Browser validation

[ato_zee]

Are there any W3C validated browsers?

So if the code and style sheet validate at W3C
the it will display as intended.

It all seems arse about face, you need compliant
browsers, or even a compliant, browser before
you consider code validation.

 

[Nico Schuyt]

Are there any W3C validated browsers?
So if the code and style sheet validate at W3C
the it will display as intended.
It all seems arse about face, you need compliant
browsers, or even a compliant, browser before
you consider code validation.

Doesn't make sense to me. Browsers have to follow the W3C-specs, on the
other hand they need to render sites as much as possible as the builder has
intended to.

 

[Els]

Browsers have to follow the W3C-specs, on the
other hand they need to render sites as much as possible as the builder has
intended to.

Don't you think that this impossible task would be easier to fulfill
if the builder would code his sites following the specs, instead of
sending his best *intentions* along with his non-standard code?

 

[Nico Schuyt]

Don't you think that this impossible task would be easier to fulfill
if the builder would code his sites following the specs, instead of
sending his best *intentions* along with his non-standard code?

No It's like the accessiblity problems: It's impossible to achieve that
every webbuilder on earth will follow the specs, so it's much more practical
to use browsers that make the best of it.

 

[Paul Watt]

No It's like the accessiblity problems: It's impossible to achieve
that every webbuilder on earth will follow the specs, so it's much more
practical to use browsers that make the best of it.

Give a chimp a stick......
--
Cheers

Paul
le singe est dans l'arbre
http://www.paulwatt.info

 

[Els]

No It's like the accessiblity problems: It's impossible to achieve that
every webbuilder on earth will follow the specs, so it's much more practical
to use browsers that make the best of it.

Tell me Nico, what is easier: build a browser that does what one set
of rules says, or build a browser that does what millions of different
sets of rules say?

Do you have any suggestions for the soccer rules too? Should the
referee judge the game by the intentions of the players ("honest, I
think the rules should allow for kicking the opponent in the shins, as
it will surely stop him from preventing my team scoring a goal"), or
should the players play by the rules?

 

[ato_zee]

So your code and css validate at W3C, yet each browser renders
differently, IE can render the page different to Firefox,
and Safari on OS X different again.
So why doesn't valid code render the same on each browser?
Quite clearly the browsers are not W3C compliant.
So why doesn't W3C validate browsers?

 

[Nico Schuyt]

Tell me Nico, what is easier: build a browser that does what one set
of rules says, or build a browser that does what millions of different
sets of rules say?

The latter is easier than instruct all them webbuilders amd adjust existing
sites.

Do you have any suggestions for the soccer rules too?

I hate soccer But the comparison is wrong: every game is a new one. On
internet there are millions of non- or no longer compliant sites.

 

[Andy Dingley]

Are there any W3C validated browsers?

Amaya. Approximately, some time ago, and for somewhat stretched
meanings of "validated".

You might like to look up the ACID browser test suites and see what the
current status of the pack is. It's not that important for a browser to
be perfect, so long as it's fairly close and you know where it isn't.

 

[Els]

The latter is easier than instruct all them webbuilders amd adjust existing
sites.

What can I say? If you ever have clients who don't understand that,
feel free to send them my way.

 

[Nico Schuyt]

What can I say? If you ever have clients who don't understand that,
feel free to send them my way.

Els, forgive me my ignorance, but what clients do you mean? Neither the
browsers nor the webbuilders are my clients.

 

[Els]

Els, forgive me my ignorance, but what clients do you mean? Neither the
browsers nor the webbuilders are my clients.

Of course not. But your clients just may expect their websites to
a) work in all currently used browsers (that includes oldies like IE5
as well as text browsers and screenreaders)
b) be accessible to everybody, including the blind and those without a
mouse.

 

[Nico Schuyt]

Of course not. But your clients just may expect their websites to
a) work in all currently used browsers (that includes oldies like IE5
as well as text browsers and screenreaders)
b) be accessible to everybody, including the blind and those without a
mouse.

Ahh, I see But it's not that I deny the importance of validation and
accessibility (I try to follow the specs concscientious), the discussion is
about a practical approach of all existing non-compliant sites. A missing
alt text for example should be replaced in a browser with alt=""; fixed
fonts should be ignored (or the compromis in IE: fixed unless changed in
accessibility options).
BTW I stopped testing in IE5 and 5.5. In FF and Opera I only apply the
latest version)

 

[Bergamot]

So your code and css validate at W3C, yet each browser renders
differently, IE can render the page different to Firefox,
and Safari on OS X different again.

What are you doing that gives such a result? Are you looking at
pixel-precision, which is a futile goal, or you are attempting something
too complex for your particular CSS skill level?

Validation only means there are no syntax errors. It cannot detect logic
errors, which is often the trouble with rendering issues, especially
when the developer isn't very knowledgeable in CSS.

IE aside, a carefully thought out design developed by someone competent
in CSS should render comparably in modern graphical browsers. It should
also degrade well in other browsing situations.

 

[dorayme]

Of course not. But your clients just may expect their websites to
a) work in all currently used browsers (that includes oldies like IE5
as well as text browsers and screenreaders)
b) be accessible to everybody, including the blind and those without a
mouse.

I would fall over backwards if my clients ever said anything as
sophisticated as this. I mean it. It is just not on the radar for
most people, I am talking intelligent people here too. I would
not be surprised if most clients have to a be persuaded by their
website makers to have such thoughts... especially if they were
lucky enough to have one like you to educate them. Methinks the
push for standards is not driven by the market (proof: look at
the state of it!), but by the trail blazers at alt.html and the
like. I can tell you, Els, I am swelling with pride here to know
this brave little band... oops I just fell off my chair, swelled
too much to the right...

 

[=?ISO-8859-1?Q?G=E9rard_Talbot?=]

Followup-to set: alt.html

Are there any W3C validated browsers?

So if the code and style sheet validate at W3C
the it will display as intended.

It all seems arse about face, you need compliant
browsers, or even a compliant, browser before
you consider code validation.

Perfectly conformant browsers do not exist. They all have HTML 4.01
bugs, CSS 2.1 bugs and DOM 2 bugs. But the best conformant ones are (in
descendant order):

Firefox 2.0
Opera 9.02
Safari 2.0.4
Icab 3.03

HTML 4 Conformance tests:
http://www.robinlionheart.com/stds/html4/
http://www.robinlionheart.com/stds/html4/results

Web browser standards support summary
http://www.webdevout.net/browser_support_summary.php

Followup-to set: alt.html

Gérard

 

[=?ISO-8859-1?Q?G=E9rard_Talbot?=]

So your code and css validate at W3C, yet each browser renders
differently, IE can render the page different to Firefox,
and Safari on OS X different again.
So why doesn't valid code render the same on each browser?

For several reasons, valid markup code and valid CSS code may not render
the same layout in each browsers:

1- browsers have often different default css property values for some
elements. Recently, there has been a movement done by browser
manufacturers (Microsoft, Opera, Mozilla, Apple, etc.) to reduce the
numbers of such occurences. E.g.: IE 7, Opera 9, Firefox 1+, Safari 2.x
all have a default 8px margin on the body element... but this was not
the case just 2 years ago.

2- Browsers have bugs. Simple as that. The browsers that have less bugs
are the ones who have established public feedback systems to report
bugs. The longer they've done that, the more correct their
implementations of specifications are.

3- Valid code will have more/better chance to be correctly rendered (as
intended) when the browsers are triggered to use "standards compliant
rendering mode" (document.compatMode == "CSS1Compat"). The browsers
conform more closely to the spec of web languages (HTML and CSS) when in
standards compliant rendering mode.

4- Valid code could still be very complex code, with a large and deep
DOM tree. This is another phenomenon. Sometimes, new comers to CSS give
a class to every element (or so) and they make the DOM tree very deep.
Browsers have a tendency to produce different layout when the dom tree
gets deep and large.

Gérard

 

[Els]

I would fall over backwards if my clients ever said anything as
sophisticated as this. I mean it. It is just not on the radar for
most people, I am talking intelligent people here too. I would
not be surprised if most clients have to a be persuaded by their
website makers to have such thoughts... especially if they were
lucky enough to have one like you to educate them. Methinks the
push for standards is not driven by the market (proof: look at
the state of it!), but by the trail blazers at alt.html and the
like.

If that were true, I don't think I would get all those clients who
want accessible websites. Most of them have never heard of alt.html


Not that it's driven by the market either, but I do think the latest
laws wrt accessibility may have something to do with it. Some people
may be scared to run into problems with the law?
http://www.webproworld.com/viewtopic.php?t=67257

Of course this is about the USA, but somehow my British clients also
want accessible websites these days.

I can tell you, Els, I am swelling with pride here to know
this brave little band... oops I just fell off my chair, swelled
too much to the right...

<g>
(vision of a Martian blowfish..)

 

[Chaddy2222]

If that were true, I don't think I would get all those clients who
want accessible websites. Most of them have never heard of alt.html


Not that it's driven by the market either, but I do think the latest
laws wrt accessibility may have something to do with it. Some people
may be scared to run into problems with the law?
http://www.webproworld.com/viewtopic.php?t=67257

Of course this is about the USA, but somehow my British clients also
want accessible websites these days.

Yes, the UK have the DDA (I think that's what it's called, and we have
the DDA here in Aus as well.
You only need to look at the Sydney Olimpics case for details regarding
this.

 

[dorayme]

If that were true, I don't think I would get all those clients who
want accessible websites. Most of them have never heard of alt.html

I have never heard anyone wanting a website actually talk about
IE5 or the blind or the mouseless or the mouse-shys or PDA's or
mobile phones without leading questions by me... I have made
websites for people in organizations who have IT depts that have
written specifications recommending or requiring accessibility
criteria, true, but these may as well be double-dutch - sorry Els
- to the particular hirer. And from the look of almost every
other page at these organizations, no one seems to take a lot of
notice of these standards however generally competent and
generally useful these "non-assessible" pages are.

But I work for people who appreciate these standards, who can be
led to see they are good things. Just my experience I guess... So
I was thinking that when most people hire people to make a
website, it would be very few indeed who talked the talk.