem of transcoding of emails by browsers (in Gmail, the transcoding to
> UTF-8 is performed apparently by the web server, so Firefox does not break
> when rendering the Gmail page).
> It is posible that what is really broken is in fact another webmail
> interface that incorrectly transcodes the email to UTF-8. I cannot know if
> the broken rendering was performed on everyone's client, if he use a
> webmail or standalone email agent. Many webmail services are broken in how
> they transcode the mails received in order to embed its content in a UTF-8
> web page.
> 2014-04-02 22:31 GMT+02:00 Doug Ewell <doug at ewellic.org>:
>> Sorry, this was my mistake. IE8 on Windows 7 displayed James's "angry"
>> line like this:
>> ヾ(@゜・・・・逅察・・縫吏・€ォ€趁占コ碣€ュ処腫徐ぢヽ(#`Д´)ノ
>> - angry
>> The only "private-use" character was something that got transcoded to
>> U+E559, and IE8 displayed that as a space, not a dot. But a quick look
>> at the ISO-2022-JP source shows this isn't right at all. So I guess I
>> did have trouble viewing it, maybe not a crash, but severe mojibake.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://unicode.org/pipermail/unicode/attachments/20140403/6307b567/attachment.html>

More information about the Unicode mailing list