apostrophe in html
Andreas Prilop
ne...@hoboes.com (Jerry Stratton) wrote:
>If you mean curly quotes, use:
>
>‘ Åš
^ This character does not exist in ISO-8859-1.
>’ Ä…
^ This is a superscript 1.
>“ Å‚
^ This is a superscript 3.
>” Ë›
^ This is a superscript 2.
The expressions ‘ ’ “ ” do not exist at all.
Please don't post advice if you don't know what you are talking about.
--
Es lässt sich ohne sonderlich viel Witz so schreiben,
dass ein anderer sehr vielen haben muss, es zu verstehen.
Georg Christoph Lichtenberg
Jerry Stratton
<nhtcapri-ya0240800...@newsserver.rrzn.uni-hannover.de>,
nhtc...@rrzn-user.uni-hannover.de (Andreas Prilop) wrote:
>In article <newsw-09099...@cx38767-a.dt1.sdca.home.com>,
>ne...@hoboes.com (Jerry Stratton) wrote:
>
>>If you mean curly quotes, use:
>>
>>‘ Åš
> ^ This character does not exist in ISO-8859-1.
>
>>’ Ä…
> ^ This is a superscript 1.
>
>>“ Å‚
> ^ This is a superscript 3.
>
>>” Ë›
> ^ This is a superscript 2.
>
>The expressions ‘ ’ “ ” do not exist at all.
>Please don't post advice if you don't know what you are talking about.
Please don't respond if you don't know what you're talking about. ‘
through ” work on every browser on every platform that I've seen, from
Lynx on basic terminals (it converts them to straight quotes, to Explorer
and Netscape on Windows and Macintosh, Netscape on Linux and SunOS, iCab
on Macintosh, and WebTV.
Curly quotes carry more information than straight quotes and make a big
differenc in readability, especially of long documents or documents with
dialogue.
If you paid attention to the posting or knew what you were talking about,
you would have realized the posting was not in HTML, it was *about* HTML.
Your complaints were about characters that would never be used in
HTML--that's why we use the encodings instead. Those characters have
different encodings on different computer platforms, but browsers will
render them correctly regardless of platform.
I recommend a good book on HTML, such as HTML: The Definitive Guide, or
for a good reference, Web Design in a Nutshell.
gr...@apple2.com
ne...@hoboes.com (Jerry Stratton) wrote:
>nhtc...@rrzn-user.uni-hannover.de (Andreas Prilop) wrote:
>> The expressions ‘ ’ “ ” do not exist at all.
>> Please don't post advice if you don't know what you are talking about.
> Please don't respond if you don't know what you're talking about.
But he does.
> ‘
> through ” work on every browser on every platform that I've seen,
There's the fallacy. You're basing your answer on what you've seen rather
than the specification.
> from
> Lynx on basic terminals (it converts them to straight quotes, to Explorer
> and Netscape on Windows and Macintosh, Netscape on Linux and SunOS, iCab
> on Macintosh, and WebTV.
They are still the wrong numeric entities, and you are relying on error
recovery in those browsers to do what you mean. All the numeric entities
from &128; to &159; inclusive are explicitly undefined and are illegal
markup. The correct entities for your curly quotes are &8216;, &8217;,
&8221;, and &8220;, respectively.
Don't go with what looks right, go with what IS right.
Followups to ciwah.
--
Nine quadrillion, nine hundred ninety-nine trillion, nine hundred ninety-nine
billion, nine hundred ninety-nine million, nine hundred ninety-nine thousand,
two hundred forty bottles of beer. You take one down, pass it around, nine
quadrillion, nine hundred ninety-nine trillion, nine hundred ninety-nine
Alan J. Flavell
> All the numeric entities
> from &128; to &159; inclusive are explicitly undefined
yes
> and are illegal markup.
Unfortunately, "no". They are undefined, not illegal; and therefore
they cannot properly be rejected by a validator. This is the source of
much confusion.
cheers
pedantry: their formal name is "numerical character references".
Alan J. Flavell
> >The expressions ‘ ’ “ ” do not exist at all.
> >Please don't post advice if you don't know what you are talking about.
>
> Please don't respond if you don't know what you're talking about.
Of course he knows what he's talking about. It's you who pathetically
believes that HTML is defined by what can be discovered from experiments
on browsers, rather than reading the specifications.
> ‘
> through ” work on every browser on every platform that I've seen, from
> Lynx on basic terminals
Yup, it used to work great, by displaying the text ‘ literally, so
as to demonstrate the cluelessness of the author.
Unfortunately, it seems they've now decided to fix up the error
silently, and do what they assume the author intended. That doesn't
make it right.
And clearly your experience of standards-conforming browsers is somewhat
limited. See for example the comments at
http://www.fourmilab.ch/webtools/demoroniser/
Now, you may well disagree with that, but the fact that you are either
unaware of it or pretend that it doesn't exist means that you aren't
really well-placed to comment on the competence of others.
> Curly quotes carry more information than straight quotes
And <Q> even more so. Just that the Big Two couldn't be bothered to
support it.
> and make a big differenc in readability, especially of long documents
> or documents with dialogue.
Then write the damned things correctly.
> If you paid attention to the posting or knew what you were talking about,
> you would have realized the posting was not in HTML, it was *about* HTML.
We see that you're actually proud of your ignorance, and delight in
abusing people who do know what they're talking about.
> Your complaints were about characters that would never be used in
> HTML--that's why we use the encodings instead.
And that just hammers the nail home, since you evidently can't tell the
difference between encodings and the Document Character Set.
> Those characters
But you have followed up to a thread about undefined numerical character
references, not "those characters".
> I recommend a good book on HTML,
I recommend a look at the SGML declaration for HTML4.0,
http://www.w3.org/TR/REC-html40/sgml/sgmldecl.html
Jukka Korpela
<newsw-10099...@cx38767-a.dt1.sdca.home.com>:
>‘ through ” work on every browser on every platform that I've seen
Then you should extend your sphere of experience. See
http://www.hut.fi/u/jkorpela/www/windows-chars.html
which contains some screen shots of situations where those undefined
references don't "work". There are even more serious risks - not very
common, but really nasty when they materialize - since if e.g. raw
octet 145 is sent in terminal connection it can be treated as control
code. (After all, that's one of the reasons why code positions like
145 are unassigned in Unicode!)
>Curly quotes carry more information than straight quotes and make a big
>differenc in readability, especially of long documents or documents with
>dialogue.
No argument about the preferability of "curly quotes", though I don't
share the view that they make a _big_ difference in _readability_.
Rather, they would make the text look somewhat better.
The question is whether the esthetic gain justifies the risk of
getting data displayed all wrong (not to mention the fact that ‘
etc _are_ undefined by the specifications).
>Those characters have
>different encodings on different computer platforms, but browsers will
>render them correctly regardless of platform.
There is no applicable definition about "correctness", since ‘ is
undefined. And you're wrong in assuming that all browsers treat them
similarly.
>I recommend a good book on HTML, such as HTML: The Definitive Guide, or
>for a good reference, Web Design in a Nutshell.
Thanks for the warning. :-) (I presume that's the source from which
you got the idea of using &145.)
[f'ups trimmed]
--
Yucca, http://www.hut.fi/u/jkorpela/
Jerry Stratton
>In article <newsw-10099...@cx38767-a.dt1.sdca.home.com>,
>ne...@hoboes.com (Jerry Stratton) wrote:
>>nhtc...@rrzn-user.uni-hannover.de (Andreas Prilop) wrote:
>
>>> The expressions ‘ ’ “ ” do not exist at all.
>>> Please don't post advice if you don't know what you are talking about.
>
>> Please don't respond if you don't know what you're talking about.
>
>But he does.
1. Do the expressions ‘ ’ “ ” exist at all?
2. What is the relevance of pointing to the *unencoded* versions and
saying that they don't appear correctly in his *newsreader*?
>> ‘
>> through ” work on every browser on every platform that I've seen,
>
>There's the fallacy. You're basing your answer on what you've seen rather
>than the specification.
No, I'm basing it on what I've *found*. There's a difference. I've felt a
need for curly quotes since I began publishing on the net ten years ago.
When I discovered those nonstandard items listed in an HTML reference, I
searched through the then-current browsers to see how widely this
non-standard was adhered to. Standards are in a pretty sorry state in
HTML, unfortunately. Many standards don't work as widely as some
non-standards. I am acutely aware that this is the most widespread
nonconformity in my web pages, and as soon as browsers stop supporting
them (unlikely) or a better alternative appears that preserves the
information in them, I will make a global change to my databases. Until
then, they stay.
Note follow-ups.
John W. Baxter
Korpela) wrote:
> Hon. usenaut ne...@hoboes.com (Jerry Stratton) wrote in message
> <newsw-10099...@cx38767-a.dt1.sdca.home.com>:
>
> >I recommend a good book on HTML, such as HTML: The Definitive Guide, or
> >for a good reference, Web Design in a Nutshell.
>
> Thanks for the warning. :-) (I presume that's the source from which
> you got the idea of using &145.)
>
HTML: The Definitify Guide does list those problem entities like ‘,
carefully and clearly marked as "Defy compliance by using the nonstandard
(N) entries". This is in the July 1996 printing, which has "Minor
corrections. Updated for HTML 3.2."
I think I should get a later edition. Hmmm...this thread is hazardous to
the bank account.
--John
--
John W. Baxter Port Ludlow, WA USA j...@olympus.net
Philip Stripling
Hi, Jerry,
>
> 1. Do the expressions ‘ ’ “ ” exist at all?
The problem with this question is that the expressions "exist," but they
are undefined. (You may remember from junior high that "Division by zero is
undefined.") So the question "Do they exist" is not meaningful, as the
expressions exist, but -- like dividing by zero -- they are undefined. The
expressions are used by some operating systems as non-printing characters,
such as control codes for printers (backspace, line feeds, and so forth,
for example). While you may have found that the expressions show correctly
in all your tests, I will tell you that I never see curly quotes or curly
apostrophes on my terminal.
>SNIP<
> someone else said:
> >There's the fallacy. You're basing your answer on what you've seen rather
> >than the specification.
>
> No, I'm basing it on what I've *found*. There's a difference. I've felt a
> need for curly quotes since I began publishing on the net ten years ago.
> When I discovered those nonstandard items listed in an HTML reference, I
> searched through the then-current browsers to see how widely this
> non-standard was adhered to. Standards are in a pretty sorry state in
> HTML, unfortunately.
Well, I would disagree. In fact, I would suggest that you have it
backwards. Standards in HTML are in fine condition. It is the _browsers_
that are in a sorry state. Since the expressions you are using are
undefined, some software makers have taken advantage of the "extended
characters" used by some operating systems to have those undefined
characters render curly marks on some terminals. Just not all, but who
cares, right? If the software programmers complied with the standards, that
would not be the case; such is life.
> Many standards don't work as widely as some
> non-standards. I am acutely aware that this is the most widespread
> nonconformity in my web pages, and as soon as browsers stop supporting
> them (unlikely) or a better alternative appears that preserves the
> information in them, I will make a global change to my databases. Until
> then, they stay.
No problem from my point of view. You are aware that the expressions are
undefined, you have found that they show up on all the combinations of
browser, hardware, and operating systems you care about, and you do not
care that the expressions are undefined by the standard. So leave
them. Your position is clear; unfortunately, many will hound the point
ceaselessly, even though they are doomed to fail in convincing you that you
are wrong. It is your markup, and you get to make it the way you want. Some
do not like it, and they get to say so.
I just wish people would stop cross-posting to comp.sys.mac.comm, but I,
too, am doomed to failure in getting my wish. Such is life.
--
Phil Stripling | email to the replyto address is presumed
The Civilized Explorer | spam and read later. email to philip@
http://www.cieux.com/ | civex.com is read daily.
gr...@apple2.com
>>ne...@hoboes.com (Jerry Stratton) wrote:
>>>nhtc...@rrzn-user.uni-hannover.de (Andreas Prilop) wrote:
>>>> The expressions ‘ ’ “ ” do not exist at all.
>>>> Please don't post advice if you don't know what you are talking about.
>>> Please don't respond if you don't know what you're talking about.
>> But he does.
> 1. Do the expressions ‘ ’ “ ” exist at all?
They exist as undefined numerical character references and should not...
nay, must not be used.
> 2. What is the relevance of pointing to the *unencoded* versions and
> saying that they don't appear correctly in his *newsreader*?
That is irrelevant to this branch of the thread and is a non-issue to me.
>>> ‘
>>> through ” work on every browser on every platform that I've seen,
>> There's the fallacy. You're basing your answer on what you've seen rather
>> than the specification.
> No, I'm basing it on what I've *found*. There's a difference.
And it is still the same fallacy. Regardless how the browsers behave when
fed those numerical character references, they are still explicitly
undefined by the specification and must not be used. The Unicode
numerical references &8216;, &8217;, &8221;, and &8220; are the correct
ones, they are the standard ones, and they do work. There's no excuse for
using the invalid numbers when the valid ones are there.
> Note follow-ups.
Then why didn't you honor them when _I_ set them the same way?
--
Nine quadrillion, nine hundred ninety-nine trillion, nine hundred ninety-nine
billion, nine hundred ninety-nine million, nine hundred ninety-nine thousand,
two hundred thirty-four bottles of beer. You take one down, pass it around,
Jerry Stratton
"Alan J. Flavell" <fla...@mail.cern.ch> wrote:
>On Fri, 10 Sep 1999, Jerry Stratton wrote:
>
>> >The expressions ‘ ’ “ ” do not exist at all.
>> >Please don't post advice if you don't know what you are talking about.
>> Please don't respond if you don't know what you're talking about.
>
>Of course he knows what he's talking about. It's you who pathetically
>believes that HTML is defined by what can be discovered from experiments
>on browsers, rather than reading the specifications.
When did I say this in that message? I posted that these entities exist,
that they are sometimes useful to me, and that they might be useful for
the person searching for a way of displaying apostrophes. If you'd like to
get into a discussion about conforming with standards in the real world,
that's fine, but has nothing to do with the original posting (apostrophes)
or his response (these don't exist and the unencoded versions don't show
up correctly). I'd be happy to take up the standards discussion in e-mail
or in comp.infosystems.www.authoring.html. Andreas *didn't* know what he
was talking about. Andreas said that those entities *don't exist*, not
that they were wrong and nonstandard.
Andreas also doesn't "know what he's talking about" because Andreas
responded to a post about numeric entities in HTML by pointing out that
the *unencoded* characters didn't show up correctly on a newsreader's
screen. Go back and look at that message. The reply was pointing directly
at the unencoded characters.
Neither of those issues have anything to do with conformity vs.
nonconformity on the web.
If you would like to continue this discussion about the posting, rather
than about the current state of standards on the web, you should first
answer the following questions:
1. Do those entities exist? Not are they nonstandard or you don't like
them or don't find them useful, but that *do they exist?*.
2. How is the appearance of the *unencoded* versions on a newsreader
relevant to displaying these on the web?
Jerry Stratton
jwba...@olympus.net (John W. Baxter) wrote:
>HTML: The Definitify Guide does list those problem entities like ‘,
>carefully and clearly marked as "Defy compliance by using the nonstandard
>(N) entries". This is in the July 1996 printing, which has "Minor
>corrections. Updated for HTML 3.2."
>
>I think I should get a later edition. Hmmm...this thread is hazardous to
>the bank account.
Heh :*)
The third edition claims to cover HTML 4.0. It still lists ‘, etc.
But rather than using an "N" (Netscape?) to mark the undefined characters,
it uses three exclamation points, and the note "Characters whose
conformance column contains '!!!' are not formally defined by the HTML
standard; use them at your own risk." Personally, I take that warning very
seriously, and of the five items I use I constantly verify that the
current state of browsers I have access to display them in a reasonable
manner.
I would note with irony, however, that the sentence I've reproduced does
not appear in the book as '!!!' surrounded by straight single quotes :*)
Personally, I strongly dislike "HTML: The Definitive Guide". Relevant to
this discussion, the authors have a tendency to, in one paragraph, decry
the use of nonstandard code in general, and in the next paragraph extol
the virtues of nonstandard code in specifics. However, it is also for my
purposes the best HTML reference I've been able to find. ("It sucks
least.")
It also lists the "Q" tag, which is part of my "quote test" file. Still
only supported by Lynx among the browsers in my test suite.
At one point, O'Reilly had an "upgrade path", giving some form of discount
for getting newer versions of their books. I don't know if they still do
this, but you might look into it.
Jerry Stratton
<phil_st...@cieux.zzn.com> wrote:
>ne...@hoboes.com (Jerry Stratton) writes:
>> 1. Do the expressions ‘ ’ “ ” exist at all?
>
>The problem with this question is that the expressions "exist," but they
>are undefined. (You may remember from junior high that "Division by zero is
>undefined.") So the question "Do they exist" is not meaningful, as the
>expressions exist, but -- like dividing by zero -- they are undefined.
Remember that the heated end of this discussion was in response to the
categorical "they don't exist at all", and the part of my posting that you
took out was "If you want to keep talking about the post, rather than
talking about standards, answer these questions". Obviously from that, I
agree that the question of existence is not meaningful in the more
interesting discussion.
>expressions are used by some operating systems as non-printing characters,
>such as control codes for printers (backspace, line feeds, and so forth,
>for example). While you may have found that the expressions show correctly
>in all your tests, I will tell you that I never see curly quotes or curly
>apostrophes on my terminal.
So what do you see on the web with these entities? In my test suite, the
only browser that does not render curly quotes renders straight quotes.
That's a fine downscaling.
>> non-standard was adhered to. Standards are in a pretty sorry state in
>> HTML, unfortunately.
>
>Well, I would disagree. In fact, I would suggest that you have it
>backwards. Standards in HTML are in fine condition. It is the _browsers_
>that are in a sorry state. Since the expressions you are using are
Standards don't read Oscar Wilde. I have to write for browsers. If I could
use <q> and reasonably expect it to work for me, I would. But <q> only
downscales (visually) in one browser in my test suite (Lynx) and it
downscales horribly in all the rest: no quotes at all, making dialogue
blend in with the rest of the stories. Using the nonstandard entities
downscales fine in Lynx to straight quotes, but keeps curly quotes in
Netscape/Mac, Netscape/Windows, Netscape/Unix, Explorer/Mac,
Explorer/Windows, WebTV, and iCab/Mac.
I am not going to publish works without differentiating between opening
and closing quotes. Not just because it doesn't read as well, but also
because I can always change my mind easily if I have the differentiation,
but not if I don't. If tomorrow Netscape and WebTV start supporting <q>, I
can easily change every left quote to <q> and every right quote to </q>.
If I'm using straight quotes, I have to come up with a complex regular
expression that may end up with unwanted consequences in one of my
thousands of pages that I'm not going to notice until months later.
>No problem from my point of view. You are aware that the expressions are
>undefined, you have found that they show up on all the combinations of
>browser, hardware, and operating systems you care about, and you do not
>care that the expressions are undefined by the standard. So leave
Actually, I *do* care that they are undefined. I would much rather that
they be defined. In fact, I would much rather be able to reasonably use
the even more informative <q>. But the choice isn't between a defined
solution and an undefined solution, it's between an undefined solution and
no solution. I don't see straight quotes as a solution.
>I just wish people would stop cross-posting to comp.sys.mac.comm, but I,
>too, am doomed to failure in getting my wish. Such is life.
Well, at least this part of the discussion is currently solely in ciwah.
You'd better re-examine that ring you found in the gutter yesterday. Maybe
you can get browsers to start supporting standards :*)
nlilavois at interactive-media dot com
>And it is still the same fallacy. Regardless how the browsers behave when
>fed those numerical character references, they are still explicitly
>undefined by the specification and must not be used. The Unicode
>numerical references &8216;, &8217;, &8221;, and &8220; are the correct
>ones, they are the standard ones, and they do work. There's no excuse for
>using the invalid numbers when the valid ones are there.
You know, of course, that this entire thread could have been avoided
if instead of insulting the man you could have simply said "That
syntax is not quite right- the preferred syntax is ‘ ’
” and “ These entity references will work on all of the
same platforms and browsers as the ones you listed, but these are
standardized."
Perhaps you could take some time off from studying markup standards
and learn a little civility.
--
/*====================================================================*\
|| Nick Lilavois | Interactive Media Corporation ||
|| nick(at)lilavois.com | nlilavois(at)interactive-media.com ||
|| http://www.lilavois.com/nick/ | http://www.interactive-media.com/ ||
\*====================================================================*/
Jerry Stratton
>And it is still the same fallacy. Regardless how the browsers behave when
>fed those numerical character references, they are still explicitly
>undefined by the specification and must not be used. The Unicode
>numerical references &8216;, &8217;, &8221;, and &8220; are the correct
>ones, they are the standard ones, and they do work. There's no excuse for
>using the invalid numbers when the valid ones are there.
1. Doesn't work in WebTV
2. Doesn't work in Netscape 2.02/Mac.
3. Doesn't work in Internet Explorer 3.02/Windows.
4. Doesn't work in Sherlock 2.0/Mac.
The undefined numbers work in all of these.
Further, the defined numbers do not downscale well in any of those
browsers, but appear as question marks.
So, regardless of how browsers behave, your recommendation is that we
should use lower-propagation non-scaling standards rather than
higher-propagation non-standards for things considered by the publisher
(me) as, at the very least, useful for reading comprehension?
The ball's in your court, Spiro.
nlilavois at interactive-media dot com
That definitely changes the whole discussion- the only logical thing
to do is what works but is not standardized, not what is standardized
but does not work.
Arjun Ray
ne...@hoboes.com (Jerry Stratton) wrote:
| 1. Do the expressions ‘ ’ “ ” exist at all?
As strings of six characters each, the expressions exist. They're
character references. However, the referents are undefined. Which
means, they have no "correct" rendering; in fact, a coredump is just
as correct a rendering as any other.
| 2. What is the relevance of pointing to the *unencoded* versions and
| saying that they don't appear correctly in his *newsreader*?
He didn't say anything about correct appearance. He merely reported
what the results were, to him. The point, of course, was that such
results were *just as "correct"* as the ones you expected others to
see. (Why else did *you* put the "unencoded" versions in your post?)
| Standards are in a pretty sorry state in HTML, unfortunately.
And the conformance?
| as soon as browsers stop supporting them (unlikely) or a better
| alternative appears that preserves the information in them, I will
| make a global change to my databases. Until then, they stay.
Congratulations. You've decided to be part of the problem rather than
part of the solution. Naturally, nothing beats persisting with legacy
data to support pious platitudes about the state of standards, right?
:ar
Jerry Stratton
>In <newsw-10099...@cx38767-a.dt1.sdca.home.com>,
>ne...@hoboes.com (Jerry Stratton) wrote:
>
>| 1. Do the expressions ‘ ’ “ ” exist at all?
>
>As strings of six characters each, the expressions exist. They're
Thank you. That was the limit of the original posting. "These do not exist
at all."
>| 2. What is the relevance of pointing to the *unencoded* versions and
>| saying that they don't appear correctly in his *newsreader*?
>
>He didn't say anything about correct appearance. He merely reported
>what the results were, to him. The point, of course, was that such
>results were *just as "correct"* as the ones you expected others to
>see. (Why else did *you* put the "unencoded" versions in your post?)
I didn't expect my posting to be considered a web page? This was a thread
about special characters in web pages, not special characters in news
postings. I'll concede that Andreas might have not known what he was
talking about because he believed we were talking about specially encoding
apostrophes in news postings, but I doubt that was the case.
>| Standards are in a pretty sorry state in HTML, unfortunately.
>
>And the conformance?
In some cases (and this is one) it is better to adhere to the nonstandard,
which works, than the standard, which doesn't.
>| as soon as browsers stop supporting them (unlikely) or a better
>| alternative appears that preserves the information in them, I will
>| make a global change to my databases. Until then, they stay.
>
>Congratulations. You've decided to be part of the problem rather than
>part of the solution. Naturally, nothing beats persisting with legacy
>data to support pious platitudes about the state of standards, right?
What legacy data? Legacy implies two things: first, that something better
has come along. What is it? Second, that it is difficult to convert the
legacy data to the modern, better format. There have been two standards
proposed over this discussion, and as soon as either one of them works
reliably, it will be trivial to convert the working, nonstandard format to
the standard.
Straight quotes are closer to being legacy data than nonstandard curly
quotes. Something better, imo, *has* come along, and it is difficult to
convert straight quotes/apostrophes in thousands of documents to curly
quotes/apostrophes--standard or nonstandard--reliably. Much better to
create documents now using the nonstandard that currently works and be
better able to switch to the standard when it does work.
Or are you claiming that we should switch to standards that break current
browsers solely for the sake of adhering to standards?
Alan J. Flavell
> That definitely changes the whole discussion- the only logical thing
> to do is what works but is not standardized, not what is standardized
> but does not work.
The only logical thing to do is neither.
If you suppose that the non-standard _works_, how do you explain
what is written at http://www.fourmilab.ch/webtools/demoroniser/ ?
Nick Lilavois
>On Sat, 11 Sep 1999, it was written:
>
>> That definitely changes the whole discussion- the only logical thing
>> to do is what works but is not standardized, not what is standardized
>> but does not work.
>
>The only logical thing to do is neither.
No- the logical thing is to do what WORKS. We are in the real world
dealing with real projects for real clients- we need real results. It
is necessary to test a solution on all of the viable platforms and see
that it WORKS- even if it does not have the stamp of approval from any
standards body. Frames and color attributes existed long before they
were accepted- but they WORKED. Tables were designed to show data in
grids not layout pages, but they WORKED. THe img tag was not intended
to hold a one pixel transparent GIF to use as a tab, but it WORKED.
What was standardized for this was a <q> tag, but it doesn't WORK.
This is just another one of those real world vs fantasy discussions.
In the best of all possible worlds, do what is standardized- not
simply because it is standardized, but because by being standardized
it is more likely to WORK. If the standards don't help, then we must
do what every programmer does- problem solve.
>If you suppose that the non-standard _works_, how do you explain
>what is written at http://www.fourmilab.ch/webtools/demoroniser/ ?
Simple- the person who wrote that article didn't do his homework.
(Not that I am defending MSWord HTML output- just the character entity
references under discussion) As Jerry said:
> ‘ through ” work on every browser on every platform that
> I've seen, from Lynx on basic terminals (it converts them to straight
> quotes, to Explorer and Netscape on Windows and Macintosh, Netscape on
> Linux and SunOS, iCab on Macintosh, and WebTV.
Unless Bill Gates made any recent acquisitions with the spare change
in his sofa, Linux, Mac and Sun were not Microsoft platforms, and they
WORK.
gr...@apple2.com
nlil...@interactive-media.com wrote:
> You know, of course, that this entire thread could have been avoided
> if instead of insulting the man you could have simply said "That
> syntax is not quite right- the preferred syntax is ‘ ’
> ” and “ These entity references will work on all of the
> same platforms and browsers as the ones you listed, but these are
> standardized."
>
> Perhaps you could take some time off from studying markup standards
> and learn a little civility.
Wouldn't have helped. He was already an ass before I got involved.
--
Nine quadrillion, nine hundred ninety-nine trillion, nine hundred ninety-nine
billion, nine hundred ninety-nine million, nine hundred ninety-nine thousand,
two hundred twenty-eight bottles of beer. You take one down, pass it around,
Nick Kew
> as to demonstrate the cluelessness of the author.
>
> Unfortunately, it seems they've now decided to fix up the error
> silently, and do what they assume the author intended. That doesn't
> make it right.
But Netscape hasn't. Sure, it's changed from displaying nothing at all
to displaying a question mark. But either way it's ugly, and shows up
the author's illiteracy.
> http://www.fourmilab.ch/webtools/demoroniser/
It's good when people release useful tools like this.
--
Nick Kew
Uriel Wittenberg
You are a nuisance. Your posts are really on a par with spam. It is no wonder
you are on record as expressly supporting harassment spam. All you seem to do
here is distract and divert people trying to make progress. Could you not
somehow ... disappear? Take a long break from your intense love affair with
HTML, which appears to be too strenuous for you?
Jerry Stratton has demonstrated quite abundantly in this thread that he very
much knows what he's talking about. His points are thoroughly reasonable. On
your return from vacation I think you should emulate him.
We don't need your emotional campaign to understand that standards are
desirable. There is no reason why Jerry's opting for the only solution that
WORKS for a high proportion of (all?) the browsers in current use should do
anything to hinder standardization.
I'm not sure if this thread has addressed the question, why doesn't the standard
simply endorse what so many browsers do with these characters? Perhaps somebody
(else) could comment.
If there's something terribly wrong about using ‘ etc. for quotation marks,
then as more or all browsers support the standard, Jerry and others will be
motivated to switch to the standard. It sounds like Jerry would be quite capable
of transforming his documents with a few keystrokes.
In case you're not understanding me, let's note a few instances of your
insincerity and misplaced sarcasm in your post:
"Alan J. Flavell" wrote:
> On Fri, 10 Sep 1999, Jerry Stratton wrote:
>
> > >The expressions ‘ ’ “ ” do not exist at all.
> > >Please don't post advice if you don't know what you are talking about.
> >
> > Please don't respond if you don't know what you're talking about.
>
> Of course he knows what he's talking about. It's you who pathetically
> believes that HTML is defined by what can be discovered from experiments
> on browsers, rather than reading the specifications.
Now, Mr. F, you must know this allegation about Jerry's "pathetic belief" is
simply untrue. He evidently knows quite well what HTML is.
> And clearly your experience of standards-conforming browsers is somewhat
> limited. See for example the comments at
> http://www.fourmilab.ch/webtools/demoroniser/
I would comment on this ... but don't have the time. I've already in the past
chased info you have a habit of directing people to in this manner, and have
found the links irrelevant or unclear. If you have a point to make, have the
courtesy to make it in your post.
> > and make a big differenc in readability, especially of long documents
> > or documents with dialogue.
>
> Then write the damned things correctly.
This encapsulates your nuisance quality quite well: a measured, appropriate
remark is made (yes, curly quotes are better), and your response is puerile,
oafish, vacuous.
> > If you paid attention to the posting or knew what you were talking about,
> > you would have realized the posting was not in HTML, it was *about* HTML.
>
> We see that you're actually proud of your ignorance, and delight in
> abusing people who do know what they're talking about.
Here is that familiar trait, intimating you have secret knowledge that reveals
the other person's ignorance. But Jerry is obviously right. It was an infantile
diversion to carp about the irrelevant issue of how newsreaders display these
characters.
Etc.
Do us a favor, Mr F. Take a break.
Murray Spork
wrote:
[snip expose of the "evil" Alan Flavell]
Uriel -- has Alan has given factually incorrect information or been
off-topic in the thread in question?
So what are you complaigning about? His manner? Boo-hoo. Grow up and
learn to deal with it.
Alan made a wholly valid point about the difference btw specification
and implementation. If you don't think that is important you are
wholly entitled to ignore the standards -- but please don't assume
that the rest of the world is not interested in standards.
Murray
Paul Kienitz
are in the range of 2018 through 201F hex, what browsers support using these
correct codes?
It does work in Netscape 4.61 under original W95, I find, except that 201B and
201F display as question marks.
In looking through my fonts, I find several where the Microsoft extended
characters (128 through 159) are not present... except for the four quote
marks. When I tried a font without the quote marks, none of the encodings
worked.
I wonder how likely it is to turn out that using “ and ” work just as
well as “ and ”? If it does, it would be the more correct and durable
way to go.
Uriel Wittenberg
> I wonder how likely it is to turn out that using “ and ” work just as
> well as “ and ”? If it does, it would be the more correct and durable
> way to go.
But Jerry has already addressed this earlier in this thread:
>undefined by the specification and must not be used. The Unicode
>numerical references &8216;, &8217;, &8221;, and &8220; are the correct
>ones, they are the standard ones, and they do work. There's no excuse for
>using the invalid numbers when the valid ones are there.
1. Doesn't work in WebTV
2. Doesn't work in Netscape 2.02/Mac.
3. Doesn't work in Internet Explorer 3.02/Windows.
4. Doesn't work in Sherlock 2.0/Mac.
The undefined numbers work in all of these.
Further, the defined numbers do not downscale well in any of those
browsers, but appear as question marks.
Uriel Wittenberg
> On Sun, 12 Sep 1999 16:21:19 -0400, Uriel Wittenberg <u...@urielw.com>
> wrote:
>
> [snip expose of the "evil" Alan Flavell]
>
> Alan made a wholly valid point about the difference btw specification
> and implementation.
Please -- quote it.
Arjun Ray
| Paul Kienitz wrote:
|
| > I wonder how likely it is to turn out that using “ and ”
| > work just as well as “ and ”?
Unfortunately, that's not a well-posed question, because it conflates
two different meanings of "work".
The Unicode values work in the sense that a browser has to support
them: it not only recognizes them, but has enough in the way of local
system resources (installed fonts, knowledge of local character sets,
fallback substitutes, etc.) to render them. Not recognizing them at
all - that is, irrespective of whether they can be rendered - is just
a conformance issue; the fact remains that these character references
refer to *known* values and thus have *defined* meaning.
OTOH, byte values in the range 128-159 are *undefined*. That is, it
is impossible to say what they mean, because they do not have known
referents. So, any observed "working" is entirely an accident, and by
the same token any *other* "working" (due to a change in configuration
say) is not a "mistake" - and not a conformance issue either.
| > If it does, it would be the more correct and durable way to go.
This is a matter of how widespread conformance is (or becomes.) The
practical point is that using a fallback instead is advisable if
failure can't be tolerated.
| But Jerry has already addressed this earlier in this thread:
Yes, in all too typical cargo-cult fashion.
| 1. Doesn't work in WebTV
| 2. Doesn't work in Netscape 2.02/Mac.
| 3. Doesn't work in Internet Explorer 3.02/Windows.
| 4. Doesn't work in Sherlock 2.0/Mac.
|
| The undefined numbers work in all of these.
They work *by accident*. They do not work by design. They do not
work reliably - because reliability by definition cannot apply to the
range 128-159.
But cargo-cultists don't care.
:ar
Arjun Ray
ne...@hoboes.com (Jerry Stratton) wrote:
| In <37dff446...@news.supernews.com>, ar...@interactrx.com wrote:
| >In <newsw-10099...@cx38767-a.dt1.sdca.home.com>,
| >ne...@hoboes.com (Jerry Stratton) wrote:
| >| 2. What is the relevance of pointing to the *unencoded* versions and
| >| saying that they don't appear correctly in his *newsreader*?
| >
| >He didn't say anything about correct appearance. He merely reported
| >what the results were, to him. The point, of course, was that such
| >results were *just as "correct"* as the ones you expected others to
| >see. (Why else did *you* put the "unencoded" versions in your post?)
|
| I didn't expect my posting to be considered a web page?
When Andreas wrote:
: >’ น
: ^ This is a superscript 1.
His caret was pointing to a character that arrived on usenet as byte
decimal 185. In ISO-8859-1, that's superscript 1. Would you care to
explain *why* you bothered to post an "unencoded" version at all?
What did you see? In your newsreader? In your wowser? What do you
expect to see in your newsreader or wowser when you give it a byte
decimal 185?
You mean you didn't intend to send byte decimal 185 to usenet? Gee,
somthing musta happenned?
Any clue how? Any clue why?
| Or are you claiming that we should switch to standards that break
| current browsers solely for the sake of adhering to standards?
Good grief.
:ar
gr...@apple2.com
>> I wonder how likely it is to turn out that using “ and ” work
>> just as well as “ and ”? If it does, it would be the more correct
>> and durable way to go.
> But Jerry has already addressed this earlier in this thread:
>> undefined by the specification and must not be used. The Unicode
>> numerical references &8216;, &8217;, &8221;, and &8220; are the correct
>> ones, they are the standard ones, and they do work. There's no excuse for
>> using the invalid numbers when the valid ones are there.
You really shouldn't mix-and-match text in followups like that. You've
made it look like either Paul or Jerry wrote the above and you wrote what
is below. Text included from another source should be quoted in a
distinctive fashion (such as with "| " instead of "> ") and have its own
attribution lines.
> 1. Doesn't work in WebTV
Boo hoo. That's why I use " for quotes and ' for apostrophes. They work
in everything.
> 2. Doesn't work in Netscape 2.02/Mac.
Twice obsolete, memory leaks, security holes, and crashes. Particularly
when dealing with frames (which I avoid anyway).
> 3. Doesn't work in Internet Explorer 3.02/Windows.
Twice obsolete. Security holes. Twice evil. :-)
> 4. Doesn't work in Sherlock 2.0/Mac.
Displays search engine results, not pages. Use a META tag for summaries
which use safe " and ' and you don't have to worry.
> The undefined numbers work in all of these.
You mean all of those are uniformly broken in their recognition of the
undefined numbers.
> Further, the defined numbers do not downscale well in any of those
> browsers, but appear as question marks.
There's also ‘ ’ “ and ”, but I'd think you'd
prefer the question marks (or slugs) to those being displayed in the
clear, since their support is very dismal.
If it isn't worth doing right, it isn't worth doing at all. I stick with
" and ' and have never had a complaint.
--
Nine quadrillion, nine hundred ninety-nine trillion, nine hundred ninety-nine
billion, nine hundred ninety-nine million, nine hundred ninety-nine thousand,
two hundred fourteen bottles of beer. You take one down, pass it around, nine
Paul Mitchum
wrote:
>It occurrs to me to wonder... since the correct unicode values for curly quotes
>are in the range of 2018 through 201F hex, what browsers support using these
>correct codes?
iCab, for one.
Boris Ammerlaan
>"Alan J. Flavell" <fla...@mail.cern.ch> wrote:
>
>>On Sat, 11 Sep 1999, it was written:
>>
>>> That definitely changes the whole discussion- the only logical thing
>>> to do is what works but is not standardized, not what is standardized
>>> but does not work.
>>
>>The only logical thing to do is neither.
>
>No- the logical thing is to do what WORKS.
That might be "the" practical thing to do, but not the logical.
> We are in the real world
>dealing with real projects for real clients- we need real results.
I.e.: we need to deceive our clients.
> It
>is necessary to test a solution on all of the viable platforms and see
And who is to decide what platforms are viable? You, who cannot
know every possible platform/browser combination? The client,
who is relying on *your* expertise?
>that it WORKS- even if it does not have the stamp of approval from any
>standards body. Frames and color attributes existed long before they
>were accepted- but they WORKED.
With a few dozen caveats and preconditions, maybe.
>What was standardized for this was a <q> tag, but it doesn't WORK.
Of course it does not. But under some specifications, you
are allowed to insert appropriate quotes into the Q
elements.
>This is just another one of those real world vs fantasy discussions.
>In the best of all possible worlds, do what is standardized- not
>simply because it is standardized, but because by being standardized
>it is more likely to WORK. If the standards don't help, then we must
>do what every programmer does- problem solve.
Then use something that does work, and that falls back gracefully -
curly quotes do not with those undefined nce's.
--
Boris Ammerlaan <bo...@stack.nl>, http://www.stack.nl/%7Eboris/
* HTML FAQ: posted bi-weekly & <URL:http://www.htmlhelp.com/faq/html/>
* c.i.w.a.h. FAQ List Pointer: posted twice a week &
<URL:http://www.stack.nl/%7Eboris/HTML/ciwahfaq.html>
Bertil Wennergren
Jukka Korpela:
> >Those characters have
> >different encodings on different computer platforms, but browsers will
> >render them correctly regardless of platform.
> There is no applicable definition about "correctness", since ‘ is
> undefined.
Well, if they're undefined, then _any_ rendering or non-rendering is
correct, and thus any browser (or toaster...) on any platform (or anything
else) does render them correctly (at least if not being incorrect means
being correct...).
Nitpicking aside, I think that it's always best to do what both works and
is correct. In this case that means using ordinary ASCII-quotes.
--
=====================================================================
Bertilo Wennergren
<bert...@hem1.passagen.se>
<http://purl.oclc.org/net/bertilo>
=====================================================================
Paananen Tero
>Mr. Flavell:
>You are a nuisance. Your posts are really on a par with spam.
Mr. Wittenberg,
Learn to use your killfile.
HTH
-TPP
Boris Ammerlaan
>I didn't expect my posting to be considered a web page? This was a thread
>about special characters in web pages, not special characters in news
>postings.
What is the difference?
>What legacy data? Legacy implies two things: first, that something better
>has come along.
Well, no. The "legacy" part in "legacy data" implies that it was
created earlier, and that you have to work with it.
> What is it?
Erm... HTML 2.0?
> Second, that it is difficult to convert the
>legacy data to the modern, better format.
Why "difficult"? At most, that it has not yet been converted.
>Straight quotes are closer to being legacy data than nonstandard curly
>quotes.
Wrong. Following your definition of legacy data, _your_ curly quotes
are to be considered legacy data, since in a WWW context it is
difficult to convert them to the standard curly quotes.
>Something better, imo, *has* come along, and it is difficult to
>convert straight quotes/apostrophes in thousands of documents to curly
>quotes/apostrophes--standard or nonstandard--reliably.
Hmm... use sed. Should not be difficult.
Philip Stripling
> Murray Spork wrote:
>
> > On Sun, 12 Sep 1999 16:21:19 -0400, Uriel Wittenberg <u...@urielw.com>
> > wrote:
> >
>SNIP<
Is it possible to trim comp.sys.mac.comm from this spat?
--
Phil Stripling | email to the replyto address is presumed
The Civilized Explorer | spam and read later. email to philip@
http://www.cieux.com/ | civex.com is read daily.
Andreas Prilop
gr...@apple2.com
je...@hoboes.com (Jerry Stratton) wrote:
>gr...@apple2.com wrote:
>>je...@hoboes.com (Jerry Stratton) wrote:
>>> 1. Doesn't work in WebTV
>> Boo hoo. That's why I use " for quotes and ' for apostrophes. They work
>> in everything.
> That's a valid choice for you. For those of us who are not willing to
> write without the added information of curly quotes, the choice is between
> something that works but isn't standard, and at least two things that are
> standard but don't work.
Exactly what information do curly quotes add that isn't already there?
Assuming you know where whitespace belongs relative to quotation marks, "
can signify opening and closing quotations unambiguously.
>> There's also ‘ ’ “ and ”, but I'd think you'd
>> prefer the question marks (or slugs) to those being displayed in the
>> clear, since their support is very dismal.
> Great. Now we have three standards that don't work ganging up on one
> nonstandard that does :*) (Are those really part of the standard? I've
> seen <Q> and I've seen xxx, but I've never seen the above mentioned
> anywhere.)
My immediate reference is the source code for HTML Tidy. I only mention
them to be more complete. To finish it off, there's also the hex notation
alternative, also with dismal support.
--
Nine quadrillion, nine hundred ninety-nine trillion, nine hundred ninety-nine
billion, nine hundred ninety-nine million, nine hundred ninety-nine thousand,
two hundred ten bottles of beer. You take one down, pass it around, nine
Paul Kienitz
> You mean all of those are uniformly broken in their recognition of the
> undefined numbers.
It sounds like what we have is a choice between a past-oriented method and a
future-oriented one, which is a pretty commonplace choice in lots of areas of
computing.
> There's also ‘ ’ “ and ”, but I'd think you'd
> prefer the question marks (or slugs) to those being displayed in the
> clear, since their support is very dismal.
What's the story with these? I've never seen these mentioned in the various
(admittedly limited) HTML docs I've read. Are they a recent addition? I tried
them and NS4.61 just displayed them literally.
This would be the ideal solution if it worked, since on systems like Amigas
which use strict Latin-1 charsets without any curly quotes available, it would
make the straight kind.
Paul Kienitz
> | > work just as well as “ and ”?
>
> two different meanings of "work".
What? No it doesn't.
> Yes, in all too typical cargo-cult fashion.
Haven't you guys ever heard of courtesy?
Nick Lilavois
>In <37da529d...@news.mindspring.com>, Nick Lilavois wrote:
>>"Alan J. Flavell" <fla...@mail.cern.ch> wrote:
>>
>>>On Sat, 11 Sep 1999, it was written:
>>>
>>>> That definitely changes the whole discussion- the only logical thing
>>>> to do is what works but is not standardized, not what is standardized
>>>> but does not work.
>>>
>>>The only logical thing to do is neither.
>>
>>No- the logical thing is to do what WORKS.
>
>That might be "the" practical thing to do, but not the logical.
No, when dealing with the real world and requiring a real
world solution, it is logical to be practical.
>> We are in the real world
>>dealing with real projects for real clients- we need real results.
>
>I.e.: we need to deceive our clients.
Deceive? Yes, I deceive my clients by giving them something
that works. Damn me. I bet they really wanted something that
didn't work. I guess to be truly deceptive, I could tell
them it doesn't work, when it really does.
>> It
>>is necessary to test a solution on all of the viable platforms and see
>
>And who is to decide what platforms are viable? You, who cannot
>know every possible platform/browser combination? The client,
>who is relying on *your* expertise?
Both actually- for Intranet projects (most of what I do),
the client decides. If all of their employees use MSIE5 on
98, then it only needs to work in MSIE5 on 98. For Internet
projects, statistics decide. Once again in the REAL WORLD,
we don't give a damn about *every possible platform/browser
combination*, just about 98% of them. And 98% of them are
using MSIE or Netscape, versions 3 or higher, on Windows
3.1/95/98/NT and Macintosh.
>>that it WORKS- even if it does not have the stamp of approval from any
>>standards body. Frames and color attributes existed long before they
>>were accepted- but they WORKED.
>
>With a few dozen caveats and preconditions, maybe.
Like being in the 98% of browsers and OS, yes.
>>This is just another one of those real world vs fantasy discussions.
>>In the best of all possible worlds, do what is standardized- not
>>simply because it is standardized, but because by being standardized
>>it is more likely to WORK. If the standards don't help, then we must
>>do what every programmer does- problem solve.
>
>Then use something that does work, and that falls back gracefully -
>curly quotes do not with those undefined nce's.
As the previous poster demonstrated, they DO work- even on
obscure browsers on rare platforms, while the "proper" way
of doing things won't even work in MSIE3.0 on 95, which
still has about 3%.
Jerry Stratton
>in everything.
That's a valid choice for you. For those of us who are not willing to
write without the added information of curly quotes, the choice is between
something that works but isn't standard, and at least two things that are
standard but don't work.
>> 2. Doesn't work in Netscape 2.02/Mac.
>Twice obsolete, memory leaks, security holes, and crashes. Particularly
>when dealing with frames (which I avoid anyway).
And used approximately 36 times on my site last week, not including
variants that hit less than ten times. Netscape 2 is the last Netscape
that will reasonably work under the lower-memory conditions of older
computers. I deal with a lot of people who use old Macintoshes with 10 meg
of RAM or less and use those computers to access the web. We continue to
have a demand for donating ours to grade schools. One of my most vocal
audiences is grade school students who have waited until the last minute
to read "The Importance of Being Earnest". (Why couldn't *I* have gone to
a grade school--or even a high school--that read Oscar Wilde?)
>> 3. Doesn't work in Internet Explorer 3.02/Windows.
>Twice obsolete. Security holes. Twice evil. :-)
Explorer 3 of various versions (not including the AOL variant, whatever
that is) was used approximately 1,990 times on my site last week. It was
included with a *lot* of Windows computers, and for most people with
modems, Netscape and Explorer take a long time to download.
>> 4. Doesn't work in Sherlock 2.0/Mac.
>Displays search engine results, not pages. Use a META tag for summaries
>which use safe " and ' and you don't have to worry.
This has two problems. The most basic is that while it compartmentalizes
the problem of what is to me legacy data into a smaller area, the data is
still there. " is not trivially convertable to whatever smart quote
standard ends up widely propagated. If you don't care about smart quotes,
that's fine, but I do.
Second, while I use meta tag summaries religiously, they seem to be only
rarely used by search engines. I saw my first one today in altavista, for
example. Most of the time, all I see in results that hit my pages are
lines from the page itself. And in some cases, I'm not sure that's all
wrong. I don't return them on my dedicated search engine: if the user
searches for Superman, I show them what the matches in that file were, so
they can tell if the hit mentioned Nietsche or Swan. I'm considering
showing them both, if it doesn't get to be too much on the page.
>> The undefined numbers work in all of these.
>You mean all of those are uniformly broken in their recognition of the
>undefined numbers.
That's fine, but I still want to write for the people using these broken
browsers.
>> Further, the defined numbers do not downscale well in any of those
>> browsers, but appear as question marks.
>There's also ‘ ’ “ and ”, but I'd think you'd
>prefer the question marks (or slugs) to those being displayed in the
>clear, since their support is very dismal.
Great. Now we have three standards that don't work ganging up on one
nonstandard that does :*) (Are those really part of the standard? I've
seen <Q> and I've seen xxx, but I've never seen the above mentioned
anywhere.)
>If it isn't worth doing right, it isn't worth doing at all. I stick with
>" and ' and have never had a complaint.
I don't find that in question. But I also don't believe that straight
quotes are "doing it right". If it's worth doing, it's worth doing right,
and for me that means including curly quotes that work.
I suppose it also depends on your definition of "doing it right". My goal
is to provide things for people to read. It is almost always best to
adhere to standards while doing so, because this helps ensure a wider
audience. But given my perceived need for curly quotes, this is one of the
cases where it is obviously not best to use the standards.
Jerry Stratton
>In <newsw-11099...@cx38767-a.dt1.sdca.home.com>, Jerry Stratton wrote:
>
>>I didn't expect my posting to be considered a web page? This was a thread
>>about special characters in web pages, not special characters in news
>>postings.
>
>What is the difference?
Given that the subject of this thread is "apostrophe in html", and that I
was taken to task for presenting numerical encodings that "don't exist at
all", you have to ask?
>>Something better, imo, *has* come along, and it is difficult to
>>convert straight quotes/apostrophes in thousands of documents to curly
>>quotes/apostrophes--standard or nonstandard--reliably.
>
>Hmm... use sed. Should not be difficult.
The tool isn't the issue. Determining which "s and single quotes
should be converted is. If you have a 100% solution (remember that *I*
already have one), contact Microsoft with your solution. You'll make
millions :*)
Jerry Stratton
>In <newsw-11099...@cx38767-a.dt1.sdca.home.com>,
>ne...@hoboes.com (Jerry Stratton) wrote:
>When Andreas wrote:
>
>: >’ น
>: ^ This is a superscript 1.
>
>His caret was pointing to a character that arrived on usenet as byte
>decimal 185. In ISO-8859-1, that's superscript 1. Would you care to
>explain *why* you bothered to post an "unencoded" version at all?
>
>What did you see? In your newsreader? In your wowser? What do you
>expect to see in your newsreader or wowser when you give it a byte
>decimal 185?
>
>You mean you didn't intend to send byte decimal 185 to usenet? Gee,
>somthing musta happenned?
Thanks for answering the question for me, even if you are completely
wrong. I would have thought the reason that I posted byte decimal 185 in a
message to a person posting on a Macintosh newsgroup about apostrophes
would be obvious, but apparently not. Since it appears to be the only
leverage I hold, if I let you in on the secret, will you let me know why
decimal 185 not appearing correctly in a usenet posting is relevant to my
lack of knowledge on "apostrophe in html"?
Arjun Ray
je...@hoboes.com (Jerry Stratton) wrote:
| In <37f89eda....@news.supernews.com>, ar...@interactrx.com wrote:
| > What do you expect to see in your newsreader or wowser when you
| > give it a byte decimal 185?
| >
| >You mean you didn't intend to send byte decimal 185 to usenet? Gee,
| >somthing musta happenned?
|
| Thanks for answering the question for me, even if you are completely
| wrong.
Thanks for proving that you're a cargo-cultist. Next time you post
pieties about standards and conformance, I hope I'll remember to plonk
you.
| I would have thought the reason that I posted byte decimal 185 in a
| message to a person posting on a Macintosh newsgroup about apostrophes
| would be obvious, but apparently not.
Ah yes. Of course. The usual DWIM.
| if I let you in on the secret, will you let me know why decimal 185
| not appearing correctly in a usenet posting
^^^^^^^^^
This is usenet. Which planet are you on?
| is relevant to my lack of knowledge on "apostrophe in html"?
As I said, it proves that you're a cargo-cultist. Not a care beyond
"Hey, it worked for me!". Not merely clueless, but savagely clueless
and insolently proud of it.
:ar
Arjun Ray
wrote:
| > Unfortunately, that's not a well-posed question, because it
| > conflates two different meanings of "work".
|
| What? No it doesn't.
Yes it does. It conflates "work by design" and "work by accident".
These differ radically in their negations, and so are not the same.
| > Yes, in all too typical cargo-cult fashion.
|
| Haven't you guys ever heard of courtesy?
Not to fools who claim to know better.
:ar
Arjun Ray
Paul Kienitz <k...@pacbell.net> wrote:
| > There's also ‘ ’ “ and ”, but I'd think
| > you'd prefer the question marks (or slugs) to those being displayed
| > in the clear, since their support is very dismal.
|
| What's the story with these? I've never seen these mentioned in the
| various (admittedly limited) HTML docs I've read. Are they a recent
| addition?
They're among the entities defined for HTML 4.0. See
<URL:http://www.w3.org/TR/REC-html40/sgml/entities.html#h-24.4>
(Most of these have been around for quite a while, in the sense that
these "standardized" names have been known. Only their incorporation
into HTML is recent.)
| I tried them and NS4.61 just displayed them literally.
That's the "recommended" behavior, as per RFC 1866, for unrecognized
entity references. There's a little bit of a fudge here, in that it
could be useful to distinguish between entity references that *should*
be recognized and truly undefined references. But since HTML browsers
don't even know how to refer to DTDs, they're restricted to the names
hard-coded into them.
(For instance, if you look at the entity declarations in the URL cited
above, you'll notice that the definitions of the names resolve to
Unicode character references. It's conceivable that a browser has a
greater repertoire of character references it can handle appropriately
than of names it recognizes.)
| This would be the ideal solution if it worked, since on systems like
| Amigas which use strict Latin-1 charsets without any curly quotes
| available, it would make the straight kind.
Exactly. Allowing for *intelligent* local translation or fallback.
:ar
Jerry Stratton
>Assuming you know where whitespace belongs relative to quotation marks, "
>can signify opening and closing quotations unambiguously.
" and '.
Can you provide a regular expression or other completely automated means
of converting *all* straight double and single quotes and apostrophes to
their correct (assuming workable standard in the future) curliness? It
must be able to work without any human intervention other than setting the
process in motion.
(I assume you already know the answer to the first question, since you
alluded to it in the next two lines...)
Jerry Stratton
>In <jerry-13099...@cx38767-a.dt1.sdca.home.com>,
>je...@hoboes.com (Jerry Stratton) wrote:
>>if I let you in on the secret, will you let me know why decimal 185
> ^^^^^^^^^
>>lack of knowledge on "apostrophe in html"?
>
The planet in which this thread started because someone told me I "didn't
know what I was talking about" when replying to a question about
apostrophes in HTML, because those numeric entities "don't exist at all",
and the appearance of non-HTML characters in a non-HTML posting in a
non-HTML environment somehow proved this.
Perhaps if you were to explain your reason for being in this thread I
might be able to better assist you.
Robert Gormley
Arjun Ray <ar...@nmds.com> wrote in message
news:37efd106....@news.supernews.com...
> In <37DDCD70...@pacbell.net>, Paul Kienitz <k...@pacbell.net>
> wrote:
>
> | > Unfortunately, that's not a well-posed question, because it
> | > conflates two different meanings of "work".
> |
> | What? No it doesn't.
>
> Yes it does. It conflates "work by design" and "work by accident".
> These differ radically in their negations, and so are not the same.
They still work. End of story. Whether or not they work 'by design' or 'by
accident' or 'act of god' or 'simply in spite of developer stupidity' does
not invalidate the point.
> | > Yes, in all too typical cargo-cult fashion.
> |
> | Haven't you guys ever heard of courtesy?
>
> Not to fools who claim to know better
And yet you berate others who do the same.
Jukka....@hut.fi
wrote:
>For those of us who are not willing to
>write without the added information of curly quotes, the choice is between
>something that works but isn't standard, and at least two things that are
>standard but don't work.
The first option does not exist. I think I have mentioned my document
http://www.hut.fi/u/jkorpela/www/windows-chars.html
which contains actual screen shots where the way you are most probably
referring to as "working" fails miserably. And they are not the worst
case. (The behavior illustrated in the screen shots is the most
common, though not the only one, browser treatment of the characters
listed in the document when the browser does not happen to support the
Windows-specific ways under discussion.)
Are you willing to take the risk of not having _any_ quotation marks
or apostrophes displayed, or something worse?
--
Yucca, http://www.hut.fi/u/jkorpela/ | http://yucca.hut.fi/yucca.html
To join the Clueless Club, send a followup to this message
quoting everything up to and including this sig!
Jerry Stratton
>On Tue, 14 Sep 1999 02:52:14 GMT, je...@hoboes.com (Jerry Stratton) wrote:
>>For those of us who are not willing to
>>write without the added information of curly quotes, the choice is between
>>something that works but isn't standard, and at least two things that are
>>standard but don't work.
>The first option does not exist. I think I have mentioned my document
>http://www.hut.fi/u/jkorpela/www/windows-chars.html
>which contains actual screen shots where the way you are most probably
I can't find any screen shots. I see very few images on that page. I also
see no mention of a specific browser, other than the vague "many others
get". Are you sure that's the right page?
>referring to as "working" fails miserably. And they are not the worst
>case. (The behavior illustrated in the screen shots is the most
>common, though not the only one, browser treatment of the characters
>listed in the document when the browser does not happen to support the
>Windows-specific ways under discussion.)
>
>Are you willing to take the risk of not having _any_ quotation marks
>or apostrophes displayed, or something worse?
Given the choice between using straight quotes and using curly quotes that
break on some browsers, I'm going to use curly quotes and minimize the
breaking. If you can show me that something other than ‘-” breaks
less, I will switch to that alternative. So far, I have yet to see an
alternative with anything near the acceptance of ‘-”.
In this particular case (curly quotes), acceptance is not limited to
Windows. Presentation appears correct on a variety of Macintosh browsers
and at least one Unix browser (Netscape 4.6).
I have so far found only one case, other than pre-Netscape 2 browsers,
where ‘-” doesn't work. This was Sun's HotJava 1.01. HotJava
showed the ‘-8 numbers as spaces, whereas it showed the Unicode
characters as boxes. I have not found any browser that supports one of the
standard alternatives but does not support the nonstandard ‘-8. I've
seen lots of words like "many" and "may break" used, but no specific
examples.
If I create my documents now with ‘-8, it will be trivial to convert
the ‘-8 characters to whatever standard wins out. It is much less
trivial to convert straight quotes.
Arjun Ray
In <jerry-13099...@cx38767-a.dt1.sdca.home.com>,
je...@hoboes.com (Jerry Stratton) wrote:
| In article <37ecbdce....@news.supernews.com>, ar...@interactrx.com wrote:
| >In <jerry-13099...@cx38767-a.dt1.sdca.home.com>,
| >je...@hoboes.com (Jerry Stratton) wrote:
[ Restoring the original context of quote and response ]
| >| if I let you in on the secret, will you let me know why decimal 185
| >| not appearing correctly in a usenet posting
| > ^^^^^^^^^
| >This is usenet. Which planet are you on?
|
| The planet in which this thread started
Could be, but it's definitely a planet that allows you to define your
own meaning of correct appearance for a usenet posting. (Do you make
it a rule to overlook or misunderstand carets?)
| because someone told me I "didn't know what I was talking about" when
| replying to a question about apostrophes in HTML,
That's right. You didn't. You were talking about what you thought
you knew. You could have reported it as observation or personal
experience, but instead you cast it as advice.
| because those numeric entities "don't exist at all",
Which was indeed what he said. What he almost certainly meant - that
there are no defined referents - was clear to people who are familiar
with Andreas' posting history. I suppose everyone - except perhaps
you - got that clarified during the course of the thread.
| and the appearance of non-HTML characters in a non-HTML posting in
| a non-HTML environment somehow proved this.
Actually, it proved that HTML isn't the only thing you don't know much
about.
| Perhaps if you were to explain your reason for being in this thread
On usenet, it's a minor public service, shall we say, to counter
strident advocacy of bogosities.
| I might be able to better assist you.
Try learning something first. You may find these resources and the
other resources pointed to therein helpful:
<URL:http://www.hut.fi/~jkorpela/chars.html>
<URL:http://ppewww.ph.gla.ac.uk/%7eflavell/iso8859/iso8859-pointers.html>
<URL:http://ppewww.ph.gla.ac.uk/%7eflavell/iso8859/iso8859-1.html>
<URL:http://ppewww.ph.gla.ac.uk/%7eflavell/iso8859/isotable.html>
:ar
Robert Gormley
>
> In <jerry-13099...@cx38767-a.dt1.sdca.home.com>,
> je...@hoboes.com (Jerry Stratton) wrote:
> | In article <37ecbdce....@news.supernews.com>, ar...@interactrx.com
wrote:
> | >In <jerry-13099...@cx38767-a.dt1.sdca.home.com>,
> | >je...@hoboes.com (Jerry Stratton) wrote:
...
> | because someone told me I "didn't know what I was talking about" when
> | replying to a question about apostrophes in HTML,
>
> That's right. You didn't. You were talking about what you thought
> you knew. You could have reported it as observation or personal
> experience, but instead you cast it as advice.
Something no-one else here could ever be possibly guilty of...
> | because those numeric entities "don't exist at all",
>
> Which was indeed what he said. What he almost certainly meant - that
> there are no defined referents - was clear to people who are familiar
> with Andreas' posting history. I suppose everyone - except perhaps
> you - got that clarified during the course of the thread.
No. It's obvious he knows perfectly well. What he was saying was that he got
annoyed with "someone's" response that anything he said was invalid because
it 'did not exist', which is rather facetious.
...
> | Perhaps if you were to explain your reason for being in this thread
>
> On usenet, it's a minor public service, shall we say, to counter
> strident advocacy of bogosities.
Stunning arrogance. Your existence here is something you classify as a
public service? I think most people would take a more modest, and realistic
view.
Jukka....@hut.fi
wrote:
>I can't find any screen shots.
Then we have different meanings for the expression "screen shot".
> I see very few images on that page.
That's because I have included informative images only.
> I also see no mention of a specific browser, other than the vague "many others
>get". Are you sure that's the right page?
I am. Does it really matter on which browsers that happens? It was
probably some version of Netscape I used. For me, it is sufficient to
know that the undefined numeric character references do not work in
WWW terms. I personally haven't used Web browsers on Unix for quite
some time. It seems that you have never used them. This does not mean
that they are nonexistent on the Web.
>>Are you willing to take the risk of not having _any_ quotation marks
>>or apostrophes displayed, or something worse?
>
>Given the choice between using straight quotes and using curly quotes that
>break on some browsers, I'm going to use curly quotes and minimize the
>breaking - -
Your choice, but I wonder why you previously wrote that the curliness
carries some information. Here you are willing to accept a much more
serious loss of information.
>So far, I have yet to see an
>alternative with anything near the acceptance of ‘-”.
I can see that you have some difficulties in accepting the reality
that there is _no_ way to present the curly quotes in a manner which
works on the WWW. Don't shoot the messenger; I'm just explaining the
facts, I did not cause them. I'd be very happy if we could use a
richer character repertoire reliably on the Web.
>If I create my documents now with ‘-8, it will be trivial to convert
>the ‘-8 characters to whatever standard wins out. It is much less
>trivial to convert straight quotes.
It is trivial to write your documents using the _correct_ numeric
character references and programmatically convert them to use straight
quotes in the present situation. You could then make both versions
available, happily sitting down and waiting for development; the
number of browsing situations where the better presentation can be
used is considerable and increasing.
Robert Gormley
<Jukka....@hut.fi> wrote in message
news:37ddf985.2833334729@news.cs.hut.fi...
> On Tue, 14 Sep 1999 07:02:49 GMT, je...@hoboes.com (Jerry Stratton)
> wrote:
>
> >I can't find any screen shots.
>
> Then we have different meanings for the expression "screen shot".
Your quote was "actual screen shot" - I don't see one either, if I were to
be as pedantic as you.
...
> It is trivial to write your documents using the _correct_ numeric
> character references and programmatically convert them to use straight
> quotes in the present situation. You could then make both versions
> available, happily sitting down and waiting for development; the
> number of browsing situations where the better presentation can be
> used is considerable and increasing
Whilst trivial, it is also very unwise, increasing the factor of maintenance
by two. For every one document previously being modified, you now need two,
and to be keeping track of where that version is being used. Extremely
unwise. How long before someone copies/saves a file over the other, etc,
etc?
Boris Ammerlaan
>In <greg-13099...@oshar.war-of-the-worlds.org>, gr...@apple2.com wrote:
>>Exactly what information do curly quotes add that isn't already there?
>>Assuming you know where whitespace belongs relative to quotation marks, "
>>can signify opening and closing quotations unambiguously.
>
>" and '.
>
>Can you provide a regular expression or other completely automated means
>of converting *all* straight double and single quotes and apostrophes to
>their correct (assuming workable standard in the future) curliness?
Erm... 2/ "/{curly open}/g
And so on.
>It
>must be able to work without any human intervention other than setting the
>process in motion.
>
>(I assume you already know the answer to the first question, since you
>alluded to it in the next two lines...)
No, in the next two lines he explained that they did not convey any
new information.
Boris Ammerlaan
>Whilst trivial, it is also very unwise, increasing the factor of maintenance
>by two. For every one document previously being modified, you now need two,
>and to be keeping track of where that version is being used. Extremely
>unwise. How long before someone copies/saves a file over the other, etc,
>etc?
Erm, why not generate both documents from a common parent?
Only edit the parent...
Jukka....@hut.fi
<rgor...@expert.com.au> wrote:
>Your quote was "actual screen shot" - I don't see one either, if I were to
>be as pedantic as you.
Sorry for not including the entire 1280x1024 screen. I always thought
it was sufficient for the illustration of the rendering of some
constructs to include the relevant part.
>> It is trivial to write your documents using the _correct_ numeric
>> character references and programmatically convert them to use straight
>> quotes in the present situation.
- -
>Whilst trivial, it is also very unwise, increasing the factor of maintenance
>by two.
I've always said that using " and ' is the simplest and most robust
way at present, and for quite some time to come. My comment was
related to automatic conversions from the incorrect form to correct
form, and I pointed out that it is just as simple to use the correct
form and generate the other correct form from it.
As regards to maintenance factors, if such generation of various
formats would double your maintenance work, you're not investing much
on maintenance now.
Boris Ammerlaan
>bo...@snail.stack.nl (Boris Ammerlaan) wrote:
>No, when dealing with the real world and requiring a real
>world solution, it is logical to be practical.
Practical is to make sure you will not have to go through all
your documents in a few (day|month|year)s whenever a new
browser comes out.
>>> We are in the real world
>>>dealing with real projects for real clients- we need real results.
>>
>>I.e.: we need to deceive our clients.
>
>Deceive?
Yes, by suggesting that it works in all situations, or even that
you know in which situations it does not.
>didn't work. I guess to be truly deceptive, I could tell
>them it doesn't work,
It does not, in WWW terms.
> when it really does.
Only in some limited range of browsers, and even then only by accident.
>>> It
>>>is necessary to test a solution on all of the viable platforms and see
>>
>>And who is to decide what platforms are viable? You, who cannot
>>know every possible platform/browser combination? The client,
>>who is relying on *your* expertise?
>
>Both actually- for Intranet projects (most of what I do),
>the client decides. If all of their employees use MSIE5 on
>98, then it only needs to work in MSIE5 on 98.
And what if they happen to switch to Win'2000 and MSIE6? Can you
*guarantee* it will continue to work?
>For Internet
>projects, statistics decide.
No, only for intranet projects.
>Once again in the REAL WORLD,
>we don't give a damn about *every possible platform/browser
>combination*, just about 98% of them. And 98% of them are
>using MSIE or Netscape, versions 3 or higher, on Windows
>3.1/95/98/NT and Macintosh.
Heh. *Sure* they are.
>>>that it WORKS- even if it does not have the stamp of approval from any
>>>standards body. Frames and color attributes existed long before they
>>>were accepted- but they WORKED.
>>
>>With a few dozen caveats and preconditions, maybe.
>
>Like being in the 98% of browsers and OS, yes.
No, like not being able to write a frameset as a URL, or like not
specifying font attributes inside a table.
>As the previous poster demonstrated, they DO work- even on
>obscure browsers on rare platforms, while the "proper" way
>of doing things won't even work in MSIE3.0 on 95, which
>still has about 3%.
For the moment, the proper way is to not use curly quotes at all.
Toby Speight
0> In <URL:news:pAnD3.1$hj6...@vic.nntp.telstra.net>, Robert wrote:
Robert> Your quote was "actual screen shot" - I don't see one either,
Robert> if I were to be as pedantic as you.
Free clue: examine all the images. Choose the yellow one, and its
immediate successor.
>> It is trivial to write your documents using the _correct_ numeric
>> character references and programmatically convert them to use straight
>> quotes in the present situation.
Robert> Whilst trivial, it is also very unwise, increasing the factor
Robert> of maintenance by two.
Free clue: note the word "programmatically" above. Ever heard of make?
--
"You can ... sell your soul for complete control -
is that really what you need?" (Pink Floyd)
Miguel Cruz
> Can you provide a regular expression or other completely automated means
> of converting *all* straight double and single quotes and apostrophes to
> must be able to work without any human intervention other than setting the
> process in motion.
QuarkXPress does a very good job of this when importing text files. So it's
obviously possible. I don't recall offhand how it handles inches and feet,
and unfortunately I'm a few miles from my DTP machine right now.
miguel
Nick Lilavois
>In <37e09ca3...@news.mindspring.com>, Nick Lilavois wrote:
>>bo...@snail.stack.nl (Boris Ammerlaan) wrote:
>
>>No, when dealing with the real world and requiring a real
>>world solution, it is logical to be practical.
>
>Practical is to make sure you will not have to go through all
>your documents in a few (day|month|year)s whenever a new
>browser comes out.
And forethought is using a database. Then any change is just
a simple SQL statement.
>>>> We are in the real world
>>>>dealing with real projects for real clients- we need real results.
>>>
>>>I.e.: we need to deceive our clients.
>>
>>Deceive?
>
>Yes, by suggesting that it works in all situations, or even that
>you know in which situations it does not.
Here is the key- I NEVER suggested it works in all
situations, because the vast majority of clients don't give
a damn about all situations. When a project begins, target
browsers and platforms are decided up-front. If someone says
they want all platforms, they won't get anything more
advanced than a <h1> tag. And guess what? their biggest
quandary is if they need to support version 3 browsers or
not, not Lynx on an terminal.
>>didn't work. I guess to be truly deceptive, I could tell
>>them it doesn't work,
>
>It does not, in WWW terms.
It does, in REAL WWW terms.
Nanny-nanny-Boo-bo.
>> when it really does.
>
>Only in some limited range of browsers, and even then only by accident.
If "limited range"=98% sure. It's amazing how they happened
to have the same accident...
>>>> It
>>>>is necessary to test a solution on all of the viable platforms and see
>>>
>>>And who is to decide what platforms are viable? You, who cannot
>>>know every possible platform/browser combination? The client,
>>>who is relying on *your* expertise?
>>
>>Both actually- for Intranet projects (most of what I do),
>>the client decides. If all of their employees use MSIE5 on
>>98, then it only needs to work in MSIE5 on 98.
>
>And what if they happen to switch to Win'2000 and MSIE6? Can you
>*guarantee* it will continue to work?
No- I don't have to, and no one can anyway. No one can
predict what will occur with Win2000 and MSIE6- anyone who
thinks they can is fooling themselves. Microsoft might win
the anti-monopoly suit, continue to grow their monopoly
power, and dump HTML for RTF for all we know. The marvelous
thing about using a database to create WBT, is it can be
changed easily when it needs to be changed.
>>For Internet
>>projects, statistics decide.
>
>No, only for intranet projects.
Um, no- I really know what I do, you aren't in any of my
office meetings.
>>Once again in the REAL WORLD,
>>we don't give a damn about *every possible platform/browser
>>combination*, just about 98% of them. And 98% of them are
>>using MSIE or Netscape, versions 3 or higher, on Windows
>>3.1/95/98/NT and Macintosh.
>
>Heh. *Sure* they are.
Yes, they are. Have you checked any stats?
>>As the previous poster demonstrated, they DO work- even on
>>obscure browsers on rare platforms, while the "proper" way
>>of doing things won't even work in MSIE3.0 on 95, which
>>still has about 3%.
>
>For the moment, the proper way is to not use curly quotes at all.
No, once again, if your client says "I want curly quotes
Dammit! And I have seen plenty of pages that have them, so
don't tell me we can't do it!"
Hey, it's not like I haven't had arguments like that- I've
worked with people who want the whole site in Flash and I
hopelessly try to talk them out of it. But everything in
life is a balance- the damage caused by those ASCII values
for curly braces is so insignificant on so EXTREMELY RARE a
platform that it just doesn't mater.
Anyone using such a hazardous platform would by now be quite
aware that they can't view the majority of sites, which use
technology far more advanced than curly quotes.
Jerry Stratton
(Boris Ammerlaan) wrote:
>In <jerry-13099...@cx38767-a.dt1.sdca.home.com>, Jerry Stratton wrote:
>>Can you provide a regular expression or other completely automated means
>>of converting *all* straight double and single quotes and apostrophes to
>>their correct (assuming workable standard in the future) curliness?
>
>And so on.
Yeah, it's the "and so on" where things get sticky :*)
>No, in the next two lines he explained that they did not convey any
>new information.
Well, obviously he believed that they conveyed opening and closing info,
but that this info should be easily inferrable.
Nick Lilavois
>As I said, it proves that you're a cargo-cultist. Not a care beyond
>"Hey, it worked for me!". Not merely clueless, but savagely clueless
>and insolently proud of it.
Since when did "cargo-cultist" enter the popular vernacular?
Do you know what a cargo cult is? It has nothing to do with
this.
Uriel Wittenberg
> It conflates "work by design" and "work by accident".
> These differ radically in their negations, and so are not the same.
>
> | Haven't you guys ever heard of courtesy?
>
> Not to fools who claim to know better.
"differ radically in their negations"?
Are they, however, relatively similar in their affirmativeness?
By the way, what are the odds of all browsers displaying something
identically *by accident*?
I wonder what the law of gravity would look like if Newton had regarded
the common downward tendency of objects as a coincidence.
Maybe you should "conflate" some common sense into your brain. That
might alleviate your sense that you are besieged by fools.
Steve Pugh
>No, once again, if your client says "I want curly quotes
>Dammit! And I have seen plenty of pages that have them, so
>don't tell me we can't do it!"
Funny, I've never had a client ask for that. Anytime clients send me
copy with curly quotes in it I change them to straight quotes and I've
had any client even notice it.
Just a thought, why has no one suggested having the curly quotes
directly in the document, not encoded at all? With the proper charset
header of course. Considering that everything but the kitchen sink has
been mentioned in this thread...
Steve
--
"My theories appal you, my heresies outrage you,
I never answer letters and you don't like my tie." - The Doctor
Steve Pugh <st...@pugh.net> <http://steve.pugh.net/>
Jerry Stratton
>Could be, but it's definitely a planet that allows you to define your
>own meaning of correct appearance for a usenet posting. (Do you make
>it a rule to overlook or misunderstand carets?)
Thank you! I was suffering from an acute case of "l'esprit d'escalier"
last night, and I was afraid you might respond to my post in a relevant
manner, depriving me of the opportunity of using this.
Since we're obviously just going in circles, and pretty small ones at
that, I'm going to end my half of what is apparently not an html
discussion. (You are of course welcome to continue arguing with yourself
:*)
I'd like to bow out with a little story. We can call it "The Browsers Must
Be Crazy".
Once upon a time there was a poor tribe living in the Southern Desert. The
tribe had nothing to drink except something they called "brackwater",
which tasted and nourished about as you would expect from something with
that name.
One day a cargo plane's pilot misread the latitude and longitude of a
drop-off, and they dropped their cargo right in the middle of the tribe's
encampment.
There isn't supposed to be any tension or anxiety in this story, so I'll
tell you right off, the box contained bottles of fruit juices. It
contained juices mainly because I'm not going to write a story about
people drinking Coke. Can't stand the stuff myself, unless mixed heavily
with rum, which wouldn't be quite the story I'm going for.
Someone tried the mango-pineapple juice, and immediately a smile spread
across his face. Soon, everyone was drinking the juice from the sky.
Except one tribemember, who held back and shook his head.
The pilot kept making the same mistake. Every day, the cargo plane dropped
a load of fruit juice near the tribal camp. Every day, everyone except one
tribemember drank fruit juice instead of brackwater. Finally, the holdout
could stand it no longer.
"You shouldn't be drinking that!" cried tribemember Nujra. "It doesn't
exist at all! And the box is made of wood!"
"But it satisfies," said one tribesman.
"It nourishes," said another tribeswoman.
"It cools," said a tribal child.
"What does the box have to do with it?" asked another child.
"What do *you* think?" asked Nujra.
"Would you like to try some juice?" they all asked.
"No," he said, "I prefer brackwater. Juice doesn't exist. But if it did,
you should be drinking locally made tribal juice from the Tribal Standard
Juice Company."
"When did the Tribal Standard Juice Company start making juices?" everyone
asked.
"They haven't," Nujra said, "but if you're going to drink juices, that's
what you should drink."
"How can we drink juices that no one makes?" asked the people.
"That's none of my business," said Nujra. "I prefer brackwater."
So the people went to the Tribal Standard Juice Company representative.
"Why haven't you started making juices?" everyone asked her.
"We're waiting for the fruit to grow," said the Juice Company Rep. "Give
us time."
"That's cool," said the people, and they went back to drinking the juice
from the sky.
The boxes kept coming.
"You shouldn't be drinking that!" cried Nujra, as he held his nose and
drank brackwater. "It doesn't exist! And besides, the box is made of
wood!"
"What does the box have to do with it?" one child asked.
"What do *you* think?" Nujra asked.
"What should we be drinking?" everyone asked.
"You should be drinking brackwater," he said.
Everyone groaned.
"Juice doesn't exist, but if it did exist, you should be drinking Tribal
Juice from the Tribal Standard Juice Company."
"But they are waiting for the fruit to grow," everyone said.
"The fruit doesn't exist," said Nujra. "But if it did, it has already grown."
So everyone headed off to see the Tribal Standard Fruit Company representative.
"Yes," she said, "the fruit has grown."
"May we try some of your fruit juice?" they asked.
"No," she said, "it isn't ready yet. We must wait for the fruit to ripen."
"Very well," the people said, and they went back to drinking the juice
from the sky.
The boxes continued to come, filled with their fine juices.
"You shouldn't be drinking that!" cried Nujra, as everyone except him
enjoyed a fine Peach-Banana fruit juice. "It doesn't exist! And the box is
made of wood!"
One child started to ask him what the box had to do with it, but his
mother put her hand over his mouth.
"Has the fruit ripened yet?" someone asked him.
"Fruit doesn't exist," said Nujra. "But if it did, it has ripened."
So everyone marched off, skyjuice in hand, to see the Tribal Standard
Juice Company representative.
"Yes," she said, "the fruit has ripened. But we still have no juice. Now
we must press it and bottle it."
"For crying out loud," they asked, "how will we know when your juice is ready?"
"Just drop by once in a while and ask," she said.
So they did. Everyone continued to drink the juice from the sky, and
occasionally dropped in on the Tribal Standard Juice Company to see if the
Tribal Fruit Juice was ready.
One day, it was.
Soon, everyone had dropped by the Tribal Standard Juice Company to try
their new juice. The Tribal Standard Juice was very good, and cool to the
throat. It nourished. Everyone loved it!
"See!" cried Nujra to anyone in earshot whenever he had the chance. "I
told you so! That other juice didn't exist at all. And the box was made of
wood!"
"Would you like some fine, cool, nourishing Tribal Standard Juice?"
everyone asked Nujra when he said this to them.
"No," he said, "I prefer brackwater."
One cargo cultist signing off...
Boris Ammerlaan
>In article <slrn7ts7t9...@snail.stack.nl>, bo...@snail.stack.nl
>(Boris Ammerlaan) wrote:
>>In <jerry-13099...@cx38767-a.dt1.sdca.home.com>, Jerry Stratton wrote:
>>>Can you provide a regular expression or other completely automated means
>>>of converting *all* straight double and single quotes and apostrophes to
>>>their correct (assuming workable standard in the future) curliness?
>>
>>Erm... 2/ "/{curly open}/g
>>And so on.
>
>Yeah, it's the "and so on" where things get sticky :*)
Pft. All right, I'll spell them all out:
s/ "/{double curly open}/g
s/" /{double curly close}/g
s/ '/{single curly open}/g
s/' /{single curly close}/g
>Well, obviously he believed that they conveyed opening and closing info,
Ah, but then sometext{double curly open}someothertext would have
some meaning.
>but that this info should be easily inferrable.
Then it is not info, but useless data.
Doug Loss
<rgor...@expert.com.au> writes:
>
>Arjun Ray <ar...@nmds.com> wrote in message
>news:37efd106....@news.supernews.com...
>> In <37DDCD70...@pacbell.net>, Paul Kienitz
><k...@pacbell.net>
>> wrote:
>>
>> | > Unfortunately, that's not a well-posed question, because it
>> | > conflates two different meanings of "work".
>> |
>> | What? No it doesn't.
>>
>> Yes it does. It conflates "work by design" and "work by
>accident".
>> These differ radically in their negations, and so are not the
>same.
>
>They still work. End of story. Whether or not they work 'by
>design' or 'by
>accident' or 'act of god' or 'simply in spite of developer
>stupidity' does
>not invalidate the point.
>
Actually, it does. "They still work," should actually be, "They still work in
my particular browsing situation." Working by design means working to an
agreed-upon specification that one might reasonably expect to work properly on
all spec-compliant browsers, and that one will have some notice of changes to
the spec before they occur. Working by accident means that there's no
guarantee that the thing in question will work across platforms, or indeed on
previous or subsequent versions of the same platform. The point is completely
invalid.
Doug Loss The difference between the right word and
Data Network Coordinator the almost right word is the difference
Bloomsburg University between lightning and a lightning bug.
dl...@bloomu.edu Mark Twain
gr...@apple2.com
>> Exactly what information do curly quotes add that isn't already there?
>> Assuming you know where whitespace belongs relative to quotation marks, "
>> can signify opening and closing quotations unambiguously.
> " and '.
", and ` and '. Realistically, how often do you have a quotation inside a
quotation, necessitating using both?
If you're really worried, you could always use << and >> as quotation
marks. Well, << and >>. Or did you forget about them?
> Can you provide a regular expression or other completely automated means
> of converting *all* straight double and single quotes and apostrophes to
> their correct (assuming workable standard in the future) curliness?
"A question for a question. Well, you're capable of evasion, anyway."
-- Roj Blake
Does a gun to your head count?
> It
> must be able to work without any human intervention other than setting the
> process in motion.
I guess it wouldn't.
It can be done. I can do it. But I'm not going to do it for you. I see
no valid utility in such a thing, and since I've seen Netscape 2.x being
used as a justification to continue using the undefined numerical
character encodings rather than the encodings in the Unicode range (which
have defined meanings even a program can determine), and that nothing will
cause that browser to cease to exist, I doubt you ever will switch to
using the right curly quote encodings. Thus such a regexp would only
serve to help you create more invalid encodings. I refuse to help you do
so.
> (I assume you already know the answer to the first question, since you
> alluded to it in the next two lines...)
And from your evasion I assume you can't tell me "exactly what information
do curly quotes add that isn't already there" since you did not. Thus, as
Orac said, "The question is futile." If you can't answer questions put to
you except by asking another question in return to divert the burden of
proof to the questioner, there is no point in continuing this discussion.
Welcome to my killfile.
--
Nine quadrillion, nine hundred ninety-nine trillion, nine hundred ninety-nine
billion, nine hundred ninety-nine million, nine hundred ninety-nine thousand,
two hundred seven bottles of beer. You take one down, pass it around, nine
quadrillion, nine hundred ninety-nine trillion, nine hundred ninety-nine
Miguel Cruz
> Pft. All right, I'll spell them all out:
> s/ "/{double curly open}/g
> s/" /{double curly close}/g
> s/ '/{single curly open}/g
> s/' /{single curly close}/g
That fails to produce proper results in several cases:
- Back in '65 we ate meat and potatoes.
- My father never grew past 4' tall.
- Bob said, "He told me, 'shut up!'"
- Don't be ridiculous.
miguel
Panos Stokas
wrote in message news:HJlD3.33$Ef6....@vic.nntp.telstra.net...
> They still work. End of story. Whether or not they work 'by design' or 'by
> accident' or 'act of god' or 'simply in spite of developer stupidity' does
> not invalidate the point.
It only remains to be seen what the word "works" means in terms of HTML. If
it is browser behaviour, I agree. If it's the accordance to open ended (and
therefore published) specifications of an international text information
markup I think it it doesn't "work".
Choice of definition then depends on the esteem one has for the quality of
the information one is to convey. Personally, I would treasure a search
engine that would exclude such constructs from her index.
Mark Jones
news:37e09ca3...@news.mindspring.com...
> bo...@snail.stack.nl (Boris Ammerlaan) wrote:
>
> >In <37da529d...@news.mindspring.com>, Nick Lilavois wrote:
> >>"Alan J. Flavell" <fla...@mail.cern.ch> wrote:
> >>
> >>>On Sat, 11 Sep 1999, it was written:
> >>>
> >>>> That definitely changes the whole discussion- the only logical thing
> >>>> to do is what works but is not standardized, not what is standardized
> >>>> but does not work.
> >>>
> >>>The only logical thing to do is neither.
> >>
> >>No- the logical thing is to do what WORKS.
> >
> >That might be "the" practical thing to do, but not the logical.
>
> No, when dealing with the real world and requiring a real
> world solution, it is logical to be practical.
been implemented a certain way in browsers used by 95+% of
users, and the specifications say it should have been done
another way, the specs need to be changed. If they aren't
going to change the specification, a note needs to be
attached that explains that the specification does not work
in current browsers.
These guys will never accept the fact that you need to do
what is required to make the pages work as good as possible
on actual browsers that people are using.
Not some hypothetical situation that has no connection to
the reality of what people are using.
>
>
>
> >> We are in the real world
> >>dealing with real projects for real clients- we need real results.
> >
> >I.e.: we need to deceive our clients.
>
> Deceive? Yes, I deceive my clients by giving them something
> that works. Damn me. I bet they really wanted something that
> didn't work. I guess to be truly deceptive, I could tell
> them it doesn't work, when it really does.
>
None of what you are doing looks like deception to me.
I make web sites that people want to actually use and I
am practical enough to come down out of the clouds and
build web pages with features that keep the users interested.
You can only take theoretical practices so far before you
have to do something that works for as many people as
possible. The specifications mean very little if virtually
none of the users have a browser that actually follows the
letter of the specification. You have to build a page that
works for browsers in current use.
>
> >> It
> >>is necessary to test a solution on all of the viable platforms and see
> >
> >And who is to decide what platforms are viable? You, who cannot
> >know every possible platform/browser combination? The client,
> >who is relying on *your* expertise?
>
> Both actually- for Intranet projects (most of what I do),
> the client decides. If all of their employees use MSIE5 on
> 98, then it only needs to work in MSIE5 on 98. For Internet
> projects, statistics decide. Once again in the REAL WORLD,
> we don't give a damn about *every possible platform/browser
> combination*, just about 98% of them. And 98% of them are
> using MSIE or Netscape, versions 3 or higher, on Windows
> 3.1/95/98/NT and Macintosh.
>
I have started building some pages that use fairly basic
techniques with JavaScript browser detection used to
redirect to different pages for MSIE4 and NS4+ and a third
page for MSIE5. If JS is turned off, the user will stay on
the basic page. I test for the browser version and not a set
of features that might also be present on other browsers.
Since certain DHTML browser features may vary between browser
companies, I check for Netscape and Microsoft browsers only.
>
>
> >>that it WORKS- even if it does not have the stamp of approval from any
> >>standards body. Frames and color attributes existed long before they
> >>were accepted- but they WORKED.
> >
> >With a few dozen caveats and preconditions, maybe.
>
> Like being in the 98% of browsers and OS, yes.
>
Exactly, you appear to take the same approach as most people.
Do what works on the actual browsers that almost everyone is
using.
>
> >>This is just another one of those real world vs fantasy discussions.
> >>In the best of all possible worlds, do what is standardized- not
> >>simply because it is standardized, but because by being standardized
> >>it is more likely to WORK. If the standards don't help, then we must
> >>do what every programmer does- problem solve.
> >
> >Then use something that does work, and that falls back gracefully -
> >curly quotes do not with those undefined nce's.
>
> As the previous poster demonstrated, they DO work- even on
> obscure browsers on rare platforms, while the "proper" way
> of doing things won't even work in MSIE3.0 on 95, which
> still has about 3%.
The idea that people should use a technique that has been
shown to _not_ work on most browsers is strange. This type
of help is not needed.
Mark Jones
news:37de...@news.bloomu.edu...
> Actually, it does. "They still work," should actually be, "They still
work in
> my particular browsing situation." Working by design means working to an
> agreed-upon specification that one might reasonably expect to work
properly on
> all spec-compliant browsers, and that one will have some notice of changes
to
> the spec before they occur. Working by accident means that there's no
> guarantee that the thing in question will work across platforms, or indeed
on
> previous or subsequent versions of the same platform. The point is
completely
> invalid.
You need to build functional websites that work for as many
actual users as possible. Not your ivory tower view of what
we would need to do if everything followed some rigid spec.
The browser companies do not have to follow the letter of
the spec. This means that we need to build towards making
a web page work on actual browsers, not pass some DTD that
no browser in existence adheres to.
If only a few percent of people have a browser that follows
a particular spec and all the other browsers require something
that violates the spec, I will violate the spec so the page
works for almost everyone. The few people with the spec compliant
browser will suffer because of this, but better that a very
small number of people have trouble instead of a large number
of people.
Nick Lilavois
>In <37DDCD70...@pacbell.net>, Paul Kienitz <k...@pacbell.net>
>wrote:
>
>| > Unfortunately, that's not a well-posed question, because it
>| > conflates two different meanings of "work".
>|
>| What? No it doesn't.
>
>Yes it does. It conflates "work by design" and "work by accident".
>These differ radically in their negations, and so are not the same.
How about "work" as in "does what we want it to do"?
>| > Yes, in all too typical cargo-cult fashion.
>|
>| Haven't you guys ever heard of courtesy?
>
>Not to fools who claim to know better.
So you are saying we should be rude to you?
gr...@apple2.com
"Mark Jones" <ch...@sound.net> wrote:
> If only a few percent of people have a browser that follows
> a particular spec and all the other browsers require something
> that violates the spec, I will violate the spec so the page
> works for almost everyone. The few people with the spec compliant
> browser will suffer because of this, but better that a very
> small number of people have trouble instead of a large number
> of people.
You mean, "Better to annoy a few than none at all," don't you? That's the
case here when one insist on using curly quotes when ", `, and ' serve
perfectly well for _everyone_.
--
Nine quadrillion, nine hundred ninety-nine trillion, nine hundred ninety-nine
billion, nine hundred ninety-nine million, nine hundred ninety-nine thousand,
two hundred bottles of beer. You take one down, pass it around, nine
Warren Steel
> I think what would solve a lot of these debates (and not just on this issue)
> is the admission that yes, while both 'sides' agree, the question is "Does
> this situation require a real world, commonsense answer?"
There is a real world, commonsense answer to producing
quotation marks reliably on the World Wide Web. It's simple
ASCII vertical quote symbols. It's this obsession with
superficial appearance and paper typography that leads
authors to seek unsatisfactory adhockery. I haven't yet
heard anyone claim that "these quotation marks" can't be
read and understood by anyone using the Web.
--
Warren Steel mu...@olemiss.edu
Joel Shepherd
need much context...}
Mark Jones wrote:
> These guys will never accept the fact that you need to do
> what is required to make the pages work as good as possible
> on actual browsers that people are using.
You know, for S&G (you figure out: second word is "Grins"), I went to
Amazon.com to see what _they_ do. Didn't see any curly quotes, just
'"' (that should be a straight quote) and " .
Went to Wired. Straight quotes.
CNET. Straight quotes. (Also a fun error message: "Sorry, that's an
invalid template path!")
Kind of seems that if your _content_ is compelling, nobody is really
going to give a damn which way your quotes slope.
Draw your own conclusions.
-- Joel.
gr...@apple2.com
"Robert Gormley" <rgor...@expert.com.au> wrote:
>>> If only a few percent of people have a browser that follows
>>> a particular spec and all the other browsers require something
>>> that violates the spec, I will violate the spec so the page
>>> works for almost everyone. The few people with the spec compliant
>>> browser will suffer because of this, but better that a very
>>> small number of people have trouble instead of a large number
>>> of people.
>> You mean, "Better to annoy a few than none at all," don't you? That's the
>> case here when one insist on using curly quotes when ", `, and ' serve
>> perfectly well for _everyone_.
> I think that's a rather simplistic, narrow-focused view... the point could
> made equally about many other things, not all of which have your quick
> 'answer'. As for *everyone*? I seem to recall at least one person saying his
> clients asked specifically for it. Would you be telling your client that
> "These are unnecessary. My belief is that *these* will do for you"? Hmm..
I'd say, "If you insist on invalid markup on this site, then you must pay
the surcharge for a site without my company logo on the site, because I
will not damage this firm's reputation by producing a site with invalid
markup to satisfy the ill-considered whims of any client. If you aren't
willing to pay the substantial surcharge, then you can find another firm
to produce your site. From scratch."
As soon as I start my own web design company that is. Right now I'm a
computer programmer, not a web page designer. (But the program is related
to HTML authoring.)
--
Nine quadrillion, nine hundred ninety-nine trillion, nine hundred ninety-nine
billion, nine hundred ninety-nine million, nine hundred ninety-nine thousand,
one hundred ninety-seven bottles of beer. You take one down, pass it around,
Nick Lilavois
>Nick Lilavois wrote:
>
>>No, once again, if your client says "I want curly quotes
>>Dammit! And I have seen plenty of pages that have them, so
>>don't tell me we can't do it!"
>
>Funny, I've never had a client ask for that. Anytime clients send me
>copy with curly quotes in it I change them to straight quotes and I've
>had any client even notice it.
I've even had a client that complained about kerning and
anti-aliasing of text. (Hi Molly!)
>Just a thought, why has no one suggested having the curly quotes
>directly in the document, not encoded at all? With the proper charset
>header of course. Considering that everything but the kitchen sink has
>been mentioned in this thread...
I don't think that would work- perhaps if they were
foreign-language characters it would. What charset would
they be in?
Robert Gormley
Nick Lilavois <nlil...@NOSPAMinteractive-media.com> wrote in message
...
> >| > Yes, in all too typical cargo-cult fashion.
I'm curious. I wonder what your definition of cargo-cult is, the traditional
doesn't apply here...
> >| Haven't you guys ever heard of courtesy?
> >
> >Not to fools who claim to know better.
>
> So you are saying we should be rude to you?
Do unto others as you would have them do unto you ;-)
Robert Gormley
Mark Jones <ch...@sound.net> wrote in message
news:37ded...@news.sound.net...
...
> > As the previous poster demonstrated, they DO work- even on
> > obscure browsers on rare platforms, while the "proper" way
> > of doing things won't even work in MSIE3.0 on 95, which
> > still has about 3%.
> The idea that people should use a technique that has been
> shown to _not_ work on most browsers is strange. This type
> of help is not needed
I think this the crux of the issue. What works in the (yes, I know, a
cliche) "hallowed halls of academia and theory" versus "the real world" - I
remember a debate a while back where someone stated that "this group is for
WWW - if you want intranet applications, form CIIAH" - we are developing
HTML for the world wide web and are in a situation where 95%+ of the
browsers used do something (say we call it "X"), whether or not it is "spec.
correct". We cannot afford to stick to the spec out of stubbornness when we
know what we can do will work if we do it using "Y" method. This is not
callous disregard, just the reality of everyday work/life and compromise.
YES - it WOULD be ideal if all browsers supported HTML 4.0 (insert latest
spec revision here) uniformly. That's not going to happen. Why? Because
developers (browser) are people, and people cannot do things identically -
we all have our individual touch, which will affect the end result.
I think what would solve a lot of these debates (and not just on this issue)
is the admission that yes, while both 'sides' agree, the question is "Does
this situation require a real world, commonsense answer?" (and please, don't
use the rhetoric that "following spec. exactly is commonsense, nothing
else" - see above)...
Regards,
Robert
Arjun Ray
nlil...@NOSPAMinteractive-media.com (Nick Lilavois) wrote:
| Since when did "cargo-cultist" enter the popular vernacular?
Surely you're joking, Mr. Lilavois?;)
| Do you know what a cargo cult is? It has nothing to do with
| this.
<URL:http://www.elsewhere.org/jargon_search/TAG253.html>
:ar
Robert Gormley
<gr...@apple2.com> wrote in message
news:greg-14099...@lnk2-jberigan-1.binary.net...
> In article <37ded...@news.sound.net>,
Dan McGarry
news:37ded...@news.sound.net...
> I would like to see the mythical "spec-compliant browsers".
Take a look at Mozilla. The Milestone 9 build for WinXX is stable enough to
give one an idea what they're working towards. I'd say that in addtion to
complying with most every relevant finished spec, it also looks like a
promising and extensible client platform.
> You need to build functional websites that work for as many
> actual users as possible. Not your ivory tower view of what
> we would need to do if everything followed some rigid spec.
On the contrary, the the ivory tower view is that which says, "This is the
way things are today, so I can logically conclude that they will always be
this way." This in spite of the many indications to the contrary.
You might want to read Eric S. Raymond's writings
(http://www.tuxedo.org/~esr/writings/) on open source software projects,
specifically on the tendency for development (and use) of software to
cluster around what is perceived to be the best alternative. As more people
use (and develop for) this 'best of breed', it tends to perpetuate -- even
increase -- its own popularity. Mozilla is a strong candidate for such
treatment. It's a pretty well thought-out 1.0 version of an application that
everyone uses.
Given the above, I'd say keeping at least one eye on the current published
standards will position you well not only now, but in the future as well.
> The browser companies do not have to follow the letter of
> the spec. This means that we need to build towards making
> a web page work on actual browsers, not pass some DTD that
> no browser in existence adheres to.
There is a great deal of HTML 4.0 that is supported across every major
browser.
> If only a few percent of people have a browser that follows
> a particular spec and all the other browsers require something
> that violates the spec, I will violate the spec so the page
> works for almost everyone.
Can you show us a popular browser that "requires" non-compliant HTML in
order to display?
I suggest you reformulate your argument. As constructed, it's really not
credible.
--
Dan McGarry
Libraxus, Inc.
Gil Harvey
<bla...@moodindigo.com> wrote:
>> The browser companies do not have to follow the letter of
>> the spec. This means that we need to build towards making
>> a web page work on actual browsers, not pass some DTD that
>> no browser in existence adheres to.
>
>There is a great deal of HTML 4.0 that is supported across every major
>browser.
Was that supposed to be an answer to the paragraph above it?
Robert Gormley
Warren Steel <mu...@olemiss.edu> wrote in message
news:37DF09...@olemiss.edu...
> Robert Gormley wrote:
>
> > I think what would solve a lot of these debates (and not just on this
issue)
> > is the admission that yes, while both 'sides' agree, the question is
"Does
> > this situation require a real world, commonsense answer?"
>
>
> quotation marks reliably on the World Wide Web. It's simple
> ASCII vertical quote symbols. It's this obsession with
> superficial appearance and paper typography that leads
> authors to seek unsatisfactory adhockery. I haven't yet
> heard anyone claim that "these quotation marks" can't be
> read and understood by anyone using the Web.
Quotation marks are an issue of aesthetics, not *paper* typesetting.
Granted, I believe this issue a bit contrived, but I did say "not just this
issue" ;-)
Arjun Ray
| Boris Ammerlaan <bo...@stack.nl> wrote:
| > Pft. All right, I'll spell them all out:
| > s/ "/{double curly open}/g
| > s/" /{double curly close}/g
| > s/ '/{single curly open}/g
| > s/' /{single curly close}/g
It isn't that easy. There's enough variation in leading and trailing
contexts - whitespace isn't the only word boundary! - to pretty much
guarantee that regexps aren't good enough to get the job done - e.g.,
a "close" predicated on a prior "open", a leading context dependency.
I suspect at least a LR(2) grammar would be needed for consistently
styled text, but in some ways that could be overengineering the
problem. If the original were composed with automated translation in
mind, it would have been possible to choose the consistent style that
made the job easiest. (e.g. explicitly use " and/or ' for
the otherwise "hard" cases.)
| That fails to produce proper results in several cases:
|
| - Back in '65 we ate meat and potatoes.
This is the only "hard" one, because it's a false positive on leading
context (whitespace is a word boundary but a quote isn't intended.)
Note that your examples involved single quotes, which in general are
much more overloaded in common use. Whereas " would rate to work
for all non-quote usage (e.g. coordinate seconds), it might be better
to use ' strictly for quoting intent. (This is an observation;
I'm sure in practice relative frequency in the particular context
could argue the other way.)
:ar
Robert Gormley
Gil Harvey <g...@netquick.net> wrote in message
news:37df153c...@news.interpath.net...
It's almost like a self-contradictory statement (well, it is) - others
chastised for 'making do' in certain aspects of implementation, but then
saying we should 'make do' on other aspects :) *shrug*
Robert
Robert Gormley
...
> I'd say, "If you insist on invalid markup on this site, then you must pay
> the surcharge for a site without my company logo on the site, because I
> will not damage this firm's reputation by producing a site with invalid
> markup to satisfy the ill-considered whims of any client. If you aren't
> willing to pay the substantial surcharge, then you can find another firm
> to produce your site. From scratch."
>
> As soon as I start my own web design company that is. Right now I'm a
> computer programmer, not a web page designer. (But the program is related
> to HTML authoring.)
I do admire your steadfastness to your beliefs :) And to each their own...
But I think we've done this subject to death :)
Robert Gormley
opinion - but differences in approach" - not a case of Last Word Syndrome,
ie "We've done this to death - x is right z is wrong" :)
Arjun Ray
Uriel Wittenberg <u...@urielw.com> wrote:
| Arjun Ray wrote:
|
| > It conflates "work by design" and "work by accident".
| > These differ radically in their negations, and so are not the same.
| "differ radically in their negations"?
|
| Are they, however, relatively similar in their affirmativeness?
How else do you imagine a conflation could happen?
| By the way, what are the odds of all browsers displaying something
| identically *by accident*?
The same odds as that of the accident occuring. It has two parts: the
"code page" (or whatever else it might be called) in use on the local
machine, and the treatment of character references in the browser
(e.g. is a character reference blindly replaced by a "corresponding"
8-bit value?)
And, btw, not only was it never "all browsers", but also the number of
browsers is getting smaller over time.
| I wonder what the law of gravity would look like if Newton had
| regarded the common downward tendency of objects as a coincidence.
-sigh-
Another one so much in stupefied slackjawed awe of his wowser, that he
treats its behavior as oracular, and proceeds to theorize about it.
Into the killfile you go.
:ar
Paul Mitchum
> There is a real world, commonsense answer to producing
>quotation marks reliably on the World Wide Web. It's simple
>ASCII vertical quote symbols. It's this obsession with
>superficial appearance and paper typography that leads
>authors to seek unsatisfactory adhockery. I haven't yet
>heard anyone claim that "these quotation marks" can't be
>read and understood by anyone using the Web.
Wouldn't it be cool if it were implemented, though?
In fact, wouldn't it be cool if about ten kerjillion other things were
implemented, as well?
Steve Pugh
nlil...@NOSPAMinteractive-media.com (Nick Lilavois) wrote:
>Steve Pugh <st...@pugh.net> wrote:
>
>>Nick Lilavois wrote:
>>
>>>No, once again, if your client says "I want curly quotes
>>>Dammit! And I have seen plenty of pages that have them, so
>>>don't tell me we can't do it!"
>>
>>Funny, I've never had a client ask for that. Anytime clients send me
>>copy with curly quotes in it I change them to straight quotes and I've
>>had any client even notice it.
>
>I've even had a client that complained about kerning and
>anti-aliasing of text. (Hi Molly!)
On graphical headers, etc? Sure why not, they often take their cue
from the company logo and hence that can be important. On HTML text?
Good one to laugh about in the pub later.
>>Just a thought, why has no one suggested having the curly quotes
>>directly in the document, not encoded at all? With the proper charset
>>header of course. Considering that everything but the kitchen sink has
>>been mentioned in this thread...
>
>I don't think that would work- perhaps if they were
>foreign-language characters it would. What charset would
>they be in?
Well, Windows-1252 leaps instantly to mind. It's where you, perhaps
indirectly and unknowingly, got the 145-148 positions for the curly
quotes from anyway. Doing this would be perfectly valid, but would
probably be advised against as Windows character sets aren't as widely
supported as ISO-8859 character sets. See Jukka's article on this
subject at http://www.hut.fi/u/jkorpela/chars.html#win
How a browser running on a system that doesn't support the Windows
character set manages to convert your ‘ into a left single quote
is something I'm not sure about. So maybe all those browsers you
verifed that ‘ works on either (a) have access to said Windows
character set or (b) have hard coded error recovery for this common
error.
And then there's Unicode.
Steve
--
"My theories appal you, my heresies outrage you,
I never answer letters and you don't like my tie." - The Doctor
Steve Pugh <st...@pugh.net> <http://steve.pugh.net/>
Chris Gray
> Yes it does. It conflates "work by design" and "work by accident".
To say nothing of final and efficient causes ...
--
Eur. Ing. Chris Gray MBCS C. Eng. chris...@smartmove.be
Chris Gray
> I've even had a client that complained about kerning and
> anti-aliasing of text. (Hi Molly!)
Heaven help us if browser support for unicode gets so good that we
start using �xfb00; et al. "The f's are stuck together!" ;>
Boris Ammerlaan
>I liked your three line statement above. If a feature has
>been implemented a certain way in browsers used by 95+% of
>users, and the specifications say it should have been done
>another way, the specs need to be changed.
That is actually tho conservatist approach. I.e.: "It works
this way, why do we need a better way."
> If they aren't
>going to change the specification, a note needs to be
>attached that explains that the specification does not work
>in current browsers.
Maybe, although I think notes about crappy browser support for
character sets do not belong in an HTML definition document.
>These guys will never accept the fact that you need to do
>what is required to make the pages work as good as possible
>on actual browsers that people are using.
Thereby enforcing a monopoly?
>None of what you are doing looks like deception to me.
Then you have not done much work for clients. You need to
deceive your clients all the time; by not informing them
of every (im)possible consideration, you are suggesting
that fewer exist than you actually know exist.
>I make web sites that people want to actually use and I
>am practical enough to come down out of the clouds and
>build web pages with features that keep the users interested.
Then use something that falls back gracefully - using
some undefined character repertoire does not.
>You can only take theoretical practices so far before you
>have to do something that works for as many people as
>possible.
Straight quotes work for everybody; curly quotes do not.
>I have started building some pages that use fairly basic
>techniques with JavaScript browser detection used to
>redirect to different pages for MSIE4 and NS4+ and a third
>page for MSIE5. If JS is turned off, the user will stay on
>the basic page.
Isn't it better to use CGI for this purpose?
>> As the previous poster demonstrated, they DO work- even on
>> obscure browsers on rare platforms, while the "proper" way
>> of doing things won't even work in MSIE3.0 on 95, which
>> still has about 3%.
>The idea that people should use a technique that has been
>shown to _not_ work on most browsers is strange. This type
>of help is not needed.
But using curly quotes (145 etc.) does *not* work on most browsers.
At best, it works on a very limited set of browsers a lot of people
happen to be using at the moment.
--
Boris Ammerlaan <bo...@stack.nl>, http://www.stack.nl/%7Eboris/
* HTML FAQ: posted bi-weekly & <URL:http://www.htmlhelp.com/faq/html/>
* c.i.w.a.h. FAQ List Pointer: posted twice a week &
<URL:http://www.stack.nl/%7Eboris/HTML/ciwahfaq.html>
Gil Harvey
>I'd say, "If you insist on invalid markup on this site, then you must pay
>the surcharge for a site without my company logo on the site, because I
>will not damage this firm's reputation by producing a site with invalid
>markup to satisfy the ill-considered whims of any client. If you aren't
>willing to pay the substantial surcharge, then you can find another firm
>to produce your site. From scratch."
>
>As soon as I start my own web design company that is. Right now I'm a
>computer programmer, not a web page designer. (But the program is related
>to HTML authoring.)
Ah, ever notice how easy it is to say "I'd do..." when some
one is depending on some one else to pay the rent, electric, payroll,
etc. ??
Panos Stokas
> >what is required to make the pages work as good as possible
> >on actual browsers that people are using.
>
> Thereby enforcing a monopoly?
This is the only part to which I disagree. Monopoly on browsers (that is, IE
dominance) has been enforced *by* the Internet specifications. An adequate
HTML4.0 + CSS implementation is too heavy for a small group of
programmers...
--quoting this line killfiles your message