cam.* pubmeet soon!

1 view
Skip to first unread message

Mark Carroll

unread,
Sep 1, 2002, 12:35:03 AM9/1/02
to
The cam.* pubmeet will be upstairs in the Castle up Castle Hill on
Tuesday 10th September, starting at 7pm, although some may turn up
earlier - whenever I've gone, they've done nice food, so maybe you
don't need dinner first. Lurkers welcome, regulars' mugshots at
http://www.chaos.x-philes.com/home/mark/cam.html

Apologies for the delay in posting this; we were out of town, driving
over three thousand miles in under a week, and I'm only just
recovering. I also won't be able to post a last-minute reminder, being
in England imminently, so maybe someone else can. AFAICT, everyone
(except me!) could make the 10th, and the Castle was the clear
favourite, the Maypole scraping in as possibly being acceptable.

-- Mark

Mark Carroll

unread,
Sep 1, 2002, 9:15:10 AM9/1/02
to
In article <2042758.F...@robinton.llondel.org>,
Dave {Reply Address in.sig} <noone$@llondel.org> wrote:
(snip)
>It was retrying at regular intervals, when it had a go at 11:51 it was
>trusted enough to deliver successfully. I was just interested to see that
>message, not had that one before. Why does it say [Irritated] at the end?

Because that's what it is. (-: IMLE it seems to report what mood the
MTA conversation put it in. Some information about the system is
available from http://www.chiark.greenend.org.uk/~ian/sauce/ including
that "mail from previously-unknown sources is delayed to give them a
chance to try a bait address or get their account cancelled".

It's the most effective antispam I've ever encountered, TBH, and is
free!

-- Mark

Nick Wagg

unread,
Sep 2, 2002, 4:43:37 AM9/2/02
to
Mark Carroll wrote:
>
> The cam.* pubmeet will be upstairs in the Castle up Castle Hill on
> Tuesday 10th September, starting at 7pm, although some may turn up
> earlier - whenever I've gone, they've done nice food, so maybe you
> don't need dinner first. Lurkers welcome, regulars' mugshots at
> http://www.chaos.x-philes.com/home/mark/cam.html

Hmm, don't like the colour scheme. Can't read white text on a light
green background. Black text on a dark blue/grey background is also
pretty difficult.
--
Nick Wagg
TranscenData Europe Ltd, Oakington House, Oakington, Cambridge CB4 5AF
Email: nick...@transcendata.com URL: www.transcendata.com
Tel: +44 (0)1223 237111 Fax: +44 (0)1223 234192

Gareth Marlow

unread,
Sep 2, 2002, 5:04:09 AM9/2/02
to
In article <3D7324B9...@transcendata.com>,

Nick Wagg <n...@fegs.co.uk> wrote:
>Mark Carroll wrote:
>>
>> The cam.* pubmeet will be upstairs in the Castle up Castle Hill on
>> Tuesday 10th September, starting at 7pm, although some may turn up
>> earlier - whenever I've gone, they've done nice food, so maybe you
>> don't need dinner first. Lurkers welcome, regulars' mugshots at
>> http://www.chaos.x-philes.com/home/mark/cam.html
>
>Hmm, don't like the colour scheme. Can't read white text on a light
>green background. Black text on a dark blue/grey background is also
>pretty difficult.

Looking at the source, all of the presentation of the page is done in CSS
rather than HTML. So the only way you can be seeing those colours is if
you're using a CSS-compliant browser. Which means you can turn the colours
off and revert to your browser default if they're causing you problems.

Gareth

--
http://www.chiark.greenend.org.uk/~garethm/ : Gareth Marlow
______________________________________________________________________

" Will you be dinging and donging merrily on high tonight sir?"

Tony Finch

unread,
Sep 2, 2002, 5:38:16 AM9/2/02
to
"Dave {Reply Address in.sig}" <noone$@llondel.org> wrote:
>
>It was retrying at regular intervals, when it had a go at 11:51 it was
>trusted enough to deliver successfully. I was just interested to see that
>message, not had that one before. Why does it say [Irritated] at the end?

SAUCE does "teergrubing", i.e. it slows down the SMTP conversation by
delaying its responses in order to slow down spammers. The size of delay
is a measure of its dislike for the sending host: the longer the delay,
the more angry.

Tony.
--
f.a.n.finch <d...@dotat.at> http://dotat.at/
FORTIES CROMARTY FORTH: SOUTHWEST BACKING SOUTH 5 OR 6 DECREASING 4. MAINLY
FAIR. GOOD.

Mike Pitt

unread,
Sep 2, 2002, 6:30:36 AM9/2/02
to
In article <UBd*3s...@news.chiark.greenend.org.uk>,

Mark Carroll <ma...@chiark.greenend.org.uk> wrote:
>It's the most effective antispam I've ever encountered, TBH, and is
>free!

It also has a past hisotry of being too effective for some. Shall we say
it has proved a little controversial? :-)


Mike

Nick Wagg

unread,
Sep 2, 2002, 9:35:54 AM9/2/02
to
Gareth Marlow wrote:
>
> In article <3D7324B9...@transcendata.com>,
> Nick Wagg <n...@fegs.co.uk> wrote:
> >Mark Carroll wrote:
> >>
> >> The cam.* pubmeet will be upstairs in the Castle up Castle Hill on
> >> Tuesday 10th September, starting at 7pm, although some may turn up
> >> earlier - whenever I've gone, they've done nice food, so maybe you
> >> don't need dinner first. Lurkers welcome, regulars' mugshots at
> >> http://www.chaos.x-philes.com/home/mark/cam.html
> >
> >Hmm, don't like the colour scheme. Can't read white text on a light
> >green background. Black text on a dark blue/grey background is also
> >pretty difficult.
>
> Looking at the source, all of the presentation of the page is done in CSS
> rather than HTML. So the only way you can be seeing those colours is if
> you're using a CSS-compliant browser. Which means you can turn the colours
> off and revert to your browser default if they're causing you problems.

I could and will...but surely it is not too much to expect sensible
defaults?

Mark Carroll

unread,
Sep 2, 2002, 10:05:17 AM9/2/02
to
In article <3D7324B9...@transcendata.com>,
Nick Wagg <n...@fegs.co.uk> wrote:
(snip)

>Hmm, don't like the colour scheme. Can't read white text on a light
>green background. Black text on a dark blue/grey background is also
>pretty difficult.

You're not using Netscape 4 with JavaScript enabled, are you? ISTR it
has some serious bugs when displaying CSS documents (doing things
wrongly rather than just not doing them). If you actually look at the
HTML and CSS you should see that nowhere do I specify white text on
light green, or black text. Really, I'd be very interested to learn
what browser and platform you're using, so I can start to get an idea
what's going on. (cam.misc experts: am I right in thinking that
something is seriously broken at Nick's end, or is it all really my
fault?)

FWIW, the combinations with poorest contrast on that page should be
#ccccff on #003333 and #6666ff on #000000.

-- Mark

Mark Carroll

unread,
Sep 2, 2002, 10:12:41 AM9/2/02
to
In article <3D73693A...@transcendata.com>,
Nick Wagg <n...@fegs.co.uk> wrote:
>Gareth Marlow wrote:
(snip)

>> you're using a CSS-compliant browser. Which means you can turn the colours
>> off and revert to your browser default if they're causing you problems.
>
>I could and will...but surely it is not too much to expect sensible
>defaults?

You're absolutely correct. I don't have much CSS experience, so I may
have written broken CSS that I'll be happy to try to fix.
Unfortunately, none of the browser/platform combinations on which I've
tested the page produce the problems you describe IIRC, which is
puzzling. Feel free to e-mail me a screenshot of what you're seeing if
you're stuck for what to do on a rainy day; my account can take large
messages.

-- Mark

Gareth Marlow

unread,
Sep 2, 2002, 10:20:30 AM9/2/02
to
In article <Lww*2X...@news.chiark.greenend.org.uk>,

Mark Carroll <ma...@chiark.greenend.org.uk> wrote:
>
>You're absolutely correct. I don't have much CSS experience, so I may
>have written broken CSS that I'll be happy to try to fix.

I think the CSS in and of itself is fine - I think Nick was complaining
simply about your choice of colours. I'm colour blind so can't comment :)

G

--
http://www.chiark.greenend.org.uk/~garethm/ : Gareth Marlow
______________________________________________________________________

I recognised your foul stench when I was brought on board

Jonathan Amery

unread,
Sep 2, 2002, 10:33:52 AM9/2/02
to
In article <S8v*Z9...@news.chiark.greenend.org.uk>,

On the third hand, it's now configurable! :)

--
Jonathan Amery. Nor till that hour shall God's whole will be done.
##### Now, even now, once more from earth to sky,
#######__o Peals forth in joy man's old undaunted cry -
#######'/ 'Earth shall be fair, and all her folk be one!' - Bax

Mark Carroll

unread,
Sep 2, 2002, 10:34:23 AM9/2/02
to
In article <ATx*RZ...@news.chiark.greenend.org.uk>,
Gareth Marlow <gar...@chiark.greenend.org.uk> wrote:
(snip)

>I think the CSS in and of itself is fine - I think Nick was complaining
>simply about your choice of colours. I'm colour blind so can't comment :)

Unfortunately, the colour combinations he describes aren't AFAIK
anything like ones I chose! (-: Not that we have any chances of seeing
exactly the same colours anyway, with different monitor settings and
eyes and things, but I fear that something is more deeply wrong than
just my taste in this instance.

-- Mark

Mark Goodge

unread,
Sep 3, 2002, 8:41:28 AM9/3/02
to

It's fine in IE, Opera and Mozilla, and your CSS is perfectly valid, but
it's severely broken in Netscape 4.7 due to that browser's buggy
handling of CSS. See http://www.markshouse.net/misc/cammisc.png for a
screenshot.

Looking at the stylesheet, there are three main issues:

1. Using the generic "font-family: cursive" isn't really a good idea, as
it seems to produce inconsistent results. Elsewhere, it's being rendered
as Comic Sans, which is OK, but on Netscape 4.7 it's Brush Script, which
is close to unreadable.

2. Using "background-color: inherit" in the links will cause problems
for Netscape 4.7. Either remove this clause, or specify the background
colour explicitly. If you want to have different backgrounds for links
in different places, then you'll need to specify different classes and
reference them explicitly using <a class=xxx href=zzz> tags.

3. As Netscape 4.7 doesn't inherit properly, it means you need to
specify a font colour for TD elements if you want it to be anything
other than black (it won't inherit the body font colour). The same
applies for other elements of the table cell text styles - they'll need
to be specified, not inherited.

Mark

Mark Carroll

unread,
Sep 3, 2002, 9:57:51 AM9/3/02
to
[ Warning, rant ahead, although not directed at present company. ]

Gosh, thanks! That's all quite insightful. I was afraid that it was
Netscape 4's complete crapness with CSS that was the problem, because
I gave up on that some time ago so that's why I wouldn't have noticed.
The screen shot is such a breathtaking departure from what the CSS
says that I wonder how people can use Netscape 4 with CSS enabled for
general browsing. Maybe it's just that most pages are still putting
physical markup in the page itself instead of using CSS.

It's easy for me to rant, of course - you'd think Netscape would pick
an actually-readable font (I hate to get into specifying names given
the wide variety of what's available on different systems: for
instance, I have no Comic Sans here), and the way it gets CSS wrong is
absolutely dire. Your description of exactly how it gets CSS wrong is
helpful, though; it almost seems like one could write a little program
that takes valid CSS and turns it into something Netscape 4 won't
vomit at.

In a couple of hours I'll be leaving town for a few weeks so I'll not
get to doing anything on it for a little while. In the meantime, to
get an idea of what I should be inclined to do: Just how many people
rely on Netscape 4 with CSS for browsing anyway? The page should be
fine with CSS disabled, and there are so many better (not necessarily
later) browsers available for some time now - I don't remember the
last time I saw Netscape 4 being used. (The page should even work well
with lynx, modulo its handling of tables. (-:)

I think part of my irritation is just that I'm sick of pandering to
broken stuff. I don't mind if it doesn't meet the latest specs - I try
to do stuff that degrades gracefully, so you don't need a recent
browser - but it's frustrating to have to work around other people's
bugs all the time, where they claim to offer functionality but instead
do something different. Take Java, for example - I try to keep just to
1.1 with no Swing, and still my code manages to break some JVMs that I
then have to code workarounds in for (and leave little comments
explaining why it's curiously arranged so someone doesn't helpfully
"fix" it back). I should admit that, to keep my stress level down, I
gave up on Netscape 4 some time ago, which is why I didn't test the
current page on it - it may even be one of the reasons I had to leave
our company pages in mostly transitional instead of strict.

Obviously, this isn't a rant at Nick, it's a rant at the shoddy state
of software today - especially, trying to write code that works on
many different interpreters (if you can think of HTML and CSS like
that). From the viewer's point of view, it's simply that the regulars
page doesn't work well, it doesn't much matter why.

If there really are a lot of Netscape 4 users out there who don't
realise you need to disable JavaScript (which disables CSS!) for a
sane[1] viewing experience, I'll be willing to do what I can to try to
make the page work better with the help of what Mark wrote here, and
just wonder how they cope with other HTML4/CSS pages. For now, I've
simply removed the reference to the stylesheet, temporarily.

[1] well, saner - it has other non-CSS bugs too to a much greater
extent than I've ever seen in any other browser

-- Mark

Gareth Marlow

unread,
Sep 3, 2002, 10:13:24 AM9/3/02
to
In article <lhC*3-...@news.chiark.greenend.org.uk>,

Mark Carroll <ma...@chiark.greenend.org.uk> wrote:
>[ Warning, rant ahead, although not directed at present company. ]

<rant snipped>

Are you aware of <URL: http://www.alistapart.com/stories/tohell/ >?

The gist is that you should separate content from presentation (as you
have done) using CSS. You should make efforts to structure the information
architecture of your site to ensure that its content is accessible to
everyone. If you want to change presentation from the browser default, use
CSS. And if your CSS fecks up (as it has in this case) due to buggy,
intermediate support in older browsers which make a half-assed attempt to
implement the standard, shrug your shoulders and move on. You have better
things to do than to spend your time working around bugs in old browsers.

I have some sympathy with this POV.

Gareth


--
http://www.chiark.greenend.org.uk/~garethm/ : Gareth Marlow
______________________________________________________________________

Oh but Sir! It's only waffer thin.

peej

unread,
Sep 3, 2002, 11:50:18 AM9/3/02
to
> http://www.alistapart.com/stories/tohell/

> I have some sympathy with this POV.

:-)
Having been flamed by some of those here on this before I'm kind of ever so
slightly amused that a page with this content is suffering from this
problem. Sorry to see the original page just went all boring though...
surely you're not giving in to the hordes of netscape 4 users who can't
figure out how to upgrade...

http://webstandards.org/act/campaign/buc/


Having worked in Mountain View for a while I'm surprised Netscape's stuff
works at all. Shame on them for letting MS grab that ball.

Phil


Gareth Marlow

unread,
Sep 3, 2002, 10:58:08 AM9/3/02
to
In article <lhC*3-...@news.chiark.greenend.org.uk>,
Mark Carroll <ma...@chiark.greenend.org.uk> wrote:
>
>In a couple of hours I'll be leaving town for a few weeks so I'll not
>get to doing anything on it for a little while. In the meantime, to
>get an idea of what I should be inclined to do: Just how many people
>rely on Netscape 4 with CSS for browsing anyway? The page should be
>fine with CSS disabled, and there are so many better (not necessarily
>later) browsers available for some time now - I don't remember the
>last time I saw Netscape 4 being used. (The page should even work well
>with lynx, modulo its handling of tables. (-:)

There is a kludge. If you link your stylesheet (I nearly ytpo'd
"styleshit" then...) with the following:

<style type="text/css" media="all">@import "foo.css";</style>

NS4 ignores the stylesheet whereas newer browsers don't. This is valid,
standard HTML/XHTML so everything which will render as you intended with
interpret the link correctly, whereas the biggest offender will just
ignore it.

j...@durge.org

unread,
Sep 3, 2002, 11:03:14 AM9/3/02
to
Mark Carroll <ma...@chiark.greenend.org.uk> wrote:
: absolutely dire. Your description of exactly how it gets CSS wrong is

: helpful, though; it almost seems like one could write a little program
: that takes valid CSS and turns it into something Netscape 4 won't
: vomit at.

What you might like to do instead is do browser detection and produce a
very basic site for Netscape 4 users and produce a standards-compliant
one for everyone else who uses something that hasn't got the worst
layout capabilities the world has ever seen.

Jon
--
-------------------------------------------------------------------------
Durge: j...@durge.org http://www.durge.org/~jon/
OnStream: acco...@rowing.org.uk http://www.rowing.org.uk/
-------------------------------------------------------------------------

Gareth Marlow

unread,
Sep 3, 2002, 11:03:39 AM9/3/02
to
In article <0of*ao...@news.chiark.greenend.org.uk>,

Gareth Marlow <gar...@chiark.greenend.org.uk> wrote:
>
>This is valid,
>standard HTML/XHTML so everything which will render as you intended with
>interpret the link correctly, whereas the biggest offender will just
>ignore it.

I had originally intended to write this sentence in English. What I meant
was "If you do this, NS4 will ignore the stylesheet." or some
such. Wibble.

Mark Goodge

unread,
Sep 3, 2002, 11:12:11 AM9/3/02
to
Mark Carroll wrote:

> Your description of exactly how it gets CSS wrong is
> helpful, though; it almost seems like one could write a little program
> that takes valid CSS and turns it into something Netscape 4 won't
> vomit at.

That's actually reasonably simple; you can do it by following two basic
rules:

1. Don't use generics except as the final member of a list.
2. Don't rely on inheritance - always specify explicitly.

The only problem with this approach is that it can make stylesheets
unnecesarily longwinded, especially if you have a lot of different
places where colours and/or fonts need to be described.

But I agree that maybe it's time to stop worrying about long-obsolete
browsers. It's not as if IE is the only decent CSS-compliant browser any
more (which, in the past, was a good reason to provide support for
alternatives, just so that no-one was forced into the clutches of
Microsoft). Current versions of Mozilla, Opera and Netscape all render
CSS equally well.

Mark

Mark Goodge

unread,
Sep 3, 2002, 12:23:17 PM9/3/02
to
Richard Kettlewell wrote:

>
> Gareth Marlow <gar...@chiark.greenend.org.uk> writes:
> > There is a kludge. If you link your stylesheet (I nearly ytpo'd
> > "styleshit" then...) with the following:
> >
> > <style type="text/css" media="all">@import "foo.css";</style>
> >
> > NS4 ignores the stylesheet whereas newer browsers don't. This is
> > valid, standard HTML/XHTML so everything which will render as you
> > intended with interpret the link correctly, whereas the biggest
> > offender will just ignore it.
>
> Err. What's the point? NS4 users can turn off style sheets anyway.
> If they actually want to use its broken CSS implementation, why make
> it hard for them?

Because a lot of them don't realise how broken it is, and blame the
website authors for the problems they have viewing some sites. Sometimes
it's easier to simply deal with such complaints by stopping them at
source.

Mark

Gareth Marlow

unread,
Sep 3, 2002, 12:42:54 PM9/3/02
to
In article <wwvwuq3...@rjk.greenend.org.uk>,

Richard Kettlewell <richard+n...@sfere.greenend.org.uk> wrote:
>
>Err. What's the point? NS4 users can turn off style sheets anyway.
>If they actually want to use its broken CSS implementation, why make
>it hard for them?

In what way is this making it hard for them? This change is at the page
level. The NS4 user is going to see a perfectly legible, well-structured
page, just using the default fonts and sizes for their browser. The NS4
user will see what the NS2 user, or the lynx user sees.

By making that small change to the way that the style sheet is linked to
his page, Mark could ensure that, even though their browser is buggy, NS4
users could still read his page, which they'll find pretty difficult to do
at the moment. If you relied on them going in to switch off the CSS to see
the page, a third of them wouldn't know what the problem was in the first
place, a third of them wouldn't know how to do it even if they knew the
problem was with their browser and a third of them wouldn't be
arsed.

Chris

unread,
Sep 3, 2002, 1:18:57 PM9/3/02
to
Gareth Marlow wrote:
> In article <lhC*3-...@news.chiark.greenend.org.uk>,
> Mark Carroll <ma...@chiark.greenend.org.uk> wrote:
>
>>[ Warning, rant ahead, although not directed at present company. ]
>
>
> <rant snipped>
>
> Are you aware of <URL: http://www.alistapart.com/stories/tohell/ >?

Yeah. What a load of shite.

The problem with most of the arguments in favour of using
CSS (and physical formatting in general) for web pages is
that they are written by people who, by and large, believe
that their graphic design skills improve the appearance
(and, for those who know the word, usability) of their
sites. This attitude is also responsible for the prevalence
of `Flash'.

But it's not true. Not even a little bit. Most of these
sites are equally ugly -- and harder to read -- with the
web design turned on as with it turned off. A particular
irritation is attempts to control font size, making pages
entirely unreadable until you press ^+ or whatever.

Unfortunately, stripping all physical design from web pages
seems quite difficult to do usefully. I wrote code to do
this for the BBC news site when they broke the `low graphics'
version of their pages:

http://ex-parrot.com/~chris/images/bbc-vs-unbbc.png

but doing this in general will probably break many of the
sites which one would want to be able to use. (The above
example is done by a program which knows quite a few
heuristics about the BBC news web design, which is a bit
of a maintenance nightmare.)


It's nice to see browser authors getting to grips with
things like CSS, though. Once they've done that, they might
want to work on something actually important, such as the
abysmal state of character set handling in form submissions:

http://ppewww.ph.gla.ac.uk/~flavell/charset/form-i18n.html


--
Chris Lightfoot, chris at ex dash parrot dot com; http://ex-parrot.com/~chris/
``I have several monomanias.'' (Edward Teller)

Mark Goodge

unread,
Sep 3, 2002, 1:41:55 PM9/3/02
to
On Tue, 03 Sep 2002 18:18:57 +0100, Chris put finger to keyboard and
typed:

>Gareth Marlow wrote:
>> In article <lhC*3-...@news.chiark.greenend.org.uk>,
>> Mark Carroll <ma...@chiark.greenend.org.uk> wrote:
>>
>>>[ Warning, rant ahead, although not directed at present company. ]
>>
>>
>> <rant snipped>
>>
>> Are you aware of <URL: http://www.alistapart.com/stories/tohell/ >?
>
>Yeah. What a load of shite.
>
>The problem with most of the arguments in favour of using
>CSS (and physical formatting in general) for web pages is
>that they are written by people who, by and large, believe
>that their graphic design skills improve the appearance
>(and, for those who know the word, usability) of their
>sites.

Which is precisely *why* they should be using CSS for the visual
markup rather than HTML. That way, if it's poorly done or doesn't work
on your browser, you can turn it off or use your own preferred
stylesheet. Visual markup in HTML (using <font color=xx> and <font
size=xxx>, etc is what really sucks.

>This attitude is also responsible for the prevalence
>of `Flash'.

Flash and CSS are diametrically opposed, in terms of web design. The
former is an attempt to create a set of open standards for visual
markup that will work on any browser and degrade gracefully where they
are not supported. Flash is a proprietory tool that requires a plugin
in order to be viewed.

>But it's not true. Not even a little bit. Most of these
>sites are equally ugly -- and harder to read -- with the
>web design turned on as with it turned off.

That's a subjective opinion, on the whole. For example I prefer the
high-graphics version of the BBC site, as it's much easier to navigate
than the low-graphics version. But the low-graphics version is handy
when connecting via a dial-up, where the loss of navigability is more
than compensated for by the quicker loading. A lot depends on the
circumstances in which a page is viewed.

>A particular
>irritation is attempts to control font size, making pages
>entirely unreadable until you press ^+ or whatever.

That, unfortunately, is a consequence of a major cockup perpetrated by
the designers of the first graphical browsers which has been copied by
all the other browser designers ever since. But, again, CSS offers the
best practical solution to this in that it allows the web author to
suggest font sizes (both relative and absolute) while leaving final
control firmly in the hands of the user.

Mark
--
http://www.good-stuff.co.uk
"Did I tell you it was wine when really it was water?"

Gareth Marlow

unread,
Sep 3, 2002, 2:00:06 PM9/3/02
to
In article <al2qu1$p9e$1...@pegasus.csx.cam.ac.uk>,

Chris <ch...@x.invalid> wrote:
>Gareth Marlow wrote:
>> In article <lhC*3-...@news.chiark.greenend.org.uk>,
>> Mark Carroll <ma...@chiark.greenend.org.uk> wrote:
>>
>>>[ Warning, rant ahead, although not directed at present company. ]
>>
>>
>> <rant snipped>
>>
>> Are you aware of <URL: http://www.alistapart.com/stories/tohell/ >?
>
>Yeah. What a load of shite.

Huh? on <URL: http://ex-parrot.com/~chris/html.html >, you explain why
your site is written in vanilla html 4.0 - which is completely fair
enough. The article referenced above advocates exactly the same thing. It
suggests to people that if they want to do presentational stuff they
should do it in CSS. That way if people's browsers can't do CSS, it won't
matter, and if people don't want to use CSS, or use their own style
sheets, that won't matter either.

>The problem with most of the arguments in favour of using
>CSS (and physical formatting in general) for web pages is
>that they are written by people who, by and large, believe
>that their graphic design skills improve the appearance
>(and, for those who know the word, usability) of their
>sites. This attitude is also responsible for the prevalence
>of `Flash'.

We *are* talking about the same article, right? You did read it? Including
the paragraph under "Why have you done this terrible thing?" which begins
"Designing in accordance with these standards..."?

Please forgive me if I'm misrepresenting your position here. Your page
above suggests that, as the majority of people who are designing web pages
don't add any value to the content of the page, they shouldn't
bother. Fair enough. But given that almost all of the web has more
presentational content than this, whether you like it or not, is it not
more sensible to encourage web designers to go forward in the way
described above?

BTW, why *are* your hyperlinks red? :)

G


--
http://www.chiark.greenend.org.uk/~garethm/ : Gareth Marlow
______________________________________________________________________

They've got lumps of it round the back

Chris

unread,
Sep 3, 2002, 2:23:51 PM9/3/02
to
Mark Goodge wrote:
> On Tue, 03 Sep 2002 18:18:57 +0100, Chris put finger to keyboard and
> typed:

> stylesheet. Visual markup in HTML (using <font color=xx> and <font


> size=xxx>, etc is what really sucks.

Nevertheless, web designers using CSS do just as badly as
those using <font...> and <table...>.

>>This attitude is also responsible for the prevalence
>>of `Flash'.
>
>
> Flash and CSS are diametrically opposed, in terms of web design.

Both are frequently used to achieve inappropriate and
unhelpful physical design effects. Flash at least works
under Netscape 4, which I suppose is one aspect in which
it is diametrically opposed to CSS.

(Yes, I know what CSS and Flash are, and how they are
designed. It's probably worth noting that Flash is an
open standard in some limited sense. The problem is not
exclusively with the technologies or their intention,
but with the uses which are made of them. In particular,
from time to time one sees an entertaining use of Flash,
such as the recent spoof traffic safety film which was
widely circulated. But that doesn't really compensate
for sites which use Flash for their entire content.)

>>But it's not true. Not even a little bit. Most of these
>>sites are equally ugly -- and harder to read -- with the
>>web design turned on as with it turned off.
>
>
> That's a subjective opinion, on the whole. For example I prefer the
> high-graphics version of the BBC site, as it's much easier to navigate
> than the low-graphics version.

I find this surprising. In particular, the high-graphics
version uses smaller fonts and yet gets fewer stories per
page, hides links to other sections behind images, and
commits various other sins. It also has atrocious
accessibility and produces very large pages. For instance,
the front page of the BBC in the high graphics page is
66k (not including images); with design stripped out, it
is 12k. Naturally, the version with `design' takes longer
to render, too, since it is full of physical markup and
CSS stuff.

> But the low-graphics version is handy
> when connecting via a dial-up, where the loss of navigability is more
> than compensated for by the quicker loading. A lot depends on the
> circumstances in which a page is viewed.

I'm still surprised that you find the high graphics version
is easier to navigate. It certainly isn't in my experience,
and nor can I see that it more closely follows current
usability dogma.

>>A particular
>>irritation is attempts to control font size, making pages
>>entirely unreadable until you press ^+ or whatever.
>
>
> That, unfortunately, is a consequence of a major cockup perpetrated by
> the designers of the first graphical browsers which has been copied by
> all the other browser designers ever since. But, again, CSS offers the
> best practical solution to this in that it allows the web author to
> suggest font sizes (both relative and absolute) while leaving final
> control firmly in the hands of the user.

Really? I thought it was the Microsoft Windows screen
resolution problem.


It seems to me that you are arguing for a particular way
to achieve physical markup on the web. My view broadly is
that every time anybody tries it, it turns out to be a
cock-up. Generally, the pages which work best are those
which have the absolute minimum of physical design and
allow the user to choose how they should appear by
configuring their browser. This is why lynx and links are
such an effective way to browse the web.

Now, CSS can be turned off or overridden with the user's
own choice of styles. (Or, at least, it ought to be
possible to -- I can't find the option in this version of
Mozilla. There's something plausible in the `Profile
Manager' but I can't see how to make it ignore other
peoples' style sheets entirely.) But the problem is that
many web sites are built entirely on the assumption that
the user is not going to do this.

--
Chris Lightfoot, chris at ex dash parrot dot com; http://ex-parrot.com/~chris/

``A little inaccuracy can save tons of explanation.'' (Saki)

peej

unread,
Sep 3, 2002, 3:40:38 PM9/3/02
to
Chris wrote:
> The problem with most of the arguments in favour of using CSS [...] is

that they
> are written by people who, by and large, believe that their graphic design
skills
> improve the appearance (and, for those who know the word, usability) of
their sites. [...]

I'm terribly sorry, but I don't see what bearing that has on the issue under
discussion. Your artistic judgement, good though it may be, shouldn't impact
the quality of the systems used to display web pages. Are you sure you're
not confusing the content of your argument with the presentation of their
web pages?

I've been here long enough not to be surprised that you'd try to argue
against CSS and web standards, but I'm impressed that you can suggest that
such things run against "usability". Rock on!

Phil


Mark Goodge

unread,
Sep 3, 2002, 2:56:14 PM9/3/02
to
On Tue, 03 Sep 2002 19:23:51 +0100, Chris put finger to keyboard and
typed:

>Mark Goodge wrote:
>> On Tue, 03 Sep 2002 18:18:57 +0100, Chris put finger to keyboard and
>> typed:
>
>> stylesheet. Visual markup in HTML (using <font color=xx> and <font
>> size=xxx>, etc is what really sucks.
>
>Nevertheless, web designers using CSS do just as badly as
>those using <font...> and <table...>.

Except that I can override it when it's done with CSS.

I dislike bad design as much as anyone. But if bad design is going to
be inflicted in an undeserving website, then at least inflict it in
such a way as to allow users to mitigate its effects.

>>
>> That's a subjective opinion, on the whole. For example I prefer the
>> high-graphics version of the BBC site, as it's much easier to navigate
>> than the low-graphics version.
>
>I find this surprising. In particular, the high-graphics
>version uses smaller fonts and yet gets fewer stories per
>page, hides links to other sections behind images, and
>commits various other sins. It also has atrocious
>accessibility and produces very large pages. For instance,
>the front page of the BBC in the high graphics page is
>66k (not including images); with design stripped out, it
>is 12k. Naturally, the version with `design' takes longer
>to render, too, since it is full of physical markup and
>CSS stuff.

Over broadband, the size is rarely an issue. The smaller number of
stories on the front page is countered by the easy to use links to the
various sections. I also like the small boxes down the right that go
straight to the more whimsical stories.

I accept that this entirely my subjective impression, and I'm not
arguing with anyone who prefers the low-graphics version. All I'm
trying to point out here is precisely this: that design is primarily a
subjective issue.

>
>>>A particular
>>>irritation is attempts to control font size, making pages
>>>entirely unreadable until you press ^+ or whatever.
>>
>>
>> That, unfortunately, is a consequence of a major cockup perpetrated by
>> the designers of the first graphical browsers which has been copied by
>> all the other browser designers ever since. But, again, CSS offers the
>> best practical solution to this in that it allows the web author to
>> suggest font sizes (both relative and absolute) while leaving final
>> control firmly in the hands of the user.
>
>Really? I thought it was the Microsoft Windows screen
>resolution problem.

It's a bit of both. It could be fixed by either, but in fact it's
fixed by neither.

>Now, CSS can be turned off or overridden with the user's
>own choice of styles. (Or, at least, it ought to be
>possible to -- I can't find the option in this version of
>Mozilla. There's something plausible in the `Profile
>Manager' but I can't see how to make it ignore other
>peoples' style sheets entirely.) But the problem is that
>many web sites are built entirely on the assumption that
>the user is not going to do this.

Which, in something like 90% of cases, is an entirely accurate
assumption. Sometimes it's necesary to take account of the fact that
the majority of your users are tecfhnically illiterate.

Mark
--
http://www.good-stuff.co.uk
"Sometimes everything is wrong"

Chris

unread,
Sep 3, 2002, 3:02:53 PM9/3/02
to
Gareth Marlow wrote:
> In article <al2qu1$p9e$1...@pegasus.csx.cam.ac.uk>,
> Chris <ch...@x.invalid> wrote:
>
>>Gareth Marlow wrote:

>>>Are you aware of <URL: http://www.alistapart.com/stories/tohell/ >?
>>
>>Yeah. What a load of shite.
>
>
> Huh? on <URL: http://ex-parrot.com/~chris/html.html >, you explain why
> your site is written in vanilla html 4.0 - which is completely fair
> enough. The article referenced above advocates exactly the same thing. It
> suggests to people that if they want to do presentational stuff they
> should do it in CSS. That way if people's browsers can't do CSS, it won't
> matter, and if people don't want to use CSS, or use their own style
> sheets, that won't matter either.

The problem is not with using a browser which does not
support CSS. Often, this is an effective way to access
the content of a CSS site. However, there are many sites
which rely heavily on their physical design and which
do not make much sense without it. I don't know what
fraction of users do specify their own style sheet. I
would guess that it is vanishingly small-- probably far
smaller than the fraction of users who would find it
easier to browse the web in this mode.

>>The problem with most of the arguments in favour of using
>>CSS (and physical formatting in general) for web pages is
>>that they are written by people who, by and large, believe
>>that their graphic design skills improve the appearance
>>(and, for those who know the word, usability) of their
>>sites. This attitude is also responsible for the prevalence
>>of `Flash'.
>
>
> We *are* talking about the same article, right? You did read it? Including
> the paragraph under "Why have you done this terrible thing?" which begins
> "Designing in accordance with these standards..."?

Sure. That bit is about how designing things in CSS is
so much better than buggering about with font tags and
tables. Fair enough; I guess I'd rather shoot my foot
off than lose the whole leg, but it's hardly a ringing
endorsement. Claiming that the CSS standard `makes the
web accessible to all' is a nonsense: the web was always
`accessible to all', except for the best efforts of web
designers to render it inaccessible through the use of
physical formatting. CSS might help, but so far it
doesn't seem to be doing so.

> Please forgive me if I'm misrepresenting your position here. Your page
> above suggests that, as the majority of people who are designing web pages
> don't add any value to the content of the page, they shouldn't
> bother. Fair enough. But given that almost all of the web has more
> presentational content than this, whether you like it or not, is it not
> more sensible to encourage web designers to go forward in the way
> described above?

I'm not certain about that. What is really needed is for
the majority of web designers to appreciate that there exist
platforms other than the one they are using. That, or some
way to enforce accessibility requirements. Yes, I know that
CSS is supposed to fix this, but it doesn't.


The other annoying thing about this is the preaching tone
of the `web standards upgrade' outfit or whatever they are
called. For instance,

Standards-compliant browsers need not be bloated and
processor-intensive. Many require less computing power
than the 4.0 browsers did. We don't want to name names, but
if you check the WaSP's Browser Upgrades page, you will
find one or more browsers that can run on your existing
system just fine.

Well, you might believe that, until you visit the page
they suggest. There you discover that the available browsers
are Internet Explorer, Mozilla and Konqueror. The reason
that they are reluctant to `name names' is presumably that
the available selection is so poor. Internet Explorer is
probably OK if you use Microsoft Windows and don't care
about your personal data; Mozilla is big and bloated, if
now relatively free of catastrophic bugs; and Konqueror is
only available for those using KDE (or at least who are
prepared to install it). None of those browsers runs on
small machines such as Palm Pilots (though I see that
`Pocket' Internet Explorer now has CSS) nor those which
run on mobile 'phones. So, for all the wittering about how
great CSS will be for users of downscale systems, you're
still going to get that patronising homily about getting
`a browser which supports web standards' when you try to
view these sites on your mobile 'phone.

Of course, in the future these things may all be fixed.
No doubt I should try to do a patch for Mozilla which puts
a big button on the toolbar to turn off all of the
CSS and physical formatting in the page being viewed.
Perhaps soon the web browsers in mobile 'phones will begin
to support CSS, or alternatively all of the web designers
who put that web standards homily on their pages will
become so pleased with themselves that they disappear up
their own behinds and suffocate.

> BTW, why *are* your hyperlinks red? :)

Heh. A good question.

No very good reason, I'm afraid. I should probably fix them
but that requires writing one of those irritating and
easy-to-go-wrong perl one-liners.


--
Chris Lightfoot, chris at ex dash parrot dot com; http://ex-parrot.com/~chris/

A person is only as big as the thing that makes him angry.

Chris

unread,
Sep 3, 2002, 3:17:52 PM9/3/02
to
peej wrote:
> Chris wrote:
>
[quoting fixed]

>>The problem with most of the arguments in favour of using
>>CSS [...] is that they are written by people who, by and
>>large, believe that their graphic design skills improve
>>the appearance (and, for those who know the word, usability)
>>of their sites. [...]
>
> I'm terribly sorry, but I don't see what bearing that has on the issue under
> discussion. Your artistic judgement, good though it may be, shouldn't impact
> the quality of the systems used to display web pages. Are you sure you're
> not confusing the content of your argument with the presentation of their
> web pages?

I'm not quite sure what you mean here. Ah, I see, it's an
incoherent flame. Splendid!


It's not really, in my view, an argument about artistic
judgment. Putting text in images is inferior to putting
text in text; fixing the colour of text is inferior to
allowing the user to choose it; forcing a user to view
text in a particular page width or font size is inferior
to allowing the user to select it. These are technical
matters and have nothing to do with artistic judgment.

The fact that most of the web design I see turns out to
be shite obviously is a judgment, although no doubt
opinions as to whether it is `artistic' judgment will
vary.

Now, in principle, I agree that CSS should make these
choices available to users, but browsers are not good
at exposing this functionality and most web designers
don't bother to make their pages work well when the
user has chosen to override their design choices. (A
typical example is the floating menu thingies which
are now popular.) I don't suppose that CSS has made
the situation much worse, but it's hardly a significant
improvement, especially given the holier-than-thou
attitude of its adherents.

> I've been here long enough not to be surprised that you'd try to argue
> against CSS and web standards, but I'm impressed that you can suggest that
> such things run against "usability". Rock on!

A web designer who constrains my ability to choose the
appearance of a web site has decreased its usability.
That's pretty obvious, isn't it?

--
Chris Lightfoot, chris at ex dash parrot dot com; http://ex-parrot.com/~chris/

``The reason that the sun never set on the British Empire
is that God wouldn't trust an Englishman in the dark.'' (Duncan Spaeth)

Mark Goodge

unread,
Sep 3, 2002, 4:07:28 PM9/3/02
to
On Tue, 03 Sep 2002 20:17:52 +0100, Chris put finger to keyboard and
typed:

>


>A web designer who constrains my ability to choose the
>appearance of a web site has decreased its usability.
>That's pretty obvious, isn't it?

Only if you can improve on its usability by doing things that you have
control over. Which is only possible if the author gives you that
control by using stylesheets to suggest visual design, instead of
including it directly in HTML.

Unless you're suggesting that all webpages are better off with no
attempt at visual design at all (which would be a ludicrous position
and one I'm certain you don't hold), I can't see any other reason to
oppose the use of CSS. Or am I missing something that you're taking
for granted but haven't actually stated yet?

Mark
--
http://www.good-stuff.co.uk
"And when you play you feel all right"

Gareth Marlow

unread,
Sep 3, 2002, 4:12:49 PM9/3/02
to
In article <al31t0$n6$1...@pegasus.csx.cam.ac.uk>, Chris <ch...@x.invalid> wrote:

>peej wrote:
>
>Now, in principle, I agree that CSS should make these
>choices available to users, but browsers are not good
>at exposing this functionality and most web designers
>don't bother to make their pages work well when the
>user has chosen to override their design choices.

I think that's because most web design at the moment *aren't* doing what's
being advocated. They're using a bastard hybrid of HTML 3.2 tables-based
page layout and some of the typographical tweaking of CSS to alter
line spacing and margins. This is slightly better than 1x1 transparent gif
layout but only slightly. If you do this kind of thing, pages degrade very
badly the further away from IE 8.5/Netscape 10 you get.

As soon as you completely separate the presentation from the structure you
get rid of all of these problems, save for buggy browsers which only half
implement CSS. Such as NS4, as we've already seen.

>(A
>typical example is the floating menu thingies which
>are now popular.) I don't suppose that CSS has made
>the situation much worse, but it's hardly a significant
>improvement, especially given the holier-than-thou
>attitude of its adherents.

You are confusing proponents with ignorant users. Proponents would say
that, if you want to do pop-up, cascading menus, do them like this:

<URL: http://www.dhtmlcentral.com/tutorials/tutorials.asp?page=1&id=14 >
using CSS *properly*. That way, you don't shaft your site if the browser
doesn't support CSS or it's turned off. I think this *is* a significant
improvement on the previous situation where these things would be
implemented in such a way that they would only work on one browser, or
you'd get 4 different versions of the same code because of nuances in
browser support.

>A web designer who constrains my ability to choose the
>appearance of a web site has decreased its usability.
>That's pretty obvious, isn't it?

I'm not sure we're going to get anywhere with this conversation because I
think you've already decided that anyone who alters anything about the
presentation of html to cause it to deviate from the browser default is
imposing their design choices on you, and that it's equally bad whether
this is done with a whole load of tables/font tags/image maps as it is
when done with CSS2 absolute positioning. The first of these points is
valid, although I disagree with it. The second is not.

Gareth

Gareth Marlow

unread,
Sep 3, 2002, 4:19:13 PM9/3/02
to
In article <wwvd6ru...@rjk.greenend.org.uk>,
Richard Kettlewell <richard+n...@sfere.greenend.org.uk> wrote:

>Gareth Marlow <gar...@chiark.greenend.org.uk> writes:
>> Richard Kettlewell <richard+n...@sfere.greenend.org.uk> wrote:
>
>>> Err. What's the point? NS4 users can turn off style sheets
>>> anyway. If they actually want to use its broken CSS
>>> implementation, why make it hard for them?
>>
>> In what way is this making it hard for them?
>
>You were suggesting using a mechanism which made it ignore the style
>sheet. This makes it hard for them to use the software that
>interprets it, i.e. NS4's CSS implementation.

I think I see what you mean; I'm suggesting that the means justify the
ends, i.e. "I know this page looks fine in a non-CSS-supporting browser; I
know this page looks fine in a CSS-supporting browser; I know this page is
fecked in Netscape 4; therefore, I'll make Netscape 4 think that there
isn't a style sheet for this page". Net result is a page which is legible
in more browsers, despite one of them having serious flaws, requiring no
intervention from the user. Isn't that making it easier for the user, not
harder?

Gareth

--
http://www.chiark.greenend.org.uk/~garethm/ : Gareth Marlow
______________________________________________________________________

They've got lumps of it round the back

Gareth Marlow

unread,
Sep 3, 2002, 4:23:55 PM9/3/02
to
In article <nd5anusc8m4luu2h8...@4ax.com>,

Mark Goodge <ma...@good-stuff.co.uk> wrote:
>
>Unless you're suggesting that all webpages are better off with no
>attempt at visual design at all (which would be a ludicrous position
>and one I'm certain you don't hold),

That *is* what he's suggesting. See:

<URL: http://ex-parrot.com/~chris/html.html >

j...@durge.org

unread,
Sep 3, 2002, 4:36:10 PM9/3/02
to
Chris <ch...@x.invalid> wrote:
: The problem with most of the arguments in favour of using

: CSS (and physical formatting in general) for web pages is
: that they are written by people who, by and large, believe
: that their graphic design skills improve the appearance
: (and, for those who know the word, usability) of their
: sites. This attitude is also responsible for the prevalence
: of `Flash'.

You're not clumsily lumping people together into categories that just
don't fit them then....

: But it's not true. Not even a little bit. Most of these


: sites are equally ugly -- and harder to read -- with the
: web design turned on as with it turned off. A particular
: irritation is attempts to control font size, making pages
: entirely unreadable until you press ^+ or whatever.

You're not clumsily lumping lots of sites together into categories that
just don't fit them then....

Sorry but I can see where you're coming from only you're painting a
picture with far too broad a brush.

j...@durge.org

unread,
Sep 3, 2002, 4:43:17 PM9/3/02
to
Chris <ch...@x.invalid> wrote:
: It's not really, in my view, an argument about artistic

: judgment. Putting text in images is inferior to putting
: text in text; fixing the colour of text is inferior to
: allowing the user to choose it; forcing a user to view
: text in a particular page width or font size is inferior
: to allowing the user to select it. These are technical
: matters and have nothing to do with artistic judgment.

All overridden easily in browsers when you're using CSS on the page.

: Now, in principle, I agree that CSS should make these


: choices available to users, but browsers are not good
: at exposing this functionality

Sorry that's nonsense. IE and Mozilla/Netscape both make it very easy
for me to ignore colours, fonts, use my own style sheet etc.
So that's 96% of web browsers covered then...

: A web designer who constrains my ability to choose the


: appearance of a web site has decreased its usability.
: That's pretty obvious, isn't it?

Yes. And CSS helps prevent this.

Chris

unread,
Sep 3, 2002, 5:10:57 PM9/3/02
to
Mark Goodge wrote:

> Unless you're suggesting that all webpages are better off with no
> attempt at visual design at all (which would be a ludicrous position
> and one I'm certain you don't hold), I can't see any other reason to
> oppose the use of CSS. Or am I missing something that you're taking
> for granted but haven't actually stated yet?

I think you're probably mischaracterising my position on
CSS. I don't oppose its use completely; I just think that
the vast majority of designers using it are not managing
to do anything at all useful with it, and much of what
they are doing is counterproductive.

Nor, I think, would I want to make the blanket statement
that there should be no visual design at all on the web.
But it is not a medium which is conducive to visual
design, and I observe that the most usable sites are
typically those which use the least visual design.

--
Chris Lightfoot, chris at ex dash parrot dot com; http://ex-parrot.com/~chris/

``You can't say that, because it's true.''
(unnamed Russian censor, to Malcom Muggeridge, 1933)

Mark Goodge

unread,
Sep 3, 2002, 5:12:12 PM9/3/02
to
On 03 Sep 2002 21:23:55 +0100 (BST), Gareth Marlow put finger to
keyboard and typed:

>In article <nd5anusc8m4luu2h8...@4ax.com>,


>Mark Goodge <ma...@good-stuff.co.uk> wrote:
>>
>>Unless you're suggesting that all webpages are better off with no
>>attempt at visual design at all (which would be a ludicrous position
>>and one I'm certain you don't hold),
>
>That *is* what he's suggesting. See:
>
><URL: http://ex-parrot.com/~chris/html.html >

That page uses visual design. It uses colour to differentiate links,
it uses font size to indicate headings and it has horizontal rules to
set off sections. The fact that all but one of these use the browser
defaults doesn't make it any less a case of visual design than if the
page used the latest in CSS specifications.

The difference between this kind of near-default-design page and one
which makes use of everything that CSS and bodged HTML can offer is
one of degree rather than kind. Admittedly, the difference between
this and, say, www.msn.com is a huge degree of difference, but they're
still points at opposing ends of the same continuum rather than
inhabiting different concept spaces altogether.

In fact, the nature of the web means that, unlike email or Usenet, it
cannot be an entirely text-only medium - it must include some aspect
of design which is inherent in the visual appearance rather than
merely the words themselves. The concept of hyperlinks requires that
they be visible in some way through the presentation of the page to
the reader, whether by underlining, boldface, a different colour, a
visible mouseover event or some combination of these. A screenshot of
one of the earliest browsers[1] shows that visual design was inherent
from the beginning. Even Lynx uses visual design, as far as it has the
ability (it uses bold text for links and indicates emphasis with
underlines). To suggest that the web should not incororate any visual
design at all is, therefore a totally ludicrous position and one which
no sensible person could possibly hold.

There is, though, a second position which is not so ludicrous. This is
the belief that web authors should not do anything to override the
browser defaults - that all visual design should be that which uses
the settings chosen by the browser manufacturers and (as far as they
have the ability to customise it) the browser users. This position
does have some validity, not least because of the design atrocities
perpetrated by some of those who do choose to override the defaults.
However, I still consider this position to be fundamentally
indefensible, as it ignores the reality of progress in both browser
design and developmnent of standards for visual design.

[1] Tim Berners Lee's WorldWideWeb browser, which can be seen at:
http://www.w3c.org/History/1994/WWW/Journals/CACM/screensnap2_24c.gif

Mark
--
http://www.good-stuff.co.uk
"I'm gonna be there tomorrow"

Chris

unread,
Sep 3, 2002, 5:30:24 PM9/3/02
to
Gareth Marlow wrote:
> In article <al31t0$n6$1...@pegasus.csx.cam.ac.uk>, Chris <ch...@x.invalid> wrote:
>>
>>Now, in principle, I agree that CSS should make these
>>choices available to users, but browsers are not good
>>at exposing this functionality and most web designers
>>don't bother to make their pages work well when the
>>user has chosen to override their design choices.
>
>
> I think that's because most web design at the moment *aren't* doing what's
> being advocated. They're using a bastard hybrid of HTML 3.2 tables-based
> page layout and some of the typographical tweaking of CSS to alter
> line spacing and margins. This is slightly better than 1x1 transparent gif
> layout but only slightly. If you do this kind of thing, pages degrade very
> badly the further away from IE 8.5/Netscape 10 you get.
>
> As soon as you completely separate the presentation from the structure you
> get rid of all of these problems, save for buggy browsers which only half
> implement CSS. Such as NS4, as we've already seen.

OK. This is the `yes, it's lousy now, but come back *next
year* -- it'll be great' argument. It will be pleasing if it
turns out to be correct, but I am not hopeful.

>>(A
>>typical example is the floating menu thingies which
>>are now popular.) I don't suppose that CSS has made
>>the situation much worse, but it's hardly a significant
>>improvement, especially given the holier-than-thou
>>attitude of its adherents.
>
>
> You are confusing proponents with ignorant users. Proponents would say
> that, if you want to do pop-up, cascading menus, do them like this:
>
> <URL: http://www.dhtmlcentral.com/tutorials/tutorials.asp?page=1&id=14 >
> using CSS *properly*. That way, you don't shaft your site if the browser
> doesn't support CSS or it's turned off. I think this *is* a significant
> improvement on the previous situation where these things would be
> implemented in such a way that they would only work on one browser, or
> you'd get 4 different versions of the same code because of nuances in
> browser support.

It's not too bad. But note that it's still a huge amount of
code to implement something which isn't really any more
functional than the original bulleted list with which the
designer starts. That would change if he had a thousand
items, rather than nine, but nevertheless you often see this
trick for tiny numbers of links, and it's not really clear
that this approach would work with a much larger amount of
content.


My point is that a designer should only add visual design
elements when they actually add something to the page or to
the user experience. I accept that by its design CSS ought
to have improved the situation; I see precious little
evidence that it has.


--
Chris Lightfoot, chris at ex dash parrot dot com; http://ex-parrot.com/~chris/

Gentlemen! You can't fight in here! This is the War Room!
(from `Dr. Strangelove')

Chris

unread,
Sep 3, 2002, 5:15:51 PM9/3/02
to
Gareth Marlow wrote:
> In article <nd5anusc8m4luu2h8...@4ax.com>,
> Mark Goodge <ma...@good-stuff.co.uk> wrote:
>
>>Unless you're suggesting that all webpages are better off with no
>>attempt at visual design at all (which would be a ludicrous position
>>and one I'm certain you don't hold),
>
>
> That *is* what he's suggesting. See:
>
> <URL: http://ex-parrot.com/~chris/html.html >

Well, no. That page explains why there is almost no
visual design on my pages. (There is a little; you
have commented already on the link colour, and I
choose whether images are aligned to right or left
-- I think it's necessary to make this choice if
you don't want an inline image, but I may be
misremembering.)

Clearly I believe that there are numerous web pages
which would be improved by the removal of visual
design elements. All? Probably not. The majority?
I'm not sure.

--
Chris Lightfoot, chris at ex dash parrot dot com; http://ex-parrot.com/~chris/

``The Algernons were all from minor public schools.
In the new mood of classlessness, they could plausibly carry on as if they
came from major ones.'' (Clive James, on 1960s Cambridge)

Chris

unread,
Sep 3, 2002, 5:52:58 PM9/3/02
to
j...@durge.org wrote:

> IE and Mozilla/Netscape both make it very easy
> for me to ignore colours, fonts, use my own style sheet etc.

Hmm. This was my understanding too, but I've just spent
some time trying to select a style sheet to use in Mozilla
and Netscape, and I can't find the functionality.

Mozilla allows you to select one of a set of style sheets
specified by the web page from the `View' menu. Netscape
4 doesn't appear to support this functionality. Both
browsers allow you to turn off designer-specified colours
and fonts, and to turn of CSS completely. I can't see any
options for selecting a style sheet to override document
styles.

The closest I could find is the somewhat complex

http://www.oreillynet.com/pub/a/network/2000/06/30/magazine/mozilla_stylesheets.html

which suggests that this is at least possible, but it's
not clear how useful it would be for suppressing bad
design when it appears. In particular, I don't know
whether the user style sheet then appears as an option
under the `View' menu-- the implication is `no', but I
haven't tested it.

It's slightly disheartening that the article comments,
(my emphasis)

You can[...] *even* control how other web sites appear to you.

Aaargh.


Here's another rant on the same theme which also bemoans
the absence of widespread user stylesheet support:

http://www.ee.surrey.ac.uk/Personal/L.Wood/stylist/index.html

--
Chris Lightfoot, chris at ex dash parrot dot com; http://ex-parrot.com/~chris/

``If you can tell good advice from bad advice, you don't need any advice.''

Gareth Marlow

unread,
Sep 3, 2002, 6:10:32 PM9/3/02
to
In article <al3avr$6s4$1...@pegasus.csx.cam.ac.uk>,

Chris <ch...@x.invalid> wrote:
>
>Here's another rant on the same theme which also bemoans
>the absence of widespread user stylesheet support:
>
> http://www.ee.surrey.ac.uk/Personal/L.Wood/stylist/index.html

Which brings us neatly back to where we started, because that page,
talking about CSS, says "This is fine, when they're fully aware of what
they're doing and weigh the consequences of their actions carefully."
where "weigh the consequences of their actions carefully" is hyperlinked
to <URL: http://www.zeldman.com/ > [1] which is the source site of the
original "To Hell with Bad Browsers" link that I posted...

CSS alone isn't going to fix the minds of people who publish web pages
without understanding the issues. CSS *is* the answer for those people who
wish to augment their content with some design, while retaining
accessibility and the ability for users to reject their design choices,
without needing to use loads of javascript browser detection or coding
"text only" and "print version" content.

Gareth

[1] this'd be so much easier if we were posting html to USENET [2] <grin>
[2] yes, I know.

Chris

unread,
Sep 3, 2002, 6:43:59 PM9/3/02
to
Gareth Marlow wrote:
> In article <al3avr$6s4$1...@pegasus.csx.cam.ac.uk>,
> Chris <ch...@x.invalid> wrote:
>
>>Here's another rant on the same theme which also bemoans
>>the absence of widespread user stylesheet support:
>>
>> http://www.ee.surrey.ac.uk/Personal/L.Wood/stylist/index.html
>
>
> Which brings us neatly back to where we started, because that page,
> talking about CSS, says "This is fine, when they're fully aware of what
> they're doing and weigh the consequences of their actions carefully."

Actually, I was more interested in the bit about user style
sheets, which aren't discussed on the other sites....

> [1] this'd be so much easier if we were posting html to USENET [2] <grin>

Well, it's a novel argument, at least....

> [2] yes, I know.

I'm intrigued by the formatting of URLs in < ... > or
<URL: ... > -- why have you chosen that style? What is
its advantage over simply putting whitespace around
them?

--
Chris Lightfoot, chris at ex dash parrot dot com; http://ex-parrot.com/~chris/

The car park is liable to flooding.
If this notice is covered, do not park your car here.
(notice in car park)

Gareth Marlow

unread,
Sep 3, 2002, 7:22:52 PM9/3/02
to
In article <al3dvf$8ua$1...@pegasus.csx.cam.ac.uk>,

Chris <ch...@x.invalid> wrote:
>
>I'm intrigued by the formatting of URLs in < ... > or
><URL: ... > -- why have you chosen that style? What is
>its advantage over simply putting whitespace around
>them?

It's a habit I've tried to get into because of problems caused by
linebreaks/other white space in URLs > 80 characters which are posted or
emailed. The appendix of <URL: http://www.w3.org/Addressing/rfc1738.txt >
refers; basically it says that mail clients and news clients which wish to
automatically infer hyperlinks from messages should ignore white space
between the <URL: > delimeters.

G

j...@durge.org

unread,
Sep 3, 2002, 8:06:04 PM9/3/02
to
Chris <ch...@x.invalid> wrote:
: The closest I could find is the somewhat complex
: http://www.oreillynet.com/pub/a/network/2000/06/30/magazine/mozilla_stylesheets.html

Try this:

http://www.geocrawler.com/archives/3/138/2000/4/0/3666992/

for Mozilla.
The same will no doubt work with Netscape 6 + 7.

I believe that graphical positioning is a sliding scale. You position
things above one another by placing them first in the HTML even if you
avoid all tables and CSS.
At the other end of the scale you get complete messes like www.ft.com
As long as a design is easy to navigate, accessible and consistent there
clearly is no problem.

Even with tables used for layout.

Dan Sheppard

unread,
Sep 3, 2002, 9:23:36 PM9/3/02
to
Chris <ch...@x.invalid> wrote:
>Putting text in images is inferior to putting text in text;
>fixing the colour of text is inferior to allowing the user to choose it;

The following images (from my livejournal) are almost entirely pieces
of writing. However, I think it would have much less impact if they
were expressed as formatted text, with or without CSS, rather than as
images. In my opinion, something like PDF (but done well) is probably
a good compromise. I am aware that the last is next to unreadable.

http://www.chiark.greenend.org.uk/~dans/liverpool.png
http://www.chiark.greenend.org.uk/~dans/bang.png
http://www.chiark.greenend.org.uk/~dans/aeroplane.jpg
http://www.chiark.greenend.org.uk/~dans/irony.jpg
http://www.chiark.greenend.org.uk/~dans/prince.jpg

Dan.
--
http://www.chiark.greenend.org.uk/~dans/

Dan Sheppard

unread,
Sep 4, 2002, 2:32:20 AM9/4/02
to

peej

unread,
Sep 4, 2002, 4:24:22 AM9/4/02
to
"Chris" <ch...@x.invalid> wrote in message
news:al31t0$n6$1...@pegasus.csx.cam.ac.uk...

[... other stuff handled elsewhere I think ...]

> A web designer who constrains my ability to choose the
> appearance of a web site has decreased its usability.
> That's pretty obvious, isn't it?

No.

If I build a web site which is an exhibition of photographs, I don't want
viewers to change either the pictures or the way they're exhibited. Those
things are my choice. It's 100% a subjective issue, it's my responsibility,
and I understand the consequences. In this case lack of user interactivity
reduces the "features" of the site by definition, but I don't think it
reduces the "usability" at all.

It seems strange for you to argue against CSS, which allows authors to
provide you with some input, should they feel this appropriate.

;-) It seems to me that ex-parrot sketch becomes hugely un-funny after a
certain number of repetitions. Are you sure that black text on white with
red links doesn't become just a little bit boring eventually?

Phil


Nick Wagg

unread,
Sep 4, 2002, 4:00:17 AM9/4/02
to
peej wrote:
>
> surely you're not giving in to the hordes of netscape 4 users who can't
> figure out how to upgrade...

Not so. At work we are obliged to stick to an ancient version of
SunOS so have to rely on an equally ancient version of Netscape.
I have tried downloading Opera but have yet to find a version which
will link with our versions of the libs.
--
Nick Wagg
TranscenData Europe Ltd, Oakington House, Oakington, Cambridge CB4 5AF
Email: nick...@transcendata.com URL: www.transcendata.com
Tel: +44 (0)1223 237111 Fax: +44 (0)1223 234192

Jonathan Amery

unread,
Sep 4, 2002, 3:58:21 AM9/4/02
to
In article <al2unn$s0j$1...@pegasus.csx.cam.ac.uk>,

Chris <ch...@x.invalid> wrote:
>Really? I thought it was the Microsoft Windows screen
>resolution problem.

At least part of the problem is that webdevs tend to use pt rather
than px, which is silly. Not that I care - since my screen resolution
is actually higher (in px/in) than Windows' (in pt/in).

--
Jonathan Amery. Well, I could dance with you honey, if you think its funny,
##### Does your mother know that you're out?
#######__o And I could chat with you baby, flirt a little maybe,
#######'/ Does your mother know that you're out? - ABBA.

Rupert Moss-Eccardt

unread,
Sep 4, 2002, 4:38:08 AM9/4/02
to
Nick Wagg wrote:
>
> peej wrote:
> >
> > surely you're not giving in to the hordes of netscape 4 users who can't
> > figure out how to upgrade...
>
> Not so. At work we are obliged to stick to an ancient version of
> SunOS so have to rely on an equally ancient version of Netscape.
> I have tried downloading Opera but have yet to find a version which
> will link with our versions of the libs.

If you are using 2.7 then why not install Internet Explorer?
I think there is an older version for even older versions of Solaris.

Steven Kitson

unread,
Sep 4, 2002, 4:35:49 AM9/4/02
to
Chris <ch...@x.invalid> wrote:
>Nor, I think, would I want to make the blanket statement
>that there should be no visual design at all on the web.
>But it is not a medium which is conducive to visual
>design,

True.

Is there any chance of some standard which actually allows proper web
design? Something like writing web pages in PDF?
--
They say his relatives are chimps and apes

Gareth Marlow

unread,
Sep 4, 2002, 4:53:19 AM9/4/02
to
In article <v-C*5f...@news.chiark.greenend.org.uk>,

Steven Kitson <ski...@chiark.greenend.org.uk> wrote:
>
>Is there any chance of some standard which actually allows proper web
>design? Something like writing web pages in PDF?

What do you mean by proper web design? I believe that the existing
standards; HTML 4.01, XHTML 1 and CSS 2, allow the structure of underlying
content to be represented in a rigorous way, and for this content then to
be formatted in an aesthetically-pleasing manner, while still keeping it
capable of being rendered in a meaningful way on a whole load of devices
and in ways not intended by the original author.

Furthermore, the user has a high degree of control over the presentation.
This can happen either automatically (using a text-only browser; a braille
display; a text-to-speech reader) or manually (over-riding style sheets
with large-print, high-contrast presentation).

Someone who is writing pages following the spirit and letter of these
standards is going to produce this kind of output. That's what I call
proper web design. It's not something I do at the moment, although I hope
to with any new material I produce.

Of course, there's nothing in the standards to impart an intelligent
appreciation of their intent to web designers. So it's still possible for
someone to write a page with all content rendered as a single GIF file (or
worse, 20 GIF files, sliced using Fireworks and then re-assembled using
tables). Just as it's possible for someone to do the same thing in PDF.

Gareth


--
http://www.chiark.greenend.org.uk/~garethm/ : Gareth Marlow
______________________________________________________________________

" Will you be dinging and donging merrily on high tonight sir?"

Steven Kitson

unread,
Sep 4, 2002, 4:57:42 AM9/4/02
to
Gareth Marlow <gar...@chiark.greenend.org.uk> wrote:
>Steven Kitson <ski...@chiark.greenend.org.uk> wrote:
>>Is there any chance of some standard which actually allows proper web
>>design? Something like writing web pages in PDF?
>
>What do you mean by proper web design?

Design, by which I mean designing the layout of pages so that they appear
in the same layout on all viewers, which works on the web.

Gareth Marlow

unread,
Sep 4, 2002, 5:20:32 AM9/4/02
to
In article <YLp*bl...@news.chiark.greenend.org.uk>,

Steven Kitson <ski...@chiark.greenend.org.uk> wrote:
>
>Design, by which I mean designing the layout of pages so that they appear
>in the same layout on all viewers, which works on the web.

No. I think it's chasing this end which has got us into the mess which CSS
is trying to sort out. You can only achieve things like this by rendering
the page as a single image, using transparent 1x1 GIFs to nudge your
content across the screen, etc. As you know, you end up with something
which looks perfect in one browser and at one screen resolution, and it
either looks broken in other browsers, or you spend all of your time
building in tweaks and work-arounds so it will work in them.

I can think of very few situations where you need the layout of pages to
appear exactly the same on all viewers. They're usually where the
aesthetics are more important than the textual content. Some examples are
those images which Dan posted. I would do what he did under the same
circumstances.

j...@durge.org

unread,
Sep 4, 2002, 5:37:02 AM9/4/02
to
Steven Kitson <ski...@chiark.greenend.org.uk> wrote:
: Is there any chance of some standard which actually allows proper web

: design? Something like writing web pages in PDF?

Scalable Vector Graphics?
http://www.w3.org/TR/SVG/

Not quite the same I know.

Steven Kitson

unread,
Sep 4, 2002, 5:27:22 AM9/4/02
to
Gareth Marlow <gar...@chiark.greenend.org.uk> wrote:
>Steven Kitson <ski...@chiark.greenend.org.uk> wrote:
>>Design, by which I mean designing the layout of pages so that they appear
>>in the same layout on all viewers, which works on the web.
>
>No. I think it's chasing this end which has got us into the mess which CSS
>is trying to sort out. You can only achieve things like this by rendering
>the page as a single image, using transparent 1x1 GIFs to nudge your
>content across the screen, etc. As you know, you end up with something
>which looks perfect in one browser and at one screen resolution, and it
>either looks broken in other browsers, or you spend all of your time
>building in tweaks and work-arounds so it will work in them.

It's chasing it through the wrong mechanism which has got us into the
mess. HTML, with whatever extensions, doesn't do it, and needs the
single image/gifs/etc to try to do it and then it all falls apart anyway.

But that's just because the implementation is wrong for the purpose. What
HTML can't achieve, PDF seems to do just fine. So I'm wondering if
something PDF-like could be adapted for the web.

Richard Meredith

unread,
Sep 4, 2002, 5:37:00 AM9/4/02
to
In article <YLp*bl...@news.chiark.greenend.org.uk>,
ski...@chiark.greenend.org.uk (Steven Kitson) wrote:

> Gareth Marlow <gar...@chiark.greenend.org.uk> wrote:
> >Steven Kitson <ski...@chiark.greenend.org.uk> wrote:
> >>Is there any chance of some standard which actually allows proper web
> >>design? Something like writing web pages in PDF?
> >
> >What do you mean by proper web design?
>
> Design, by which I mean designing the layout of pages so that they appear
> in the same layout on all viewers, which works on the web.

That's contrary to the principle that not all viewers are equally capable:
if you need that, you're reduced to least common denominator design.

PDAs, high resolution CRT/LCD panels, TV sets and character terminals
/can't/ display the same content identically.

--
This message may contain traces of nuts. Do not refreeze once thawed.
No animals were hurt in the making of this production. Suitable for
vegetarians.


Mark Goodge

unread,
Sep 4, 2002, 5:44:29 AM9/4/02
to
Steven Kitson wrote:
>
> Gareth Marlow <gar...@chiark.greenend.org.uk> wrote:
> >Steven Kitson <ski...@chiark.greenend.org.uk> wrote:
> >>Is there any chance of some standard which actually allows proper web
> >>design? Something like writing web pages in PDF?
> >
> >What do you mean by proper web design?
>
> Design, by which I mean designing the layout of pages so that they appear
> in the same layout on all viewers, which works on the web.

No. Because you don't have any control over things like the screen size
and resolution, the gamma correction (which will significantly affect
the user's perception of your colours), etc. That's before you even take
account of different operating systems, people browsing via a shell
login, etc.

You can get something approaching it using PDF, and CSS offers a way of
achieving similar control in a web page, but the idea of making
something the same in *all* viewers is a complete non-starter. The
nearest you can manage is to make it look pretty much the same in a
subset of browsers that covers a large enough group of users. Even so,
you'd find it difficult to prevent people looking at it with software
that will render it differently - there are very few pages out there
that Lynx can't make at least readable, if not pretty.

Mark

Steven Kitson

unread,
Sep 4, 2002, 5:47:54 AM9/4/02
to
Richard Meredith <rmer...@cix.co.uk> wrote:

>ski...@chiark.greenend.org.uk (Steven Kitson) wrote:
>> Design, by which I mean designing the layout of pages so that they appear
>> in the same layout on all viewers, which works on the web.
>That's contrary to the principle that not all viewers are equally capable:
>if you need that, you're reduced to least common denominator design.

On the contrary, it's 'least common denominator design' which tells us
that the web has to cope with mobile phones and character terminals.

>PDAs, high resolution CRT/LCD panels, TV sets and character terminals
>/can't/ display the same content identically.

So don't give them the same content. Better to specifically design for the
receiver (as something like WML does) than try to provide a
one-size-fits-all.

Steven Kitson

unread,
Sep 4, 2002, 5:49:12 AM9/4/02
to
Mark Goodge <ma...@good-stuff.co.uk> wrote:

>Steven Kitson wrote:
>> Design, by which I mean designing the layout of pages so that they appear
>> in the same layout on all viewers, which works on the web.
> The
>nearest you can manage is to make it look pretty much the same in a
>subset of browsers that covers a large enough group of users. Even so,
>you'd find it difficult to prevent people looking at it with software
>that will render it differently - there are very few pages out there
>that Lynx can't make at least readable, if not pretty.

If people are going to be preverse, that's their problem.