Misplaced umlauts?

Check out the umlauts here:

<https://www.nytimes.com/2019/07/10/science/skull-neanderthal-human-europe-greece.html>

For some reason, SM displays them to the right of the target letter, 
although the source code displays them correctly. It's because the 
author has used "u" plus the combining umlaut rather than the single 
letter "ü," thus: "ü."

So "the University of Tübingen" appears as "the University of 
Tu¨bingen," and "Rainer Grün" appears as "Rainer Gru¨n."

I'm not using an odd font here, and they don't specify one. It's 
plain-vanilla Times Roman.

Copy/pasting the source code into Notepad, MS Word, or other 
applications shows that it displays correctly there. Even copy/pasting 
the text from the badly displayed browser window shows that it displays 
correctly elsewhere. Only in an SM browser window is the umlaut 
misplaced. No, wait, IE 11 has the same problem.

Any idea how to fix this?

-- 
War doesn't determine who's right, just who's left.
--
Paul B. Gallagher

0
Paul
7/11/2019 5:08:02 AM
mozilla.support.seamonkey 13151 articles. 0 followers. Post Follow

16 Replies
25 Views

Similar Articles

[PageSpeed] 14

Paul B. Gallagher wrote on 11/07/2019 3:08 PM:
> Check out the umlauts here:
> 
> <https://www.nytimes.com/2019/07/10/science/skull-neanderthal-human-europe-greece.html> 
> 
> 
> For some reason, SM displays them to the right of the target letter, 
> although the source code displays them correctly. It's because the 
> author has used "u" plus the combining umlaut rather than the single 
> letter "ü," thus: "ü."
> 
> So "the University of Tübingen" appears as "the University of 
> Tu¨bingen," and "Rainer Grün" appears as "Rainer Gru¨n."
> 
> I'm not using an odd font here, and they don't specify one. It's 
> plain-vanilla Times Roman.
> 
> Copy/pasting the source code into Notepad, MS Word, or other 
> applications shows that it displays correctly there. Even copy/pasting 
> the text from the badly displayed browser window shows that it displays 
> correctly elsewhere. Only in an SM browser window is the umlaut 
> misplaced. No, wait, IE 11 has the same problem.
> 
> Any idea how to fix this?
> 
Paul, I'm using SM 2.49.5 and both 'Tübingen' and 'Grün' looked fine to 
me .... mind you, I had to concentrate hard to see the umlauts as 
separate/above from the 'u'. ;-) .... and my spellchecker was 
complaining about the umlauts as well!

-- 
Daniel

User agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:52.0) Gecko/20100101 
SeaMonkey/2.49.1 Build identifier: 20171016030418
or
User agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:52.0) Gecko/20100101 
SeaMonkey/2.49.5 Build identifier: 20190609032134

User agent: Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101 
SeaMonkey/2.49.1 Build identifier: 20171015235623
0
Daniel
7/11/2019 6:14:56 AM
Paul B. Gallagher schrieb:
> Check out the umlauts here:
> 
> <https://www.nytimes.com/2019/07/10/science/skull-neanderthal-human-europe-greece.html> 
> 
> 
> For some reason, SM displays them to the right of the target letter, 
> although the source code displays them correctly. It's because the 
> author has used "u" plus the combining umlaut rather than the single 
> letter "ü," thus: "ü."
> 
> So "the University of Tübingen" appears as "the University of 
> Tu¨bingen," and "Rainer Grün" appears as "Rainer Gru¨n."
> 
> I'm not using an odd font here, and they don't specify one. It's 
> plain-vanilla Times Roman.
> 
> Copy/pasting the source code into Notepad, MS Word, or other 
> applications shows that it displays correctly there. Even copy/pasting 
> the text from the badly displayed browser window shows that it displays 
> correctly elsewhere. Only in an SM browser window is the umlaut 
> misplaced. No, wait, IE 11 has the same problem.
> 
> Any idea how to fix this?
> 

Hi Paul,

no such problems here with SeaMonkey 2.49.4 (Linux) and no idea how to 
fix it.

Saludos
Wolf


-- 
Linux Mint 18.3 Sylvia, MATE, 64-bit
Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101 Firefox/52.0 
SeaMonkey/2.49.4
Lightning-5.4 -sm+tb
0
Wolf
7/11/2019 6:26:28 AM
On 11.07.2019 07:08, Paul B. Gallagher wrote:
> So "the University of Tübingen" appears as "the University of Tu¨bingen," 

Hi Paul,

NOT reproducible with installation of  unofficial (by wg9s) en-US SeaMonkey 2.53 
(NT 6.1; Win64; x64; rv:56.0) Gecko/20100101 Firefox/56.0 Build 20190329130005 
(Default Classic Theme, German Language Pack active) on German WIN7 64bit

CU

Rainer

---
Diese E-Mail wurde von Avast Antivirus-Software auf Viren geprüft.
https://www.avast.com/antivirus

0
Rainer
7/11/2019 9:39:43 AM
On 11/07/19 06:08, Paul B. Gallagher wrote:
> Check out the umlauts here:
> 
> <https://www.nytimes.com/2019/07/10/science/skull-neanderthal-human-europe-greece.html> 
> 
> 
> For some reason, SM displays them to the right of the target letter, 
> although the source code displays them correctly. It's because the 
> author has used "u" plus the combining umlaut rather than the single 
> letter "ü," thus: "ü."

Interesting article, anyway.

The answer is in this Bugzilla item.

"the combining ... character [is not supported in the selected font]; so 
it falls back to a different font ..., and positioning of diacritics 
does not generally work across font-change boundaries." 
<https://bugzilla.mozilla.org/user_profile?user_id=329583>

This happens in this case because of an apparently unreported bug in SM.

In more detail, I tested, with SM 2.49.5 build 20190507045107, LXLE 
(Lubuntu) 16.04, 32-bit, no JS, the rendering of this paragraph:

<p class="css-exrw3m evys1bk0">“It’s a very good question, and I have no 
idea,” said Dr. Harvati, a paleoanthropologist at the University of 
Tübingen in Germany. “I mean, this is the first time that we’ve found 
them.”</p>

Using the document styles, "Tübingen" appears as "Tu¨bingen". The font 
style specified by NYT for this is "nyt-imperial,georgia,'times new 
roman',times,serif". The font used is matched to the 'times new roman' 
font name even though Georgia is installed: it *is* used if the name is 
"Georgia" rather than "georgia" -- perhaps "times new roman" is handled 
specially for historical reasons -- but this is a bug against "CSS Fonts 
Module Level 3 W3C Recommendation 20 September 2018" 
<https://www.w3.org/TR/css-fonts-3/#font-matching-algorithm>.

Using the browser default styles, it's "Tübingen". This corresponds to a 
browser font-family "serif", which on this platform seems to be the font 
"Bitstream Vera Serif".

"Times New Roman" v 2.82, "a valuable asset of Monotype", is one of a 
small number of system serif fonts installed in this platform that 
display the u and combining umlaut separately; most display ü correctly 
(tested in Font Manager).

After enabling @font-face and JS for nytimes.com and static01.nyt.com, I 
see what must be the nyt-imperial font 
<https://www.typesample.com/samples/nyt_8iqqn_2x> offered by NYT (oh, 
it's almost indistinguishable from Georgia), but it also fails to 
combine "u¨". Apparently NYT uses Imperial in the print edition but has 
had Georgia for the web since 2006 (absent @font-face).

Lesson 0
SM 2.49.5 (and presumably earlier versions) is at least 6 years behind 
the current CSS font matching specification.

Lesson 1
If you want to avoid this display, at least check the default fonts used 
by the browser and set the defaults to fonts that display u and 
combining umlaut properly (other combinations may also apply). Eg, don't 
have Times New Roman 2.82 as the default serif font.

Lesson 2
You won't lose much by disabling site font styles globally: clear 
Edit>Preferences>Appearance>Fonts>Allow documents to use other fonts.

Lesson 3
You can prevent sites sending you their own fonts by blocking @font-face 
with (eg) NoScript.

Lesson 4
You can overrule unsatisfactory site font styles with user CSS, directly 
or by means of an extension like Stylish. This user CSS might fix the 
display if (eg) Georgia is installed:

..css-exrw3m { font-family: Georgia,serif; }

Or a user JS could edit font-family styles to remove 'times new roman' 
and/or change "georgia" to "Georgia".

/df

-- 
London
UK
0
Dirk
7/11/2019 12:01:52 PM
On 11/07/19 10:39, Rainer Bielefeld wrote:
> On 11.07.2019 07:08, Paul B. Gallagher wrote:
>> So "the University of Tübingen" appears as "the University of Tu¨bingen," 
> 
> Hi Paul,
> 
> NOT reproducible with installation of  unofficial (by wg9s) en-US 
> SeaMonkey 2.53 (NT 6.1; Win64; x64; rv:56.0) Gecko/20100101 Firefox/56.0 
> Build 20190329130005 (Default Classic Theme, German Language Pack 
> active) on German WIN7 64bit

Which font is actually being selected for the display?

If it's not Georgia and Georgia is installed, is Georgia used if the CSS 
font-family is "Georgia,serif" instead of the NYT's list that includes 
"georgia"?

If it's Times New Roman, does it have the combining diaeresis U+0308? 
What version is it (Fonts Control Panel, Properties, when I last tried)?

thx
/df

-- 
London
UK
0
Dirk
7/11/2019 2:25:05 PM
Dirk Fieldhouse wrote:

> On 11/07/19 10:39, Rainer Bielefeld wrote:
>> On 11.07.2019 07:08, Paul B. Gallagher wrote:
>>> So "the University of Tübingen" appears as "the University of 
>>> Tu¨bingen," 
>>
>> Hi Paul,
>>
>> NOT reproducible with installation of  unofficial (by wg9s) en-US 
>> SeaMonkey 2.53 (NT 6.1; Win64; x64; rv:56.0) Gecko/20100101 
>> Firefox/56.0 Build 20190329130005 (Default Classic Theme, German 
>> Language Pack active) on German WIN7 64bit
> 
> Which font is actually being selected for the display?
> 
> If it's not Georgia and Georgia is installed, is Georgia used if the CSS 
> font-family is "Georgia,serif" instead of the NYT's list that includes 
> "georgia"?
> 
> If it's Times New Roman, does it have the combining diaeresis U+0308? 
> What version is it (Fonts Control Panel, Properties, when I last tried)?

Yes, TNR does contain the combining diaeresis at U+0308. I have version 
5.22 of February 5, 2016. I also have Georgia v. 5.00 of June 10, 2009, 
and it has this character too.

Thanks for your other much more detailed investigation. Fascinating.

I'm afraid I don't have the expertise to see at a glance whether my 
system is actually using TNR or Georgia or something else. I'm good with 
audio, enough to tell if your F# is a little too sharp or not sharp 
enough, but not so much with visuals. Still, after a bit of staring it 
seems to me (don't quote me in court!) that Georgia is being used.

-- 
War doesn't determine who's right, just who's left.
--
Paul B. Gallagher
0
Paul
7/11/2019 3:05:20 PM
On 7/10/2019 10:08 PM, Paul B. Gallagher wrote:
> Check out the umlauts here:
> 
> <https://www.nytimes.com/2019/07/10/science/skull-neanderthal-human-europe-greece.html>
> 
> For some reason, SM displays them to the right of the target letter, 
> although the source code displays them correctly. It's because the 
> author has used "u" plus the combining umlaut rather than the single 
> letter "ü," thus: "ü."
> 
> So "the University of Tübingen" appears as "the University of 
> Tu¨bingen," and "Rainer Grün" appears as "Rainer Gru¨n."
> 
> I'm not using an odd font here, and they don't specify one. It's 
> plain-vanilla Times Roman.
> 
> Copy/pasting the source code into Notepad, MS Word, or other 
> applications shows that it displays correctly there. Even copy/pasting 
> the text from the badly displayed browser window shows that it displays 
> correctly elsewhere. Only in an SM browser window is the umlaut 
> misplaced. No, wait, IE 11 has the same problem.
> 
> Any idea how to fix this?
> 

It might be fixed if the New York Times would correct the 37 HTML errors
contained in that Web page.

-- 
David E. Ross
<http://www.rossde.com/>

Donald Trump lied his way onto the Forbes 400 richest people list.
<https://tinyurl.com/yx9ebrqz>
0
David
7/11/2019 3:07:11 PM
On 7/11/2019 8:05 AM, Paul B. Gallagher wrote:
> Dirk Fieldhouse wrote:
> 
>> On 11/07/19 10:39, Rainer Bielefeld wrote:
>>> On 11.07.2019 07:08, Paul B. Gallagher wrote:
>>>> So "the University of T�bingen" appears as "the University of 
>>>> Tu�bingen," 
>>>
>>> Hi Paul,
>>>
>>> NOT reproducible with installation of� unofficial (by wg9s) en-US 
>>> SeaMonkey 2.53 (NT 6.1; Win64; x64; rv:56.0) Gecko/20100101 
>>> Firefox/56.0 Build 20190329130005 (Default Classic Theme, German 
>>> Language Pack active) on German WIN7 64bit
>>
>> Which font is actually being selected for the display?
>>
>> If it's not Georgia and Georgia is installed, is Georgia used if the CSS 
>> font-family is "Georgia,serif" instead of the NYT's list that includes 
>> "georgia"?
>>
>> If it's Times New Roman, does it have the combining diaeresis U+0308? 
>> What version is it (Fonts Control Panel, Properties, when I last tried)?
> 
> Yes, TNR does contain the combining diaeresis at U+0308. I have version 
> 5.22 of February 5, 2016. I also have Georgia v. 5.00 of June 10, 2009, 
> and it has this character too.
> 
> Thanks for your other much more detailed investigation. Fascinating.
> 
> I'm afraid I don't have the expertise to see at a glance whether my 
> system is actually using TNR or Georgia or something else. I'm good with 
> audio, enough to tell if your F# is a little too sharp or not sharp 
> enough, but not so much with visuals. Still, after a bit of staring it 
> seems to me (don't quote me in court!) that Georgia is being used.

Windows 7
SeaMonkey 2.49.4

The body of the article uses a font named "FB Web Use Only".  However,
the mangled "T�bingen" is a mixture of that font and Georgia.  The
source does not show a change of font for "T�bingen".  I see the problem
in both SeaMonkey and Internet Explorer.

If I change my preferences at [Appearance > Fonts] to uncheck the
checkbox "Allow documents to use other fonts", the problem is resolved.
This causes the page to render with my choice: Georgia 5.00.

-- 
David E. Ross
<http://www.rossde.com/>

Donald Trump lied his way onto the Forbes 400 richest people list.
<https://tinyurl.com/yx9ebrqz>
0
David
7/11/2019 8:16:56 PM
David E. Ross wrote:

> The body of the article uses a font named "FB Web Use Only". However,
> the mangled "T�bingen" is a mixture of that font and Georgia. The
> source does not show a change of font for "T�bingen". I see the problem
> in both SeaMonkey and Internet Explorer.
> 
> If I change my preferences at [Appearance > Fonts] to uncheck the
> checkbox "Allow documents to use other fonts", the problem is resolved.
> This causes the page to render with my choice: Georgia 5.00.

Same here, though my pref for Western is set to Times New Roman. Either 
way, I've never heard of "FB Web Use Only," and I'll bet dollars to 
donuts that no visitors have it installed. Wonder why they'd do that.

-- 
War doesn't determine who's right, just who's left.
--
Paul B. Gallagher
0
Paul
7/11/2019 10:01:40 PM
On 11.07.2019 22:16, David E. Ross wrote:
> "Allow documents to use other fonts"


Hello,

indeed, after having selected "Allow documents to use other fonts" I can 
reproduce the problem.


CU

Rainer

---
Diese E-Mail wurde von Avast Antivirus-Software auf Viren gepr=FCft.
https://www.avast.com/antivirus

0
Rainer
7/12/2019 7:37:48 AM
Not fixed in 2.57. This one is using stylo the new mozilla style system so 
probably in current Fx too.

FRG

Rainer Bielefeld wrote:
> On 11.07.2019 22:16, David E. Ross wrote:
>> "Allow documents to use other fonts"
> 
> 
> Hello,
> 
> indeed, after having selected "Allow documents to use other fonts" I can 
> reproduce the problem.
> 
> 
> CU
> 
> Rainer
> 
> ---
> Diese E-Mail wurde von Avast Antivirus-Software auf Viren gepr�ft.
> https://www.avast.com/antivirus
> 
0
Frank
7/12/2019 12:02:06 PM
On 12/07/19 13:02, Frank-Rainer Grahl wrote:
> Not fixed in 2.57. This one is using stylo the new mozilla style system 
> so probably in current Fx too.

I presume you mean the issue that a combining punctuation character is 
font-matched separately from its preceding character. I think this is 
fairly low priority since a) why not use actual ü instead of uU+0308, to 
use this example -- perhaps it fails better if the browser's font 
selection has no ü, and b) the problem disappears if the selected font 
has both u and U+0308.

What about failing to match the CSS font-family case-insensitively (as 
in this case, platform's "Georgia" vs site's "georgia"), which is 
definitely a spec violation?

/df

-- 
London
UK
0
Dirk
7/12/2019 1:46:40 PM
Op 11-7-2019 om 14:01 schreef Dirk Fieldhouse:
> On 11/07/19 06:08, Paul B. Gallagher wrote:
>> Check out the umlauts here:
>>
>> <https://www.nytimes.com/2019/07/10/science/skull-neanderthal-human-europe-greece.html>
>>
>>
>> For some reason, SM displays them to the right of the target letter,
>> although the source code displays them correctly. It's because the
>> author has used "u" plus the combining umlaut rather than the single
>> letter "ü," thus: "ü."
> 
> Interesting article, anyway.
> 
> The answer is in this Bugzilla item.
> 
> "the combining ... character [is not supported in the selected font]; so
> it falls back to a different font ..., and positioning of diacritics
> does not generally work across font-change boundaries."
> <https://bugzilla.mozilla.org/user_profile?user_id=329583>
> 
> This happens in this case because of an apparently unreported bug in SM.

This also happen in latest Firefox 68.0. Anyway, I don't think it's a
browser issue, but a website issue, if neither Firefox and Internet
Explorer display the page correctly, unless you specify to not let the
page override the font...

Onno
0
Onno
7/12/2019 6:25:24 PM
Paul B. Gallagher wrote:
> Check out the umlauts here:
> 
> <https://www.nytimes.com/2019/07/10/science/skull-neanderthal-human-europe-greece.html> 
> 
> 
> For some reason, SM displays them to the right of the target letter, although 
> the source code displays them correctly. It's because the author has used "u" 
> plus the combining umlaut rather than the single letter "ü," thus: "ü."
> 
> So "the University of Tübingen" appears as "the University of Tu¨bingen," and 
> "Rainer Grün" appears as "Rainer Gru¨n."
> 
> I'm not using an odd font here, and they don't specify one. It's plain-vanilla 
> Times Roman.
> 
> Copy/pasting the source code into Notepad, MS Word, or other applications 
> shows that it displays correctly there. Even copy/pasting the text from the 
> badly displayed browser window shows that it displays correctly elsewhere. 
> Only in an SM browser window is the umlaut misplaced. No, wait, IE 11 has the 
> same problem.
> 
> Any idea how to fix this?
> 
Can you not use the character map to get an accented character?  It is easier 
with MacOS because there are keyboard shortcuts to create accented characters.

0
EE
7/13/2019 8:28:00 PM
On 13/07/19 21:28, EE wrote:
> 
>... >
> Can you not use the character map to get an accented character?  It is 
> easier with MacOS because there are keyboard shortcuts to create 
> accented characters.

You need to ask the NY Times, whose web page is at issue, why they don't 
do that. One possibility is that the composite uU+0308 displays better 
if the user's fonts don't have the actual accented character, which 
might have been more likely when the NYT website production system was 
set up, probably more than a decade ago.

/df

-- 
London
UK
0
Dirk
7/13/2019 11:07:16 PM
EE wrote:
> Paul B. Gallagher wrote:
>> Check out the umlauts here:
>>
>> <https://www.nytimes.com/2019/07/10/science/skull-neanderthal-human-europe-greece.html> 
>>
>>
>> For some reason, SM displays them to the right of the target letter, 
>> although the source code displays them correctly. It's because the 
>> author has used "u" plus the combining umlaut rather than the single 
>> letter "ü," thus: "ü."
>>
>> So "the University of Tübingen" appears as "the University of 
>> Tu¨bingen," and "Rainer Grün" appears as "Rainer Gru¨n."
>>
>> I'm not using an odd font here, and they don't specify one. It's 
>> plain-vanilla Times Roman.
>>
>> Copy/pasting the source code into Notepad, MS Word, or other 
>> applications shows that it displays correctly there. Even copy/pasting 
>> the text from the badly displayed browser window shows that it 
>> displays correctly elsewhere. Only in an SM browser window is the 
>> umlaut misplaced. No, wait, IE 11 has the same problem.
>>
>> Any idea how to fix this?
>>
> Can you not use the character map to get an accented character?  It is 
> easier with MacOS because there are keyboard shortcuts to create 
> accented characters.

I know perfectly well how to make accented characters in my own 
documents, HTML and otherwise. That wasn't the question.

-- 
War doesn't determine who's right, just who's left.
--
Paul B. Gallagher
0
Paul
7/13/2019 11:16:01 PM
Reply: