Firefox problem - CSS vs FONT FACE

[B Wooster]

There is a page that shows Firefox 1.0 render text differently for CSS
as compared to FONT FACE for one particular font:
http://www.aczoom.com/fonts/xdvng.ttf is the font that does not work.

http://www.aczoom.com/isongs/0/t_ff.html
is the test page - that has shows the line "This is using CSS to
specify xdvng:" show up with the default english font instead of
xdvng.

Internet Explorer works as expected - both CSS and FONT FACE use of
xdvng displays identical text.

I am not sure if I should submit this as a bug for firefox, or if this
has something to do with the xdvng font?

[Chris Leipold]

Hi,

you should read the FAQ for this list. (http://css.nu/faq/ciwas-aFAQ.html)

Chris

B Wooster wrote:

There is a page that shows Firefox 1.0 render text differently for CSS
as compared to FONT FACE for one particular font:
http://www.aczoom.com/fonts/xdvng.ttf is the font that does not work.

http://www.aczoom.com/isongs/0/t_ff.html
is the test page - that has shows the line "This is using CSS to
specify xdvng:" show up with the default english font instead of
xdvng.

Internet Explorer works as expected - both CSS and FONT FACE use of
xdvng displays identical text.

I am not sure if I should submit this as a bug for firefox, or if this
has something to do with the xdvng font?

[B Wooster]

Chris Leipold <cl******@dietzk.de> wrote in message news:<2v*************@uni-berlin.de>...

you should read the FAQ for this list. (http://css.nu/faq/ciwas-aFAQ.html)
Chris
I did read the FAQs, re-read the page you mention - but I am not sure
what
you think in there applies to my question.

The question simply is this: take any browser, and text defined using
CSS and font-family, and alternately FONT FACE HTML tag.
Should they render differently?

Firefox renders that text differently.

I am not sure if I should submit this as a bug for firefox, or if this
has something to do with the xdvng font - in which case I will enter a
bug report for firefox.
On the other hand, if there is a reason why a CSS font-family can
safely
ignore a font that exists on the system, it is not a bug in firefox -
but
I would like to understand why!
B Wooster wrote:

There is a page that shows Firefox 1.0 render text differently for CSS
as compared to FONT FACE for one particular font:
http://www.aczoom.com/fonts/xdvng.ttf is the font that does not work.

http://www.aczoom.com/isongs/0/t_ff.html
is the test page - that has shows the line "This is using CSS to
specify xdvng:".....

[Andrew Thompson]

On 10 Nov 2004 10:16:35 -0800, B Wooster wrote:

Chris Leipold <cl******@dietzk.de> wrote in message news:<2v*************@uni-berlin.de>...

you should read the FAQ for this list. (http://css.nu/faq/ciwas-aFAQ.html)
Chris

I did read the FAQs, re-read the page you mention - but I am not sure
what
you think in there applies to my question.

Chris may well have been thinking of this (very important) section..
<http://css.nu/faq/ciwas-aFAQ.html#QA11>

Which soon leads one to..
<http://validator.w3.org/check?uri=http%3A%2F%2Fwww.aczoom.com%2Fisongs%2F0 %2Ft_ff.html&charset=iso-8859-1+%28Western+Europe%29>

HTH

--
Andrew Thompson
http://www.PhySci.org/codes/ Web & IT Help
http://www.PhySci.org/ Open-source software suite
http://www.1point1C.org/ Science & Technology
http://www.LensEscapes.com/ Images that escape the mundane

[B Wooster]

Andrew Thompson <Se********@www.invalid> wrote in message news:<i7****************************@40tude.net>.. .

On 10 Nov 2004 10:16:35 -0800, B Wooster wrote:

Chris may well have been thinking of this (very important) section..
<http://css.nu/faq/ciwas-aFAQ.html#QA11>
Which soon leads one to.. <http://validator.w3.org/check?uri=http%3A%2F%2Fwww.aczoom.com%2Fisongs%2F0 %2Ft_ff.html&charset=iso-8859-1+%28Western+Europe%29> HTH

Fair enough, though all the reported errors are in a later section of
the document, not in the part where I had the example with the
problem. Here's another page that has only the minimal text to show
the CSS vs FONT FACE difference (only a problem on Firefox, works ok
on atleast Internet Explorer - if someone with Opera or Safari can
tell me if the page works correctly on their browsers, would
appreciate it - page to test:
http://www.aczoom.com/isongs/0/ff.html

and a validator URL (also followed the CSS validation - no warnings,
no errors):

http://validator.w3.org/check?uri=ht...tern+Europe%29

[Christoph Paeper]

B Wooster <bw********@gmail.com>:

the CSS vs FONT FACE difference (only a problem on Firefox, works ok
on atleast Internet Explorer - if someone with Opera or Safari can
tell me if the page works correctly on their browsers,
OperaÂ*7, which is available in a free ad-sponsored version, does show
"English" letters everywhere. I don't have a font called "xdvng"
installed, but do have some fonts featuring the Devanagari alphabet (Arial
Unicode MS, Mangal, Titus Cyberbit Basic). (The cryptic font name and the
keyword "Hindi" tell me that that script is wanted.)
http://www.aczoom.com/isongs/0/ff.html

Your concept is broken.¹ Devanagari is in the Unicode range U+0900-097F
(dec. 2304-2431, not all used). Use those characters, either with UTF-8
(अ) or numeric character references (&#x0905;, अ). Everything else
is just hackerish and not guaranteed to work---as you have already
encountered the hard way.

I think Alan J. Flavell (<http://ppewww.ph.gla.ac.uk/~flavell/charset/>)
and Jukka Korpela (<http://www.cs.tut.fi/~jkorpela/html/chars.html>) have
some pages on that issue (not focussed on mixed Hindi and English texts,
though).

____
¹ I know that several groups have been using something like this for years
nevertheless, because nothing else worked at all.

--
"Not all chemicals are bad. Without chemicals such as hydrogen and oxygen,
for example, there would be no way to make water, a vital ingredient in beer."
Dave Barry

[Rijk van Geijtenbeek]

On Thu, 11 Nov 2004 05:32:43 +0100, Christoph Paeper
<ch**************@nurfuerspam.de> wrote:

...

http://www.aczoom.com/isongs/0/ff.html
Your concept is broken.¹

... ____
¹ I know that several groups have been using something like this for
years nevertheless, because nothing else worked at all.

A large number of Indian websites still this broken concept :(

--
Rijk van Geijtenbeek

The Web is a procrastination apparatus:
It can absorb as much time as is required to ensure that you
won't get any real work done. - J.Nielsen

[Andreas Prilop]

On Thu, 11 Nov 2004, Christoph Paeper wrote:

http://www.aczoom.com/isongs/0/ff.html
Your concept is broken.

Absolutely!
I think Alan J. Flavell (<http://ppewww.ph.gla.ac.uk/~flavell/charset/>)
and Jukka Korpela (<http://www.cs.tut.fi/~jkorpela/html/chars.html>) have
some pages on that issue (not focussed on mixed Hindi and English texts,
though).

For Devanagari text, see
<http://www.unics.uni-hannover.de/nhtcapri/multilingual1.html#nagari>
<http://www.unics.uni-hannover.de/nhtcapri/multilingual2.html#nagari>
<http://www.unics.uni-hannover.de/nhtcapri/urdu-alphabet.html>
<http://www.bbc.co.uk/hindi/>

[Andreas Prilop]

On 10 Nov 2004, B Wooster wrote:

I am not sure if I should submit this as a bug for firefox, or if this
has something to do with the xdvng font - in which case I will enter a
bug report for firefox.

File a bug report for the nonsense font xdvng - better yet trash
xdvng immediately and forget about it!
Read
<http://www.tldp.org/HOWTO/Indic-Fonts-HOWTO/>
<http://www.unicode.org/faq/indic.html>
<http://www.microsoft.com/typography/otfntdev/indicot/>
<http://docs.info.apple.com/article.html?artnum=20891>
to learn about Indic fonts.

[B Wooster]

Christoph Paeper <ch**************@nurfuerspam.de> wrote in message news:<opsg99ktxbwsrb86@crissov>...

B Wooster <bw********@gmail.com>:

http://www.aczoom.com/isongs/0/ff.html

Your concept is broken.¹ Devanagari is in the Unicode range U+0900-097F

I think Alan J. Flavell (<http://ppewww.ph.gla.ac.uk/~flavell/charset/>)
and Jukka Korpela (<http://www.cs.tut.fi/~jkorpela/html/chars.html>) have

I am quite familiar with Alan's great docs, but my question was much
simpler!
I can understand (as Alan explains) if a Browser enforces character
encodings, etc.

But - let's ignore non-ASCII characters, using Latin1 encoding.
And the system has the xdvng font installed, and that font is
advertised as a Latin1 font.
Now, if I have a word "abcd" and I use FONT FACE for xdvng, and I use
CSS font family for xdvng - should they render *differently*?

I can totally understand if both of these render using some other font
than xdvng, but to have FONT FACE render differently from CSS - that
is the only issue I was asking about!

And the page http://www.aczoom.com/isongs/0/ff.html
now shows only ASCII characters (removed all 127+ char codes), and
shows the problem....

[Andreas Prilop]

On 11 Nov 2004, B Wooster wrote:

But - let's ignore non-ASCII characters, using Latin1 encoding.
And the system has the xdvng font installed, and that font is
advertised as a Latin1 font.
Please make up your mind! Either "xdvng" is a Latin-1 font or
it is a Devanagari font.
Now, if I have a word "abcd" and I use FONT FACE for xdvng, and I use
CSS font family for xdvng - should they render *differently*?
Try to understand the difference between "character" and "glyph".
That's fundamental! With <font face> or with CSS, you can have
different glyphs (i.e. from different typefaces) for your string
"abcd" in Latin letters. But "abcd" *always* remain Latin letters.
There's *no way* to transform "abcd" by some magic into Devanagari
letters. Carefully read
<http://ppewww.ph.gla.ac.uk/~flavell/charset/fontface-harmful.html>
You cannot turn the Latin letter "q" into a Greek "theta".
I can totally understand if both of these render using some other font
than xdvng, but to have FONT FACE render differently from CSS - that
is the only issue I was asking about!
Your *approach* is broken and you seem not to understand it.
And the page http://www.aczoom.com/isongs/0/ff.html

.... is bullshit.

Here's how to display Devanagari text in a web browser:
<http://www.unics.uni-hannover.de/nhtcapri/multilingual1.html#nagari>
<http://www.unics.uni-hannover.de/nhtcapri/multilingual2.html#nagari>
<http://www.bbc.co.uk/hindi/>
<http://www.google.co.in/intl/hi/>

[Christoph Paeper]

*Andreas Prilop* <nh******@rrzn-user.uni-hannover.de>:

On 11 Nov 2004, B Wooster wrote:

But - let's ignore non-ASCII characters, using Latin1 encoding.
And the system has the xdvng font installed, and that font is
advertised as a Latin1 font.
Please make up your mind! Either "xdvng" is a Latin-1 font or
it is a Devanagari font.

We all know it's a hack long time in use (and hopefully falling out of it
soon). I find a xdvng.ttf with Google downloaded, installed and looked at
it. With it installed Opera*7 and FireBird (0.6, I really should update)
"work", i.e. show Devanagari glyphs with both, HTML's 'face' and CSS's
'font-family'. Microsoft's TTF Extension tells me its "Font Encoding Type"
is "Unicode (ISO*10646-2)", but the fields for "Supported Unicode Ranges"
and "Supported Code Pages" read both "Information not available in font".
That sounds to me like it does *neither* tell to support Devanagari nor
Basic Latin/ASCII/ISO-8859-X (which is not really lying). Some other
similar fonts I have are said to have an encoding of "Symbol".

Now, if I have a word "abcd" and I use FONT FACE for xdvng, and I use
CSS font family for xdvng - should they render *differently*?

My best guess is, that newer Geckos have some sort of quirks handling,
that still allows the 'font face' hackery to work, but is more strict on
the CSS side. There'd be /some/ sense in that, given that sites relying on
(for a reason) badly encoded fonts are probably from the time before the
wider adoption of both, Unicode and CSS.

Other browsers/programs probably just assume that xdvng.ttf has the glyphs
of Latin-1, i.e. ISO-8859-1, although it does not say so. How are search
engines supposed to deal well with that?
Try to understand the difference between "character" and "glyph".

I guess he already knows, but doesn't want to change for some reason.
Maybe it is, (that it's assumed) that among readers of Hindi it's more
likely to be using older hardware and software. However he does seem to
want to use CSS for some reason, thus not making the same assumption as
the Gecko developers (by my allegement).

--
The penis mightier than the sword.

[Alan J. Flavell]

On Thu, 11 Nov 2004, Christoph Paeper wrote:

My best guess is, that newer Geckos have some sort of quirks
handling, that still allows the 'font face' hackery to work, but is
more strict on the CSS side.
There has been a long, long argument on their bugzilla about whether
to do what authors want, or what the specification wants; for all that
I know, this argument may still be going on.

You might find the font called "Symbol" being treated specially. You
might find all fonts which claim to be of type "Symbol" treated
specially. You might find differences between standards mode and
quirks mode. ALL of that detail is bogus as far as handling
specification-conforming HTML documents is concerned: it's only a hack
which attempts to render some wrongly-made pages in the way that their
misguided author intended, in the interests of the *reader* of such
misguided documents. It should not, repeat NOT, tempt anyone to
repeat those authoring blunders.

The only authoring advice that I can offer at this time, as you can
see, is *not* to make new documents a hostage to such long-since-
obsolete tricks. This has been clear, to those who have been taking
an interest, since at least RFC2070; and more widely since HTML4.0 was
specified. That's a LONG time ago in www terms.
There'd be /some/ sense in that, given that sites relying on (for a
reason) badly encoded fonts are probably from the time before the
wider adoption of both, Unicode and CSS.

Possibly; but that's all a long time ago, as I say. Document authors
who *still* in 2004 haven't caught up with HTML4 probably are missing
a lot of other important developments; you might want to consider
whether they really still deserve to be taken seriously.

IMHO and all that.

[Andreas Prilop]

On Thu, 11 Nov 2004, Christoph Paeper wrote:

Other browsers/programs probably just assume that xdvng.ttf has the glyphs
of Latin-1, i.e. ISO-8859-1, although it does not say so. How are search
engines supposed to deal well with that?

Search engines don't give^W care about HTML markup, <font face>, CSS.
<http://www.google.com/search?q=%221+W+1+V%2FA%22> finds
pages that claim 1 W = 1 V/A and
pages that claim 1 W = 1 V·A .

Confusing, isn't it? ;-)

[Daniel R. Tobias]

bw********@gmail.com (B Wooster) wrote in message news:<20**************************@posting.google. com>...

http://www.aczoom.com/fonts/xdvng.ttf is the font that does not work.
That server is serving the font file as "text/plain", meaning that the
user agent at the receiving end should (if standards-compliant) render
it as plain text (as my copy of Mozilla in fact does).
Internet Explorer works as expected - both CSS and FONT FACE use of
xdvng displays identical text.

Usually when IE "works" and other browsers don't, that's a sign that
the site is done wrong and it is in fact the other browsers that "get
it right".

--
Dan

[B Wooster]

"Alan J. Flavell" <fl*****@ph.gla.ac.uk> wrote in message news:<Pi*******************************@ppepc56.ph .gla.ac.uk>...

Possibly; but that's all a long time ago, as I say. Document authors
who *still* in 2004 haven't caught up with HTML4 probably are missing
a lot of other important developments; you might want to consider
whether they really still deserve to be taken seriously.

IMHO and all that.

Thanks everyone for comments, I have got all the information I needed,
looks like things changed between Mozilla 1.2 and Firefox 1.0, and it
may all be that some parts for Firefox follow the new spec, but in that
case certain other parts certainly dont.

http://www.aczoom.com/isongs/0/ff.html
now makes it clear that old documents are hacks, but if anyone is
interested in deciphering an old doc that may exist, hopefully that
test page will make them clear that it is a hack, and test if their
browser will work.

Thanks, everyone, over and out from my end.

[Christoph Paeper]

*Andreas Prilop* <nh******@rrzn-user.uni-hannover.de>:

On Thu, 11 Nov 2004, Christoph Paeper wrote:

Other browsers/programs probably just assume that xdvng.ttf has the
glyphs of Latin-1, i.e. ISO-8859-1, although it does not say so. How
are search engines supposed to deal well with that?
Search engines don't give^W care about HTML markup, <font face>, CSS.

I know, that question was rhetorical.
OTOH, if you have a (language) group that heavily relies on such hacks, it
may be a /welcomed/ feature (by that group) for a search engine to "know"
this hacks. In practice, however, this knowledge of hacks basically makes
search engines ignore markup.
<http://www.google.com/search?q=%221+W+1+V%2FA%22> finds
pages that claim 1 W = 1 V/A and
pages that claim 1 W = 1 V·A .

Confusing, isn't it? ;-)

Well, it's handy if the ignored characters are pretty much interchangable,
like in "don't" and "don’t" or "800·600", "800×600" and "800*600"---only
"800x600" doesn't work.

--
"I didn't have the time to write a short letter, so I wrote a long one instead."
Mark Twain

[Andreas Prilop]

On Fri, 12 Nov 2004, Christoph Paeper wrote:

<http://www.google.com/search?q=%221+W+1+V%2FA%22> finds
pages that claim 1 W = 1 V/A and
pages that claim 1 W = 1 V·A .

Confusing, isn't it? ;-)

Well, it's handy if the ignored characters are pretty much interchangable,

I'm afraid you completely missed my point.

The equation 1 W = 1 V/A means 1 ohm = 1 volt per ampere .
The equation 1 W = 1 V·A means 1 watt = 1 volt times ampere .

[SeaPlusPlus]

Andreas Prilop wrote:

The equation 1 W = 1 V/A means 1 ohm = 1 volt per ampere .
I believe you meant...
The equation 1 W = 1 V/A means 1 watt = 1 volt per ampere.
The equation 1 W = 1 V·A means 1 watt = 1 volt times ampere .

Don't mean to get 'picky' on you... sorry...

Thanks...

Rich

[Christoph Paeper]

*Andreas Prilop* <nh******@rrzn-user.uni-hannover.de>:

On Fri, 12 Nov 2004, Christoph Paeper wrote:

pages that claim 1 W = 1 V/A and
pages that claim 1 W = 1 V·A .

Well, it's handy if the ignored characters are pretty much
interchangable,

I'm afraid you completely missed my point.

Somehow, I forgot to consider the topic of the thread and that the first
equation can become true /visually/ with something like "<font
face='symbol'>W</font>".

--
"Anyone who slaps a 'this page is best viewed with Browser X' label on a
Web page appears to be yearning for the bad old days, before the Web, when
you had very little chance of reading a document written on another computer,
another word processor, or another network." Tim Berners-Lee, 1996

[Andreas Prilop]

On Mon, 15 Nov 2004, SeaPlusPlus wrote:

The equation 1 W = 1 V/A means 1 ohm = 1 volt per ampere .

I believe you meant...
The equation 1 W = 1 V/A means 1 watt = 1 volt per ampere.

No, I mean what I write. 1 ohm = 1 volt per ampere
We can learn from Dr. Watson^W Stockton:
W is the symbol for watt.
W is the symbol for ohm.
Ö is the symbol for square root.
<http://google.com/search?ie=ISO-8859-1&q=%22ohm+symbol+W%22>
<http://google.com/search?ie=ISO-8859-1&q=%22square+root+%D6%22>

--
http://www.unics.uni-hannover.de/nht...quareroot.html