Never Mind

[Joy Beeson]

I validated my diary, just for practice. I habitually use an
almost-blank line

</p><p><center> <!-- *A* -->
&#8258;</center>
</p><p> Just in case we like the devilled eggs:

to mark a change of subject or a lapse in time. Each time these lines
occur, validator.w3.org highlights the next-to-the last ">" (That is,
the second occurrence of </p>) and says

> No p element in scope, but a p end tag seen

But the <p> is right there plainer than the nose on my face.

Could this be a side effect of putting <center></center> inside a
paragraph?

I changed a couple, the validator no longer mentioned them, I changed
all to

</p><center><p> <!-- *A* -->
&#8258;</p>
</center><p>

And now the validator contents itself with complaining that <center>
is obsolete, which bothers me not one whit.

On the other hand, the validator is happy with my changes of topic,
but I can no longer simply paste two lines above a standard paragraph
beginning when I want to put in a marked blank line. But it took me a
long time to figure out the obvious way to insert the old code; there
is no doubt an easy way that at the moment escapes me; perhaps I could
put it on a defined key such as the one I used for signatures on
letters back when I wrote enough letters that I could remember where
the signature key was. (Now I copy an old letter and delete the
body.)

And I'm going to mail this post anyway, to let you guys know that
sometimes you explain things just by being there.


--
Joy Beeson
joy beeson at comcast dot net
http://www.debeeson.net/joy/
The above message is a Usenet post.
I don't recall having given anyone permission to use it on a Web site.

[richard]

it should be <p><center>test</center></p>
or <center><p>text</p></center>
If the latter is what you are attempting, then try this
<div class="paragraph"><p>text</p></div>
and style the division as below.

better ► <p class=center"></p>
CSS: .center { text-align:center;}

The text you are attempting to center is a "comment".
That will never be shown on the page regardless of where it is.

Another method for centering an object is
margin-left:auto; margin-right:auto;

[Chris F.A. Johnson]

> better ??? <p class=center"></p>

> CSS: .center { text-align:center;}
>
> The text you are attempting to center is a "comment".

No, she's trying to center &#8258;

Though I don't understand why she doesn't use CSS.

--
Chris F.A. Johnson
<http://torontowebdesign.cfaj.ca/>

[Jukka K. Korpela]

2013-06-01 5:12, Joy Beeson wrote:

> I validated my diary, just for practice. I habitually use an
> almost-blank line
>
> </p><p><center> <!-- *A* -->
> &#8258;</center>
> </p><p>

Like most comments, the comment “*A*” is misleading, since the character
used in the content is ASTERISM, “⁂”. This character is really the
problem you should consider: have you specified for it a reasonable list
of font families, containing fonts that have a glyph for it? Cf. to the
list at
http://www.fileformat.info/info/unicode/char/2042/fontsupport.htm
Otherwise, old browsers might even be unable to display it at all.

> Each time these lines
> occur, validator.w3.org highlights the next-to-the last ">" (That is,
> the second occurrence of </p>) and says
>
>> No p element in scope, but a p end tag seen

That’s because the <center> tag, starting a block-level element,
implicitly closes the open p tag, as if the markup started with

</p><p></p><center>

And this makes the next </p> homeless. The minimal fix is to remove that
</p> tag. A better fix is to use

<p align=center>&#8258;</p>

though some people say you should used

<p class=divider>&#8258;</p>

together with a CSS rule like

.divider { text-align: center }

And some Real Purists (like a past version of me) might tell you to use
<hr> and just style it so that it appears as a centered asterism. (It’s
possible by the spec, but perhaps not well enough supported yet by all
those old browsers around.) The reason is that <hr> has always been
defined to denote “change of topic”, even though it is named so that
uneducated people would guess that it means “horizontal rule”. ☺

> Could this be a side effect of putting <center></center> inside a
> paragraph?

In a sense. It’s the effect of trying to put a center element inside a p
element, which is forbidden (by HTML specs) and impossible (browsers
behave like I described, implicitly closing the p element).

> And now the validator contents itself with complaining that <center>
> is obsolete, which bothers me not one whit.

Then why are you using <!doctype html>, which makes the validator apply
HTML5 rules, which express some people’s opinion that presentational
markup like <center> is Evil, Obsolete, and Anathema? If you wish to
keep using such markup, the HTML 4.01 Transitional doctype is a better
choice, probably.

--
Yucca, http://www.cs.tut.fi/~jkorpela/

[dorayme]

In article <jpliq8d3fi1uu4na9...@4ax.com>,

Joy Beeson <jbe...@invalid.net.invalid> wrote:

> I validated my diary, just for practice. I habitually use an
> almost-blank line
>
> </p><p><center> <!-- *A* -->
> &#8258;</center>
> </p><p>

If you do not want to make the safeguards JK recommended about fonts
and you are not worried otherwise, use an image of the asterisks. I
give you a choice for the alt text which can come in play if images
are off, you might choose yet another alt replacement if you wish.

<http://dorayme.netweaver.com.au/beetson.html>

--
dorayme

[Luuk]

This page does not show correctly in IE10, with 'Show pictures' off

It also show real big '*' when printing the page (=looking a preview,
because i did not waste a 'forrest' for actually printing...)

[dorayme]

In article <h5vo7a-...@luuk.invalid.lan>, Luuk <lu...@invalid.lan>
wrote:

Don't know much about IE10, thanks for the reminder.

If really using, one should provide for printing, my above was just
something quick for screen media. I did use a rather large asterisk
set to keep sharpness for em enlargement.

It *is* a worry using images when images are not available! Some
browsers do behave rather differently in respect to alt text...

Perhaps safest of all is

<hr style="width: 80%">

--
dorayme

[Jonathan N. Little]

Jukka K. Korpela wrote:
> And some Real Purists (like a past version of me) might tell you to use
> <hr> and just style it so that it appears as a centered asterism. (It’s
> possible by the spec, but perhaps not well enough supported yet by all
> those old browsers around.) The reason is that <hr> has always been
> defined to denote “change of topic”, even though it is named so that
> uneducated people would guess that it means “horizontal rule”. ☺

Since it is really just a decoration and not content I think that it
falls under the purview of CSS. I would use a class on the paragraph or
div it the entries encompass more than one paragraph. It won't work for
<=IE7 but since it just decoration would not impact the content IMO. Not
sure if OP had the asterism preceeding or following entries but that is
ease to adjust, and html entities will not work with generated content
so either use unicode, alternative character, or use an image:


<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01//EN"
"http://www.w3.org/TR/html4/strict.dtd">
<html><head>
<meta http-equiv="content-type" content="text/html; charset=UTF-8">
<title>Via CSS</title>

<style type="text/css">
.entry:after { content: '⁂'; display: block; text-align: center; }
</style>
</head>

<body>
<p class="entry">
This is one paragraph entry...
</p>

<div class="entry">
<p>
An entry with more than one paragraphs...
</p>
<p>
Second associated paragraph...
</p>
</div>

<p class="entry">
This is one paragraph entry...
</p>

</body></html>

--
Take care,

Jonathan
-------------------
LITTLE WORKS STUDIO
http://www.LittleWorksStudio.com

[Jukka K. Korpela]

2013-06-01 18:09, Jonathan N. Little wrote:

> Jukka K. Korpela wrote:

>> And some Real Purists (like a past version of me) might tell you to use
>> <hr> and just style it so that it appears as a centered asterism. (It’s
>> possible by the spec, but perhaps not well enough supported yet by all
>> those old browsers around.) The reason is that <hr> has always been
>> defined to denote “change of topic”, even though it is named so that
>> uneducated people would guess that it means “horizontal rule”. ☺
>
> Since it is really just a decoration and not content I think that it
> falls under the purview of CSS.

The description in the original message was “to mark a change of subject
or a lapse in time”. This sounds like a change of topic to me. (“A lapse
of time” seems to relate to a story and imply some change of topic.) In
speech rendering, I would expect this to be rendered as a considerable
pause.

That’s way <hr> would appear to be appropriate, because it has a default
rendering that is at least tolerable for this purpose. Styling <hr> is a
bit tricky, due to inconsistencies between browsers in its default
rendering. But the following would work reasonably:

hr {
border: none;
height: 1em;
text-align: center;
}
hr:before {
content: "***";
}

If you really want an asterism instead of three asterisks, contrary to
literary tradition, you can replace "***" by "\2042".

--
Yucca, http://www.cs.tut.fi/~jkorpela/

[Joy Beeson]

On Sat, 01 Jun 2013 09:38:24 +0300, "Jukka K. Korpela"
<jkor...@cs.tut.fi> wrote:

Quite a lot of useful information; thank you.

> Then why are you using <!doctype html>, which makes the validator apply
> HTML5 rules, which express some people’s opinion that presentational
> markup like <center> is Evil, Obsolete, and Anathema? If you wish to
> keep using such markup, the HTML 4.01 Transitional doctype is a better
> choice, probably.

Every time I ask which doctype to use, I get a different answer, and
<!doctype html> was the latest. Seemed logical, since it only says
"html" and doesn't say which flavor. Terms of art will mess you up
every time. Especially when two different arts use the same term, as
when doctors and knitters say "stockinet". (Doctors mean "one-on-one
ribbing")

If validators weren't as useful as spelling checkers, I'd delete
doctypes entirely.

--------------------------------

After following links from the validator site, the new latest is
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"
"http://www.w3.org/TR/html4/loose.dtd">

Seems to me that there are a great many more characters in that than
are needed.

[Joy Beeson]

On Sat, 1 Jun 2013 01:45:18 -0400, "Chris F.A. Johnson"
<cfajo...@gmail.com> wrote:

> Though I don't understand why she doesn't use CSS.

Unless there are wing-dings I haven't yet found and removed lurking in
ancient files, <center> is the *only* presentational element I ever
use. Learning CSS just for that would be like shooting a flea with an
elephant gun.

Particularly since the file I was practicing on is the draft copy for
a plain-text e-mail and all the <center>s drop out anyway -- I stick
them in because it looks nice while I'm proofreading.

[Joy Beeson]

On Fri, 31 May 2013 23:12:30 -0300, Joy Beeson

<jbe...@invalid.net.invalid> wrote:

> there
> is no doubt an easy way that at the moment escapes me;

So when I closed out May and opened June, it was easier to type

</p><center>&#8258;
</center><p> <!-- *A* -->

</p><p> Took a walk down Boys City after my nap; the


Than to copy from the old document. This makes the validator happy,
but it doesn't make me happy. The asterism is a separate paragraph
and it ought to be *marked* as a separate paragraph; I don't like
relying on deprecated tags to keep it from being appliqu&eacute;d to
the previous paragraph.

Not to mention that there are no letters between <center> and
</center>, which causes my spelling checker to report it as a repeated
word. Not good in a draft copy that gets spelling-checked after each
modification.

A little shuffling around produced:

</p><center><p> <!-- *A* -->
&#8258;</p>

</center><p> I did measure from the selvages for the two

Which keeps both me and the validator happy, and encroaches less on
the space reserved for content. (I place a VERY high value on being
able to read source code as if it were plain text.)

It also puts letters between the centers, so I could go back to using
*** to represent the asterism, but I've become accustomed to *A*.

(I just now noticed that "A" stands for "Asterism". I selected A
because it was the symmetrical letter that looked best between two
asterisks in eighty-column font.)

This version removes the ability to paste the asterism above a new
paragraph, but I think I can get used to pasting it as a new
paragraph.

And maybe there's a variation on "enter" that isn't already in use,
that I could define as "new topic".

Perhaps I should have separate signs for "sharp turn" and "I've been
away from the typewriter for a while". The document contains no
section breaks, so <hr> would do for "sharp turn". Or I could just
say "-----------------"

Now it's way past time to stop talking about talking about it and get
back to doing it.

[Doug Miller]

Joy Beeson <jbe...@invalid.net.invalid> wrote in
news:kfq3r89i33hvcr49e...@4ax.com:

> After following links from the validator site, the new latest is
> <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"
> "http://www.w3.org/TR/html4/loose.dtd">
>
> Seems to me that there are a great many more characters in that than
> are needed.

Only because you fail to understand their purpose.

http://www.w3.org/QA/2002/04/valid-dtd-list.html
http://www.w3.org/QA/Tips/Doctype
http://alistapart.com/article/doctype/

[dorayme]

In article <timstreater-E323...@news.individual.net>,
Tim Streater <timst...@greenbee.net> wrote:

> I wouldn't
> bother with standalone validators, use browsers to validate your html.

This is not good advice for a number of reasons which have been
explained a few times.

--
dorayme

[Joy Beeson]

On Sat, 08 Jun 2013 07:43:52 +1000, dorayme <dor...@optusnet.com.au>
wrote:

> This is not good advice for a number of reasons which have been
> explained a few times.

Even I could see that he was off the wall.

[se]

"Joy Beeson" <jbe...@invalid.net.invalid> skrev i meddelelsen
news:ti75r8dgje5g9rllt...@4ax.com...

> On Sat, 08 Jun 2013 07:43:52 +1000, dorayme <dor...@optusnet.com.au>
> wrote:
>
>> This is not good advice for a number of reasons which have been
>> explained a few times.
>

> Even I could see that he was off the wall.

It isn't a good idea wiping the floor with somebody here that is
trying to extend your knowledge. There's no sign from you of knowing
anything of website validating.

/se

[dorayme]

In article <ti75r8dgje5g9rllt...@4ax.com>,

Joy Beeson <jbe...@invalid.net.invalid> wrote:

> On Sat, 08 Jun 2013 07:43:52 +1000, dorayme <dor...@optusnet.com.au>
> wrote:
>
> > This is not good advice for a number of reasons which have been
> > explained a few times.
>
> Even I could see that he was off the wall.

He has one good practical point and that is that you should test your
website in different browsers to see what others will see, you should
also change the text size as you do this and you should change the
browser size - in other words take it for a test run.

But you cannot test your website in browsers that have not been
created. It is almost logically impossible. Nor can you predict
changes to browsers. Guess where the browser makers get quite a bit,
not all, of their information from? They get it from the various
documentations that go to informing validators. It's a big complicated
conversation. You can help yourself by at the very least doing the
validations that provide a sort of prima facie ok. Everything is
provisional, but it is a handle that would be foolish to ignore.

Put it another way, just as with many a construction of any object for
public use, in developing you will find it very useful indeed in the
long run to run it by the validators to pick up alleged mistakes for
your consideration. Your remark about spell-checkers was spot on!

If you understand a bit about these things, you will get to the stage
of not being too worried by some flagged errors or warnings. But if
you do not understand such things, perhaps your best bet is to make
sure it validates and that it works in all your tests too.

--
dorayme

[Jukka K. Korpela]

2013-06-08 12:24, Tim Streater wrote:

>> <!doctype html> "says" HTML5.
>
> Only to a validator. Browers only care whether you have a doctype
> or not, and will correspondingly operate either in standards or quirks
> mode.

Browsers care about the exact doctype string, which is like a magic
incantation: a single typo there probably makes it lose its power.

In the set of doctype strings that trigger "standards" mode, <!doctype
html> is by far the simplest.

Regarding validators, they let you use "doctype override", i.e. specify
the version of HTML to be checked against via menus in the validator's
user interface.

--
Yucca, http://www.cs.tut.fi/~jkorpela/

[dorayme]

In article <timstreater-0842...@news.individual.net>,
Tim Streater <timst...@greenbee.net> wrote:

> In article <dorayme-905C0E...@news.albasani.net>,
> dorayme <dor...@optusnet.com.au> wrote:
...
> > ...you cannot test your website in browsers that have not been
> > created. ... Nor can you predict changes to browsers.
>
> ...the substance of my point is that despite various
> standards bodies insisting that this or that feature of html is
> "deprecated", I'm not aware of any feature that has actually been
> removed (except perhaps things that were put in ahead of standards and
> later removed).

There is more to validation than deprecation.

--
dorayme

[Jukka K. Korpela]

2013-06-08 12:27, Tim Streater wrote:

> despite various standards bodies insisting that this or that feature of
> html is "deprecated", I'm not aware of any feature that has actually
> been removed (except perhaps things that were put in ahead of standards
> and later removed).

As regards to removing features from specs, e.g. XHTML 1.1 dropped a
large number of features (almost all presentational features), as
compared with XHTML 1.0 and HTML 4.01. It did not become a success, and
this may be related to that move, at least in part.

HTML 4 dropped <xml>, which was flagged as deprecated in HTML 3.2 (and
in HTML 2.0). But browsers have kept supporting it.

So is there features that have been in HTML specifications and have been
implemented but later removed from *browsers*? Not many. The only thing
that comes into my mind is <listing>, but even it is still supported,
but with interpretation essentially different from the original, so one
might say that the *original* <listing> has been dropped. But it has
been used very, very little.

--
Yucca, http://www.cs.tut.fi/~jkorpela/

[Christoph Becker]

Tim Streater wrote:
> In article <kovfv6$vka$1...@dont-email.me>,

> "Jukka K. Korpela" <jkor...@cs.tut.fi> wrote:
>

> And so I rest my case. Remember: there are lots more guys involved in
> writing browsers than there are writing validators.

And there are even more guys involved in writing HTML. If these would
not use deprecated features, new browser versions would not have to
support these features anymore (at least sometimes in the future), what
might simplify the codebase of the browsers, so the developers could
focus on more important aspects.

--
Christoph M. Becker

[Jukka K. Korpela]

2013-06-08 18:35, Tim Streater wrote:

> In article <kovfv6$vka$1...@dont-email.me>,

> "Jukka K. Korpela" <jkor...@cs.tut.fi> wrote:

[...]

>> So is there features that have been in HTML specifications and have
>> been implemented but later removed from *browsers*? Not many. The only
>> thing that comes into my mind is <listing>, but even it is still
>> supported, but with interpretation essentially different from the
>> original, so one might say that the *original* <listing> has been
>> dropped. But it has been used very, very little.
>

> And so I rest my case. Remember: there are lots more guys involved in
> writing browsers than there are writing validators.

Hold on... I just noticed that <basefont> has been dropped. Modern
browsers simply ignore it; they generate a DOM node, but that's all. IE
was the last to have support, but IE 10 dropped it.

<basefont> was introduced in HTML 3.2, with size=... as the only
attribute. HTML 4 added the face=... and color=... attributes but
declared the element deprecated. Support was once widespread but never
consistent, partly because the definitions were vague. HTML5 declares it
as obsolete and forbidden and describes it only syntactically and in DOM
perspective, without saying anything about its effect.

And <basefont> has been used, and is still used on legacy pages. It has
just lost effect.

--
Yucca, http://www.cs.tut.fi/~jkorpela/

[dorayme]

In article <timstreater-54E9...@news.individual.net>,

Tim Streater <timst...@greenbee.net> wrote:

> And so I rest my case. Remember: there are lots more guys involved in
> writing browsers than there are writing validators.

Your case is not as substantial as it might seem that it it is rested
so easily just because browsers have been supporting deprecated
things. It is a good idea, not a bad idea to be thrown out, for
website developers to check their work regularly with validators.

--
dorayme

[dorayme]

In article <timstreater-31A9...@news.individual.net>,
Tim Streater <timst...@greenbee.net> wrote:

> In article <kovqvd$rj8$1...@speranza.aioe.org>,
> Christoph Becker <cmbec...@gmx.de> wrote:

>
> > Tim Streater wrote:
>
> > > And so I rest my case. Remember: there are lots more guys involved in
> > > writing browsers than there are writing validators.
> >
> > And there are even more guys involved in writing HTML. If these would
> > not use deprecated features, new browser versions would not have to
> > support these features anymore (at least sometimes in the future), what
> > might simplify the codebase of the browsers, so the developers could
> > focus on more important aspects.
>

> The chance of people stopping using deprecated features, and just as
> importantly, going back and changing existing sites, is vanishingly
> small.

I see you are fixated on this deprecation business! I have been
trying, without success it seems, to communicate *past* this.
...

--
dorayme

[Jukka K. Korpela]

2013-06-09 0:48, Tim Streater wrote:

> The chance of people stopping using deprecated features, and just as
> importantly, going back and changing existing sites, is vanishingly
> small.

I'm afraid far too many people are working on existing sites to clean
them up, changing them to comply with "latest HTML standard", whatever
this might mean to someone this week. But surely most sites still don't
comply with any "HTML standard".

> There was already an attempt to enforce a cleanup: XHTML, which tried to
> insist that browsers should not do error correction in the case of badly
> formed source, but that instead the browser should just give an error
> message saying "bad page" or something. The browser writers ignored
> this,

Not quite. They just apply such Draconian error handling at most to
pages served as real XHTML, i.e. with an XML media type, like the demo page
http://www.cs.tut.fi/~jkorpela/test/error.xhtml
Firefox processes it by XHTML rules. Chrome displays the page (it should
not) but preceded by a prominent error message. Modern versions of IE
can deal with real XHTML but they fail to do it properly: they show the
page, with error recovery, and do not even display an error message,
except in the console log, which is normally invisible to users.

> for the same reason that anyone else would: there are billions of
> pages already in existence which are full of errors.

That's not really the reason. Those billions of pages are not delivered
as real XHTML, so they would not be affected.

> Instead, the browser writers figured out how to recover from HTML errors
> in a consistent way. AIUI, this is even documented now.

Well, let's say that error handling is being documented and is being
made more consistent. I guess this is a good thing in general, even
though it effectively extends as modifies HTML as a real-life language:
when an error or an oddity gets documented well and gets handled
consistently in browsers, it becomes part of the HTML language that
authors can use - no matter how much the specs declare such markup as
deprecated, obsolete, cursed, sinful, or whatever.

--
Yucca, http://www.cs.tut.fi/~jkorpela/