Since ProNews/2 doesn't work with Japanese encodings (yet, at any
rate), I've started reading sci.lang.japan using SeaMonkey Mail & News
(Peter's 1.0.1 unofficial build).
There's a mix of message encodings used; mostly ISJP, occasionally
EUC-JP, Shift_JIS, or UTF-8. It's the last one that give me trouble.
I cannot figure out how to persuade SeaMonkey to display UTF-8 messages
in anything other than Courier New. , I get a lot of question
marks in the messages.
Where is the setting to specify the fonts used for displaying UTF-8 messages?
I tried " Languages", but that doesn't do it.

Sun, 7 May 2006 15:24:31 UTC, "Alex Taylor" wrote:

Where is the setting to specify the fonts used for displaying UTF-8 messages?
I tried " Languages", but that doesn't do it.

That should be the right one. I am not sure if Mozilla checks if the
UTF-8 font is actually unicode. What fonts have you tried to set?

Alex Taylor wrote:
Since ProNews/2 doesn't work with Japanese encodings (yet, at any
rate), I've started reading sci.lang.japan using SeaMonkey Mail & News
(Peter's 1.0.1 unofficial build).

There's a mix of message encodings used; mostly ISJP, occasionally
EUC-JP, Shift_JIS, or UTF-8. It's the last one that give me trouble.

I cannot figure out how to persuade SeaMonkey to display UTF-8 messages
in anything other than Courier New. , I get a lot of question
marks in the messages.

IIRC, you also need Unicode fonts as well. The font needs a sufficient
character repertoire to display the extended character shapes (glyphs?).

Peter Weilbacher wrote:
Sun, 7 May 2006 15:24:31 UTC, "Alex Taylor" wrote:

>Where is the setting to specify the fonts used for displaying UTF-8 messages?
>I tried " Languages", but that doesn't do it.

That should be the right one. I am not sure if Mozilla checks if the
UTF-8 font is actually unicode. What fonts have you tried to set?

It's currently on 'Monotype Sans Duospace WT J' for monospaced text,
'Arial Unicode MS' for sans-serif, and 'Times New Roman WT J' for serif.

I think the only place I have 'Courier New' (which is what it seems to
be using) set is under 'Western'.

Alex Taylor wrote:

>IIRC, you also need Unicode fonts as well. The font needs a sufficient
>character repertoire to display the extended character shapes (glyphs?).

Yes, I have ' Languages' set to use Unicode fonts. The problem is,
SeaMonkey isn't _using_ those fonts. It seems to be using the 'Western'
font settings for UTF-8 messages, as far as I can see

No, no, I mean the font package itself, not the selection in the browser.
I am not even sure how to phrase what the requirements are. "Repertoire"
was one phrase. Basically it means the font package has sufficient
expressive power to display the extended characters.

Sun, 07 May 2006 23:56:04 -0700, James Moe
<jmm-listDESPAMMZ (AT) sohnen-moe (DOT) comwrote:
>Yes, I have ' Languages' set to use Unicode fonts. The problem is,
>SeaMonkey isn't _using_ those fonts. It seems to be using the 'Western'
>font settings for UTF-8 messages, as far as I can see
>
No, no, I mean the font package itself, not the selection in the browser.
I am not even sure how to phrase what the requirements are. "Repertoire"
was one phrase. Basically it means the font package has sufficient
expressive power to display the extended characters.

I know for a fact that the fonts contain the required characters, if
that's what you mean.

Alex Taylor wrote:
>>
>No, no, I mean the font package itself, not the selection in the browser.
>I am not even sure how to phrase what the requirements are. "Repertoire"
>was one phrase. Basically it means the font package has sufficient
>expressive power to display the extended characters.

I know for a fact that the fonts contain the required characters, if
that's what you mean.

Yes. So much for that idea, then.

/Alex Taylor/:

Yes, I have ' Languages' set to use Unicode fonts. The problem is,
SeaMonkey isn't _using_ those fonts. It seems to be using the 'Western'
font settings for UTF-8 messages, as far as I can see

As far as I remember there was an "Unicode" category in the previous
Mozilla versions but I guess that was dropped because it was
initially meant for use for "Unicode" encodings, i.e. UTF variants
but no matter how a character is encoded it is the same character.
Then just the character category is important, so you set specific
font to display Latin letters, you set a font preference for
Cyrillic letters, etc. " Languages" should map to characters in
other than the specified categories. I don't know what the "User
Defined" category means but there's such option to select for
character encoding for messages and documents.

/Stanimir Stamenkov/:

no matter
how a character is encoded it is the same character. Then just the
character category is important, so you set specific font to display
Latin letters, you set a font preference for Cyrillic letters, etc.

Seems like the font categories don't denote specific character
ranges but affect content marked up to be in specific language (like
using 'lang' attribute in HTML). When there's no language specified
for the content (like may be in plain text messages) the text
encoding is used as hint for the language in use. If the text
encoding is some UTF variant the default language (Preferences -
Navigator -Languages) or the language assumed from the default
encoding is used for the font preferences to use. This is only like
that as far as I've tested and I've made some guesses.

/Alex Taylor/:

There's a mix of message encodings used; mostly ISJP, occasionally
EUC-JP, Shift_JIS, or UTF-8. It's the last one that give me trouble.

I cannot figure out how to persuade SeaMonkey to display UTF-8 messages
in anything other than Courier New. , I get a lot of question
marks in the messages.

Could you forward here (as attachments) some of those messages
showing lots of question marks in them?

/Stanimir Stamenkov/:
/Alex Taylor/:

>There's a mix of message encodings used; mostly ISJP, occasionally
>EUC-JP, Shift_JIS, or UTF-8. It's the last one that give me trouble.
>>
>I cannot figure out how to persuade SeaMonkey to display UTF-8 messages
>in anything other than Courier New. , I get a lot of
>question marks in the messages.

Could you forward here (as attachments) some of those messages showing
lots of question marks in them?

Generally if a font don't have a glyph to displaying some character
it should be displayed as square box not a question mark, but
Mozilla has quite good font substitution engine and should display
the character if there's an installed font containing the glyph,
although the font may differ in appearance from the current one.

If you see ordinary question marks, that should mean those
characters where already damaged by the sender, e.g. the client
haven't been able to encode the characters using the target encoding
(which should not be a case using an UTF variant) and replaced them
with question marks. If you see (that should be U+FFFD REPLACEMENT
CHARACTER [1]) it is because the sender has generated an invalid
byte sequence or sequence representing invalid Unicode character
(not all code points represent valid Unicode characters).

[1]

/yo/:

>Generally if a font don't have a glyph to displaying some character it
>should be displayed as square box not a question mark, but Mozilla has
>quite good font substitution engine and should display the character
>if there's an installed font containing the glyph, although the font
>may differ in appearance from the current one.

Maybe I don't have a proper glyph for UTF-8. I get bunch of squares for
UTF-8 pages. (My mozilla 1.7.13 Unicode is set to Times New Roman MT30)
on eCS 1.2.

It is not only the glyphs but the presence of correct Unicode
tables, like cases such as the Windings font on Windows:

It's funny that I can read the same page correctly on Warp4 Japanese
with Mozilla 1.7.13 which have exactly the same font settings.

here is an example
%3A%2F%2Fmail.google.com%2Fmail%2F%3Fui%3Dhtml%26z y%

I must be missing something just like Alex.

I can't really reach the content at the URL you've pointed as it
seems private to your Gmail account.

/Steve Wendt/:
Stanimir Stamenkov wrote:

>Generally if a font don't have a glyph to displaying some character it
>should be displayed as square box not a question mark,

I think that is a Win32 vs S/2 difference.

But I think Mozilla should behave uniformly across platforms in this
regard and I've verified it replaces (on Windows) characters which
could not be displayed with question marks and not square boxes, indeed.

Mon, 8 May 2006 01:15:19 UTC, Alex Taylor wrote:

Peter Weilbacher wrote:
That should be the right one. I am not sure if Mozilla checks if the
UTF-8 font is actually unicode. What fonts have you tried to set?

It's currently on 'Monotype Sans Duospace WT J' for monospaced text,
'Arial Unicode MS' for sans-serif, and 'Times New Roman WT J' for serif.

Similar to what I have set. Can you post some example MsgIDs for
messages where it works and where not?

Sun, 14 May 2006 15:58:28 UTC, "Peter Weilbacher" <newsspam (AT) weilbacher (DOT) orgwrote:

That should be the right one. I am not sure if Mozilla checks if the
UTF-8 font is actually unicode. What fonts have you tried to set?

It's currently on 'Monotype Sans Duospace WT J' for monospaced text,
'Arial Unicode MS' for sans-serif, and 'Times New Roman WT J' for serif.

Similar to what I have set. Can you post some example MsgIDs for
messages where it works and where not?

>From sci.lang.japan:
<e3v6oe$5b8$1 (AT) news-est (DOT) ocn.ad.jp>
<4cbmc9F156215U1 (AT) individual (DOT) net>

see the attachments I posted in my previous message in this thread
As I mentioned, it works (strangely enough) if I View Message Source,
just not in the message reader itself.

Sun, 14 May 2006 16:25:13 UTC, "Alex Taylor" wrote:

From sci.lang.japan:
<e3v6oe$5b8$1 (AT) news-est (DOT) ocn.ad.jp>
<4cbmc9F156215U1 (AT) individual (DOT) net>

see the attachments I posted in my previous message in this thread
As I mentioned, it works (strangely enough) if I View Message Source,
just not in the message reader itself.

What is View -Character Encoding set to when you view the messages?
For me it is "Unicode (UTF-8)" and the characters seem to display just
fine. Is there a difference in this between message display and view
source? You could try to experiment with the possibilites under "Auto
Detect" (although for me it works with both "()" and "Universal").

Sun, 14 May 2006 20:31:00 UTC, "Peter Weilbacher" <newsspam (AT) weilbacher (DOT) orgwrote:

From sci.lang.japan:
<e3v6oe$5b8$1 (AT) news-est (DOT) ocn.ad.jp>
<4cbmc9F156215U1 (AT) individual (DOT) net>

Incidentally, how do I convert these into useable links?

see the attachments I posted in my previous message in this thread
As I mentioned, it works (strangely enough) if I View Message Source,
just not in the message reader itself.

What is View -Character Encoding set to when you view the messages?
For me it is "Unicode (UTF-8)" and the characters seem to display just
fine. Is there a difference in this between message display and view
source? You could try to experiment with the possibilites under "Auto
Detect" (although for me it works with both "()" and "Universal").

It shows "Unicode (UTF-8)" here as well, in both places.

Changing the value of "auto detect" has no discernable effect.

Wed, 17 May 2006 02:16:48 UTC, "Alex Taylor" wrote:

Sun, 14 May 2006 20:31:00 UTC, "Peter Weilbacher" <newsspam (AT) weilbacher (DOT) orgwrote:

From sci.lang.japan:
<e3v6oe$5b8$1 (AT) news-est (DOT) ocn.ad.jp>
<4cbmc9F156215U1 (AT) individual (DOT) net>

Incidentally, how do I convert these into useable links?

You can install the MessageIDFinder extension from
<>.

What is View -Character Encoding set to when you view the messages?
For me it is "Unicode (UTF-8)" and the characters seem to display just
fine. Is there a difference in this between message display and view
source? You could try to experiment with the possibilites under "Auto
Detect" (although for me it works with both "()" and "Universal").

It shows "Unicode (UTF-8)" here as well, in both places.

Changing the value of "auto detect" has no discernable effect.

Would have been too easy. Perhaps the environment can affect this, so
what is LANG set to and what codepage do you use?

Tue, 16 May 2006 23:32:36 -0500, Peter Weilbacher
<newsspam (AT) weilbacher (DOT) orgwrote:
><e3v6oe$5b8$1 (AT) news-est (DOT) ocn.ad.jp>
><4cbmc9F156215U1 (AT) individual (DOT) net>
>
>Incidentally, how do I convert these into useable links?
>
You can install the MessageIDFinder extension from
<>.

Interesting. Thanks!

>What is View -Character Encoding set to when you view the messages?
>Is there a difference in this between message display and view
>source?
>
>It shows "Unicode (UTF-8)" here as well, in both places.
>Changing the value of "auto detect" has no discernable effect.
>
Would have been too easy. Perhaps the environment can affect this, so
what is LANG set to and what codepage do you use?

My LANG is en_CA, and I run under codepage 850.

Wed, 17 May 2006 14:29:04 UTC, Alex Taylor wrote:

My LANG is en_CA, and I run under codepage 850.

Just tried that and in general doesn't seem to make a difference. _But_:
the very first article I looked at with LANG=en_CA was also displayed
with gibberish instead of Japanese characters. I tried to switch back
and forward but it never happened again, so way to debug. :-( For now, I
am out of ideas.