cellpadding and cellspacing problem

[K.]

Hello!

I created yet unfinished page www.slub.zmyslowski.pl about my wedding.
I would like to fit it to every browser, mostly for Firefox and Internet
Explorer.
On internet explorer everything is fine, but
when I open this page on Firefox I have small white lines between two images
and I don`t know where error is. Could you help me?

I used validator before-> http://www.htmlvalidator.com/. -> my page is
valid.

Kindest regards
Marcin

 

[Travis Newbury]

I created yet unfinished pagewww.slub.zmyslowski.pl about my wedding....
I used validator before->http://www.htmlvalidator.com/. -> my page is

Man I don't have an answer for you, but I have to give you kudos for
doing all the prerequisite work and giving us a URL and validating
your page first. This is a rare thing in alt.html

I think we should all give the OP the "doing the right thing" award,
and all the honors that come with it.

 

[Els]

On internet explorer everything is fine, but
when I open this page on Firefox I have small white lines between two images
and I don`t know where error is. Could you help me?

If you give the images in the first and second row of the table an
extra attribute: style="display:block;", then the lines will
disappear. They are caused by the space that is below text and
inline-images, to make room for descenders of letters like g,j,p,q,y.

 

[Jukka K. Korpela]

Scripsit Travis Newbury:

Man I don't have an answer for you, but I have to give you kudos for
doing all the prerequisite work and giving us a URL and validating
your page first.

Well, almost giving us a URL (the http:// part is missing) and trying to
validate. Unfortunately, he used a product that is dishonestly sold as a
validator but isn't, the infamous "CSE HTML Validator". His confusion is
understandable; the phoney validator's business isn't.

In reality, the page contains dozens of syntax errors, as a real validator
like http://validator.w3.org would tell (for free). The page has a mixture
of XHTML 1.0 and HTML 4.01 syntax, and contrary to the doctype declaration,
it uses Transitional features. It should be fixed to comply with HTML 4.01
Transitional, then perhaps improved (by replacing presentational markup by
CSS) to HTML 4.01 Strict.

The problem asked is independent of these problems. As Els writes in his
reply, the gaps are caused "by the space that is below text and
inline-images, to make room for descenders of letters like g,j,p,q,y".
Setting display: block for the images is one way of fixing. Another option,
perhaps more natural, is
img { vertical-align: bottom; }

Technically, the problem is that an image is treated as a special kind of
text element, like a letter, which is by definition vertically aligned to
the _baseline_ of text. You can see this if you put an image between
letters, e.g. <p>yep<img ...>yep</p>.

This is what CSS specifications say, and Firefox obeys this, whereas IE
doesn't. But this can fixed either by making the images blocks (as far as
rendering is considered), which aren't treated as text elements, or by
explicitly setting their vertical alignment to bottom (i.e., the bottom of
the image is aligned to the bottom of the enclosing element's box).

 

[Bergamot]

Technically, the problem is that an image is treated as a special kind of
text element, like a letter, which is by definition vertically aligned to
the _baseline_ of text.

This is what CSS specifications say, and Firefox obeys this, whereas IE
doesn't.

FYI, IE has other bugs with replaced elements, too, such as failure to
honor line-height.

 

[dorayme]

Hello!

I created yet unfinished page www.slub.zmyslowski.pl about my wedding.
I would like to fit it to every browser, mostly for Firefox and Internet
Explorer.
On internet explorer everything is fine, but
when I open this page on Firefox I have small white lines between two images
and I don`t know where error is. Could you help me?

The white line prob has been addressed and you have two excellent
answers. If you go with Els' you can simply put a line in your
style sheet for all images (rather than pick out particular
images), img {display:block;}

Not sure why you have image map for the top banner? Do you need
to have link for it?

In fact, considering the errors in the page, why not simplify it
greatly. Either as a table or as not. The table idea first:

First row would be just a cell with your Baner.gif, second row
would be a row with 5 cells, in each a simple text link, you can
recommend a font and specify a starting size. This would have the
advantage of scaling with the text size of the user, not using
more bandwidth to transmit pictures etc. If you must have images
for the links, no need for all the image map code. Imagemaps come
into their own when you have a big picture and want to link to
bits of it. There is not much point in having an image map if you
are already splitting up the image anyways into separate images.
Third row is paragraph text with a centered HeyKasiaNosowska.jpg.
Enclose text in <p>s, style them if necessary.

Use

<!DOCTYPE html PUBLIC "-//W3C//DTD html 4.01//EN"
"http://www.w3.org/TR/html4/strict.dtd">

to head your html and use a proper validator (see JK's
suggestion) to check up on your code.

If you have time, forget about the table structure for so simple
a site, consider using (as a simple suggestion), a div for the
head, an inline list for the navigation, and another div to hold
the bottom with paragraphs and images. Look up inline lists on
Google.

Just some suggestions!

 

[K.]

Scripsit Travis Newbury:


Well, almost giving us a URL (the http:// part is missing) and trying to
validate. Unfortunately, he used a product that is dishonestly sold as a
validator but isn't, the infamous "CSE HTML Validator". His confusion is
understandable; the phoney validator's business isn't.

In reality, the page contains dozens of syntax errors, as a real validator
like http://validator.w3.org would tell (for free). The page has a mixture
of XHTML 1.0 and HTML 4.01 syntax, and contrary to the doctype
declaration, it uses Transitional features. It should be fixed to comply
with HTML 4.01 Transitional, then perhaps improved (by replacing
presentational markup by CSS) to HTML 4.01 Strict.

The problem asked is independent of these problems. As Els writes in his
reply, the gaps are caused "by the space that is below text and
inline-images, to make room for descenders of letters like g,j,p,q,y".
Setting display: block for the images is one way of fixing. Another
option, perhaps more natural, is
img { vertical-align: bottom; }

Technically, the problem is that an image is treated as a special kind of
text element, like a letter, which is by definition vertically aligned to
the _baseline_ of text. You can see this if you put an image between
letters, e.g. <p>yep<img ...>yep</p>.

This is what CSS specifications say, and Firefox obeys this, whereas IE
doesn't. But this can fixed either by making the images blocks (as far as
rendering is considered), which aren't treated as text elements, or by
explicitly setting their vertical alignment to bottom (i.e., the bottom of
the image is aligned to the bottom of the enclosing element's box).

THANK YOU VERY MUCH FOR HELP

It fully helped me

Thank you one more time
Marcin

 

[Albert Wiersch]

Unfortunately, he used a product that is dishonestly sold as a validator
but isn't, the infamous "CSE HTML Validator". His confusion is
understandable; the phoney validator's business isn't.

It's only "phony" if you strictly limit your definition... the vast majority
of people would agree that CSE HTML Validator is a validator since they
understand the general meaning and function of an "HTML validator". Plus,
CSE HTML Validator can find many more problems than "real" validators. Have
a look at
http://www.htmlvalidator.com/htmlval/whycseisbetter.html

The online validator at http://www.htmlvalidator.com/ is limited since it is
based on the free lite edition, so it doesn't find as many issues as the
paid versions, but still is quite useful and can still find issues that
other validators and checkers miss.

But passing validation by any validator does not mean your page will display
"right". Different browsers can display the same HTML differently, and this
greatly complicates trying to render certain things a specific way in
multiple browsers.

Albert

 

[Jukka K. Korpela]

Scripsit Albert Wiersch:

It's only "phony" if you strictly limit your definition...

Not unexpectedly, you express your continued determinedness to sell your
product under a misleading name, i.e. to cheat customers.

Not unexpectedly, you had nothing to say about the user's confusion that
your program caused by giving false information. You don't really
participate Usenet discussions; you just advertize and present excuses for
misleading people.

 

[Albert Wiersch]

Not unexpectedly, you express your continued determinedness to sell your
product under a misleading name, i.e. to cheat customers.

Actually, our customers would feel cheated if they only got a "real"
validator as you define it. They don't want to be limited to that, but we do
include one in the current std/pro editions, though hardly anyone uses it
based on the questions and feedback (lack of that is) I get about it.

Not unexpectedly, you had nothing to say about the user's confusion that
your program caused by giving false information. You don't really
participate Usenet discussions; you just advertize and present excuses for
misleading people.

Saying (or implying) that a "real" DTD based validator is the only type of
checking worth doing is what is truly misleading. If there is any "false"
information I or the product gives out, then please feel free to bring it to
my attention, but please don't rehash the name argument or anything based on
it as there's no need to keep arguing about that. You can disgree with it,
but most people (I'd say more than 99%) have no problem with it. And like I
said earlier, the current std/pro editions include a DTD based validator for
the few who want to use one.

Albert

 

[Jukka K. Korpela]

Scripsit Albert Wiersch:

Saying (or implying) that a "real" DTD based validator is the only
type of checking worth doing is what is truly misleading.

It would. And it is disgustingly dishonest that you present, for your
commercial purposes, my view as if I had written something like that. Surely
you know better, if you have read what I have written about validation, and
you have hardly managed to avoid that because you keep commenting on it.
Thus, you make deliberate lies.

I have repeatedly said that a good checking tool that checks _many_ aspects
of web page quality would be very useful. The product you are selling is
nothing of the kind, though. If you cannot perform a simple and fairly
trivial part of the checking correctly, how could you do something
essentially more complicated? Even if some of the checks it performs are in
the right direction, users will have no way of distinguishing them from
bogus. What's the point of getting a hundred "error messages" when half of
them are just nonsense, even if some of the rest relate to some real error?

If there is
any "false" information I or the product gives out, then please feel
free to bring it to my attention,

You have repeatedly expressed your willingness to ignore such notes.

You keep selling your phony "validator", and you keep confusing people about
validation (they're surely confused enough without your efforts). The name
seems to be important to you, despite its incorrectness. Ergo, it is
important to you to mislead potential customers.

And you have not made _any_ statement about the fact from which the
discussion started: that your product once again seriously misled someone.

 

[Albert Wiersch]

It would. And it is disgustingly dishonest that you present, for your
commercial purposes, my view as if I had written something like that.
Surely you know better, if you have read what I have written about
validation, and you have hardly managed to avoid that because you keep
commenting on it. Thus, you make deliberate lies.

I've read your posts and to me at least, they imply that people should use a
"real" validator instead of CSE HTML Validator. They imply that there's no
use for a checker like CSE HTML Validator.

I have repeatedly said that a good checking tool that checks _many_
aspects of web page quality would be very useful. The product you are
selling is nothing of the kind, though. If you cannot perform a simple and
fairly trivial part of the checking correctly, how could you do something
essentially more complicated? Even if some of the checks it performs are
in the right direction, users will have no way of distinguishing them from
bogus. What's the point of getting a hundred "error messages" when half of
them are just nonsense, even if some of the rest relate to some real
error?

I suppose you don't know CSE HTML Validator if you think it doesn't perform
checks on _many_aspects of a web page. It certainly does. I don't know what
you are referring to as "bogus" except that you think the name is "bogus".
Otherwise please let me know of something specific it says that is truly
bogus.

Since you repeatedly say that a good checking tool checks _many_ aspects of
a web page, why do you continue to bash CSE HTML Validator? Now that it
includes a "real" validator (std/pro) on top of all the other checks it
does, I would think that you would be recommending it left and right!

As for having a hundred "error messages" when have are nonsense, please tell
me what messages are nonsense? Yes there may be some "cascading" errors that
may cause confusion... no checker is perfect and some people will always get
confused, especially people who are new to HTML and just learning it.

You have repeatedly expressed your willingness to ignore such notes.

I have only ignored the ones that I deemed weren't useful. Obviously it
wouldn't make sense to act on every note since not every note is useful. By
useful I mean that it will result in a positive improvement in the product
and in what people obtain from it.

You keep selling your phony "validator", and you keep confusing people
about validation (they're surely confused enough without your efforts).
The name seems to be important to you, despite its incorrectness. Ergo, it
is important to you to mislead potential customers.

I wouldn't blame me for the confusion... DTD based validators are much more
confusing that CSE HTML Validator. It seems misleading to me when the W3C
calls documents that have a lot of problems "valid". I have an example on my
site of a page with many problems, but "real" validators don't find any of
the issues. That sure sounds like it could misleading!

And you have not made _any_ statement about the fact from which the
discussion started: that your product once again seriously misled someone.

Sorry, I don't see how it misled the original poster. The only confusion I
potentially see is that the OP may not have known that just because a page
doesn't have problems (as detected by a checker) that it doesn't mean it
will render as intended. Also, like I said before, the W3C misleads the same
way, by saying a page is valid when it doesn't render as the user intended
or still contains problem. With that logic, then all checkers and validators
would be misleading.

Albert

 

[Jukka K. Korpela]

Scripsit Albert Wiersch:

Sorry, I don't see how it misled the original poster. The only
confusion I potentially see is that the OP may not have known that
just because a page doesn't have problems (as detected by a checker)
that it doesn't mean it will render as intended. Also, like I said
before, the W3C misleads the same way, by saying a page is valid when
it doesn't render as the user intended or still contains problem.
With that logic, then all checkers and validators would be misleading.

Babble, babble.

Are you pretending to be so stupid that you did not understand the
statements that described that the "CSE HTML Validator" claimed a page to be
valid when it in fact had dozens of reportable markup errors, i.e. was
invalid in the sense that is relevant in HTML context?

You have repeatedly claimed that your commercial product, "CSE HTML
Validator", is better than the free validators around. Once again, it was
pointed out that it is much _less_ and even claims that a page is valid when
it is not.

 

[Albert Wiersch]

Are you pretending to be so stupid that you did not understand the
statements that described that the "CSE HTML Validator" claimed a page to
be valid when it in fact had dozens of reportable markup errors, i.e. was
invalid in the sense that is relevant in HTML context?

I don't think I'm the stupid one here. CSE HTML Validator doesn't claim
pages to be "valid". It simply finds potential issues based on real-world
browsers. It doesn't generate as many worthless (useless in real-life)
"errors" as real validators.

You have repeatedly claimed that your commercial product, "CSE HTML
Validator", is better than the free validators around.

Because it is in a many number of ways.

Once again, it was pointed out that it is much _less_ and even claims that
a page is valid when it is not.

Once again you do not know what you're talking about because CSE HTML
Validator doesn't claim pages are "valid".

Albert

 

[Beauregard T. Shagnasty]

CSE HTML Validator doesn't claim pages to be "valid". ...

Well! Then why do you name it CSE HTML *Validator* ?

Suggest you change its name to:

CSE HTML Simply-Finds-Potential-Issues-Based-On-Real-World-Browsers.

 

[Albert Wiersch]

Well! Then why do you name it CSE HTML *Validator* ?

Because it finds problems in web pages, which is what people want when they
look for a validator.

Suggest you change its name to:

CSE HTML Simply-Finds-Potential-Issues-Based-On-Real-World-Browsers.

I think that's a bit too long.

Albert

 

[Ben C]

Because it finds problems in web pages, which is what people want when they
look for a validator.


I think that's a bit too long.

Something like "HTML Lint" would be the traditional name for such a
tool.

 

[Richard Rundle]

Because it finds problems in web pages, which is what people want when they
look for a validator.

No. What I (at least) want is for a validator to tell me what's valid.

 

[Albert Wiersch]

No. What I (at least) want is for a validator to tell me what's valid.

If you want to know if a document is technically valid, then you can use the
included DTD based validator in the standard and pro editions of CSE HTML
Validator. Of course if you only use that, then you could be missing quite a
few issues that can't be found by a DTD based validator.

Albert

 

[Andy Dingley]

It's only "phony" if you strictly limit your definition...

So Albert, your snake oil "validator" is perfect and the OP didn't
really have a problem?
What was it - evil Finnish Pixies?