Web site critique

Neo Geshel wrote:

Oli Filth wrote:

Neo Geshel wrote:

Gazza wrote:
Fails CSS validation.

Really?
http://jigsaw.w3.org/css-validator/v...css/global.css

Somehow, I can't understand how you manage to get “Failed” from
“Congratulations! Valid CSS! This document validates as CSS! ”.


However:
http://jigsaw.w3.org/css-validator/v...nentalkit.com/

Although the CSS itself *is* correct. You need to check your MIME
types etc. if you're going to use XHTML and the <?xml ... ?> directive.

Yes, but I have no control over the mime-type that the server puts out
(the site isn’t hosted with us), and a quick search of google hasn’t
turned up anything that I can use in PHP (without throwing an error,
that is). If you have any links that show me how I can get the server to
modify a normal page’s mime-type through PHP (before the page is sent to
the user), I'd be glad to implement it. So far, most everything that I
have come across is related to setting the mime-type of an upload to the
server.

header("Content-Type: ...");

Before any script output. However, there are issues with XHTML MIME
types and Internet Explorer (cos it's lame). Do a Google search for
"xhtml mime explorer" to find your options.

--
Oli

On Mon, 16 May 2005 08:26:36 GMT, Neo Geshel <gotcha (AT) geshel (DOT) org>
wrote:

Just looking for a 10,000 foot overview of this web site:

http://continentalkit.com/

Looking for several things:

=E2=80=A2 Why does IE ignore the other stylesheets? Both Mozilla and Op=
era see=20
the print and handheld stylesheets, but IE ignores them. When one does a =

print preview in Mozilla or Opera, the print preview comes out=20
correctly, but in IE it does not. Same with viewing the site on a=20
cellphone or a handheld device.

This may or may not help. For your "print.css," you've got to go
through and, basically, create a duplicate of your "screen.css" file
but use "display: none" for those classes and IDs you do not want
showing in a printed page. IE may be confused by some of the
directions it is trying to interpret from both the screen.css and
print.css files. IIRC, your print.css file has to account for
everything in the screen.css file. Remember, too, that the proper
measurement for print typefaces is "point," not "pixel".

=E2=80=A2 How is the header image coming out at the very top? Are IE pp=
l able=20
to see the full alpha-transparency? (Notice how the background slides=20
behind the blur when you scroll)

In IE 6.02, it is NOT displaying the transparency.

What's wrong with using a GIF transparency? Or, if you were to
absolutely position your IDs, simply using your existing background
tile as the background for the banner?

And why use Flash for a simple slide show? At 56k, with a great
majority of people still use in the US (~70%), it took over three
minutes for the file to load with no indication in Firefox that a
Flash file was loading. An impatient visitor will just stop loading
the page after xx number of seconds if it otherwise appears that the
page has loaded. As an alternative, a properly optimized GIF animation
would get the same job done for fewer bytes. Overall, without the
Flash file, your page still took ~22 seconds to load at 56K -- about
twice the recommended speed.

Your title header is way too large (~60k) for the purpose it serves.
Your visitors would be better served if you at least incorporated the
image as a background-image in css so as to greatly speed up the
loading of subsequent pages.

Hopefully this will all be helpful to you.

Oli Filth wrote:

Neo Geshel wrote:

Oli Filth wrote:

Neo Geshel wrote:

Gazza wrote:

Fails CSS validation.


Really?
http://jigsaw.w3.org/css-validator/v...css/global.css

Somehow, I can't understand how you manage to get “Failed” from
“Congratulations! Valid CSS! This document validates as CSS!”.


However:
http://jigsaw.w3.org/css-validator/v...nentalkit.com/

Although the CSS itself *is* correct. You need to check your MIME
types etc. if you're going to use XHTML and the <?xml ... ?> directive.

Yes, but I have no control over the mime-type that the server puts out
(the site isn’t hosted with us), and a quick search of google hasn’t
turned up anything that I can use in PHP (without throwing an error,
that is). If you have any links that show me how I can get the server
to modify a normal page’s mime-type through PHP (before the page is
sent to the user), I'd be glad to implement it. So far, most
everything that I have come across is related to setting the mime-type
of an upload to the server.

header("Content-Type: ...");

Before any script output. However, there are issues with XHTML MIME
types and Internet Explorer (cos it's lame). Do a Google search for
"xhtml mime explorer" to find your options.

I’ve already tried that. It throws a major server error that totally
destroys my stylesheet formatting and produces a line of text (the error
message) at the very top (prior to any XHTML code). I’ll try it again
soon so I can pass on the error message.

...Geshel
--
************************************************** ********************
My reply-to is an automatically monitored spam honeypot. Do not use it
unless you want to be blacklisted by SpamCop. Please reply to my first
name at my last name dot org.
************************************************** ********************

Toby Inkster wrote:

Neo Geshel wrote:
But show me a peer-reviewed research paper that indicates the usage of
browsers that have images or css turned off as being larger than the
number of TTS readers, and I’ll gladly change my tune.

Most mobile user-agents don't support CSS.

Most of the ones that I have seen in Canada do, including my relatively
low-end LG-4600 cell phone. Many do it badly, though, which is why my
mobile stylesheet has got only the bare essentials (such as disabling
material not appropriate for mobile agents, such as the footer and flash).

BTW, I don’t know how bad technology adoption is in the states (you guys
are always about 3-5 years behind the rest of the first-world countries
for personal tech), but I don’t know of any color-screen cellphone in
Canada being without a microbrowser that can’t understand CSS. Many new
cell phones in Asia and Europe have microbrowser support that even we in
Canada won’t see for a few months to a year or two.

...Geshel
--
************************************************** ********************
My reply-to is an automatically monitored spam honeypot. Do not use it
unless you want to be blacklisted by SpamCop. Please reply to my first
name at my last name dot org.
************************************************** ********************

Steve Sundberg wrote:

On Mon, 16 May 2005 08:26:36 GMT, Neo Geshel <gotcha (AT) geshel (DOT) org
wrote:
=E2=80=A2 How is the header image coming out at the very top? Are IE pp=
l able=20
to see the full alpha-transparency? (Notice how the background slides=20
behind the blur when you scroll)

In IE 6.02, it is NOT displaying the transparency.

That's not good at all. You should see the title, with a white glow
around it, and the background-image extending right underneath the glow
(and fading out because of the glow). When you scroll, the
background-pattern (because it is fixed) should slide underneath the
white glow and vanish slowly as if the glow was a fog whose strength was
proportional to its proximity to the text.

What's wrong with using a GIF transparency? Or, if you were to
absolutely position your IDs, simply using your existing background
tile as the background for the banner?

GIF cannot do alpha transparency. Besides, it is also a moral issue.
Check out the “No GIF’s” link at the bottom of the page.

I also don't use the background-image as a part of the header image
because the content is center-aligned. Any difference in screen width
would cause a discontinuity between the background-image for the page
and the background-pattern for the header image, which would NOT look
good at all. My only recourse would be a solid colour for the
background, which is not what I wanted to do.

And why use Flash for a simple slide show? At 56k, with a great
majority of people still use in the US (~70%), it took over three

Ah. Really? Dial up usage in Canada is down to about 30-35%. The rest
(65-70%) are all broadband connections. We are just behind Korea (80%)
in broadband penetration, although our broadband-over-cellphone systems
are nowhere as good as theirs (we inherited parts of the really nasty US
standards kludge that is preventing the adoption of true 3G services in
North America).

minutes for the file to load with no indication in Firefox that a
Flash file was loading. An impatient visitor will just stop loading
the page after xx number of seconds if it otherwise appears that the
page has loaded. As an alternative, a properly optimized GIF animation

GIF animation? I might as well go with MNG. Besides, how do you do a
**RANDOM** GIF animation? Look closely. All the images are in a
statistically random order, with no sequence of images appearing twice
in a row.

would get the same job done for fewer bytes. Overall, without the
Flash file, your page still took ~22 seconds to load at 56K -- about
twice the recommended speed.

Your title header is way too large (~60k) for the purpose it serves.

Good point. I’ll see about compressing it a bit more.

Your visitors would be better served if you at least incorporated the
image as a background-image in css so as to greatly speed up the
loading of subsequent pages.

Hopefully this will all be helpful to you.

...Geshel
--
************************************************** ********************
My reply-to is an automatically monitored spam honeypot. Do not use it
unless you want to be blacklisted by SpamCop. Please reply to my first
name at my last name dot org.
************************************************** ********************

Neo Geshel mumbled the following on 16/05/2005 18:52:

Gazza wrote:

Fails CSS validation.


Really?
http://jigsaw.w3.org/css-validator/v...css/global.css

Somehow, I can't understand how you manage to get “Failed” from
“Congratulations! Valid CSS! This document validates as CSS! ”.

At the time it didn't. You had a font-style: bold, instead of
font-weight: bold. Why you've changed this, and then claimed that it was
never wrong is beyond me.
Now it just shows an error to do with the mime-type, as mentioned by
another poster:
http://jigsaw.w3.org/css-validator/v...nentalkit.com/

Fails XHTML1.1 validation (right Doctype, wrong MIME type).

Really?
http://validator.w3.org/check?uri=ht...ntalkit.com%2F
Somehow, I can't understand how you manage to get "Failed" from "This
Page Is Valid XHTML 1.1!"

You need to read the specs for XHTML Media Types. It clearly says that
XHTML1.1 SHOULD use application/xhtml+xml as it's mime type. Although
the validator may say that you've passed, it's not perfect, and the
validator shouldn't be relied upon.
If you can't alter the mime-type (as you mention on another post),
(though PHP can manage it quite well), then change the Doctype to
XHTML1.0 Strict and be done with it.

Fails Automated Section 508.
Fails Automated WAI Level A, let alone Level AAA.


I don’t know what you used to do the validation on these, but this site:
http://webxact.watchfire.com/

Cynthia. But again, checking now, the page passes automated checks on
WAI and S508 - these pages have clearly been changed since the first
time I checked. As above, the validators/lints should only be used as a
first step - passing the checks is only part of the process; however,
failing the automated checks does mean there was something wrong somewhere.

says that I only fail level AAA, and by only one checkpoint which I
can’t do anything about because of the way that the feedback form must
be processed by the server-side code.

A quick check using WebXact shows the error as not providing default
text in your form elements. This is trivial to do, and you could use
JavaScript to remove the default text onfocus. Without JS enabled, it
could be left to the user to remove code, but at least they know what is
expected of them to be entered.

My hands are tied. If you really

wanna be a neurotic bitch about it, I’ll change the text to AA.

You asked for comments, I'm giving them - you not being able take those
comments and resorting to personal name-calling says more about you than
me. A fail on one checkpoint is, after all, still a fail. Otherwise we
might all claim our sites reach AAA level, except for the 10
check-points we fail on. Just because the server-side script might be
out of your control, is no excuse for claiming something that isn't
true. Whether you want to change the text from AAA to AA is left to your
own conscious and professionalism.

And as for Section 508, the same site clocks me in at having (once
again!) only one (1) error, which is a dubious issue, since I provide
exactly what they demand (the link to the plugin is INSIDE the <object
tag; users that don’t have flash should see the link... this was a
check that I was asking about. As well, there is a static image that
should also be seen if the user doesn’t have flash).

I can only suggest that perhaps the link needs to be outside the image.
Perhaps this could be included in the footer tech details somewhere.

It's all very well trying to impress people, but to those in the know,
you look a bit silly when you can't back those claims up...


I just did. Who's looking silly now?

Hey, it's your website, your initial mistakes.

With images disabled, you have no obvious <h1> and with CSS disabled
(but images on), you have it twice. On IE6 I don't see any image, even
with images on.


Point taken. But my main thrust of support is for TTS readers. Not ppl
with images turned off or CSS disabled.

So those people aren't potential customers for your product? Great sales
technique, I'm sure...

Your title for the acronym JPEG is wrong, it should be Photographic
Experts, not Photographer�s Expert.


Mea Culpa. That, at least, can be “fixed”, and it has been.

In your own words then, you've made changes to the site, and that, I
strongly suspect, include fixing the failed CSS/S508/WAI checks
mentioned above.
--
Gazza
Mobile Number Network Checker - http://mnnc.net/
Creative writing & Poems - http://garyjones.co.uk/
Leovanna Leonbergers - http://leovanna.co.uk/

kchayka wrote:

Accessibility is a heck of a lot more than just accommodating blind
users. If you really had a clue about this subject, you would know that.

I predict within 5 years (assuming it has not already happened) you will
see lawsuits based on accessibility and language. It is bound to happen.

--
-=tn=-

[follow-ups set to news:alt.html]

While the city slept, Travis Newbury (TravisNewbury (AT) hotmail (DOT) com) feverishly
typed...

kchayka wrote:
Accessibility is a heck of a lot more than just accommodating blind
users. If you really had a clue about this subject, you would know
that.

I predict within 5 years (assuming it has not already happened) you
will see lawsuits based on accessibility and language. It is bound to
happen.

Two words. Olympics, Australian. Rearrange into a common phrase or saying!
;-) http://www.webaim.org/coordination/law/australia/#case

Reading another (rather short) article on this case[1], apparantly IBM said
it would cost $2,000,000 and a year to retrofit alt text to the site[2].
Probably overexagerated, but when you consider it would have cost ~$0 and
hardly any more time to do it properly in the first place...

[1] http://slashdot.org/articles/00/08/28/1143249.shtml
[2] Yes, I know there is more to accessibility than alt text. Just saying
what I read in the article.

Cheers,
Nige

--
Nigel Moss http://www.nigenet.org.uk
Mail address will bounce. nigel (AT) DOG (DOT) nigenet.org.uk | Take the DOG. out!
"Your mother ate my dog!", "Not all of him!"

Neo Geshel wrote:

Oli Filth wrote:

header("Content-Type: ...");

Before any script output. However, there are issues with XHTML MIME
types and Internet Explorer (cos it's lame). Do a Google search for
"xhtml mime explorer" to find your options.


I’ve already tried that. It throws a major server error that totally
destroys my stylesheet formatting and produces a line of text (the error
message) at the very top (prior to any XHTML code). I’ll try it again
soon so I can pass on the error message.

Sounds like something on your server is screwy then.


--
Oli

Toby Inkster wrote:

What link? Your link styling (or lack thereof) makes it pretty difficult
to distinguish links and plain text. BTW, the patent on gifs expired
some time ago.
...In the US. Elsewhere, there are still patents on GIF.

Has anyone ever been sued over the gif patent? I mean an individual, or
company < 500 employees.

--
-=tn=-