By using this site, you agree to our updated Privacy Policy and our Terms of Use. Manage your Cookies Settings.
424,689 Members | 1,759 Online
Bytes IT Community
+ Ask a Question
Need help? Post your question and get tips & solutions from a community of 424,689 IT Pros & Developers. It's quick & easy.

Jakob Nielsen says: Change the color of visited links

P: n/a
Jakob Nielsen in his most recent Alertbox
(http://www.useit.com/alertbox/20040503.html) tells web authors to
change the color of visited links.

I agree to his purpose: Help users understand where they have been. I
also agree to the background: Too many web authors keep uniform link
colors or their pages, thereby confusing the users.

What I have a hard time agreeing to is his conclusion that web authors
should deliberately change the color of already visited links.

I do not recall having seen a GUI based browser that does not change
the color (or at least the appearance) of visited links by default. In
the browsers that I am currently using (IE6, Moz1.4, Opera7.2),
color/appearance of visited links can even be controlled by the user.

So my question is: Why should I as a web author control the color of
visited links when most browsers do so by default and when users in
most cases are allowed to set their own preferences?

--
Jan Tuxen
Remove five-three-o-o in my address if you want to
e-mail directly to me.
Jul 20 '05 #1
Share this Question
Share on Google+
18 Replies


P: n/a
On Tue, 04 May 2004 15:48:16 +0200, Jan Tuxen <ja*******@get2net.dk5300>
wrote:
Jakob Nielsen in his most recent Alertbox
(http://www.useit.com/alertbox/20040503.html) tells web authors to
change the color of visited links.
...
So my question is: Why should I as a web author control the color of
visited links when most browsers do so by default and when users in
most cases are allowed to set their own preferences?


You should make suggestions. One good reason is if your background color
would not provide sufficient contrast for the link.

Fact is, I can override every style suggestion you make. So why do
anything? Because they are suggestions, and *most* users will use them.
Jul 20 '05 #2

P: n/a
On Tue, 04 May 2004 15:48:16 +0200, Jan Tuxen wrote:
So my question is: Why should I as a web author control the color of
visited links when most browsers do so by default and when users in
most cases are allowed to set their own preferences?


To account for 'the few'. Leave as little
to chance as is practicable.

--
Andrew Thompson
http://www.PhySci.org/ Open-source software suite
http://www.PhySci.org/codes/ Web & IT Help
http://www.1point1C.org/ Science & Technology
Jul 20 '05 #3

P: n/a
Els
Jan Tuxen wrote:
Jakob Nielsen in his most recent Alertbox
(http://www.useit.com/alertbox/20040503.html) tells web authors to
change the color of visited links.

I agree to his purpose: Help users understand where they have been. I
also agree to the background: Too many web authors keep uniform link
colors or their pages, thereby confusing the users.

What I have a hard time agreeing to is his conclusion that web authors
should deliberately change the color of already visited links.
I might have misread Nielsen, but I don't think he said
anything about deliberately, other than making sure the
colors are different. If web authors would leave the colours
alone, most browsers would show difference in colour between
visited and non-visited links.

Personally I think Nielsen meant to say: make sure you don't
make the visited link colour equal to the non-visited link
colour.
Which results indeed, in deliberately changing the colour of
visited links if you changed the colour of any links, I suppose?
I do not recall having seen a GUI based browser that does not change
the color (or at least the appearance) of visited links by default. In
the browsers that I am currently using (IE6, Moz1.4, Opera7.2),
color/appearance of visited links can even be controlled by the user. So my question is: Why should I as a web author control the color of
visited links when most browsers do so by default and when users in
most cases are allowed to set their own preferences?


Because authors happen to change the default colors of the
links, and most users don't bother to set their own
preferences? And in some cases, users aren't allowed to
change the settings of their browser (i.e. in libraries,
internet-cafes)

--
Els
http://locusmeus.com/
Sonhos vem. Sonhos vo. O resto imperfeito.
- Renato Russo -

Jul 20 '05 #4

P: n/a
"Jan Tuxen" <ja*******@get2net.dk5300> wrote in
comp.infosystems.www.authoring.stylesheets:
What I have a hard time agreeing to is his conclusion that web authors
should deliberately change the color of already visited links.

So my question is: Why should I as a web author control the color of
visited links when most browsers do so by default and when users in
most cases are allowed to set their own preferences?


He points out that most browsers do it by default. The problem is
that some Web authors -- thoughtless Web authors -- tend to style
all links the same. Nielsen is saying that _if_ you assign colors to
links you should assign different colors to visited and unvisited
links.

Your other option is not to assign colors to links -- but that means
not assigning any colors at all, and few people have that much self
restraint.

--
Stan Brown, Oak Road Systems, Cortland County, New York, USA
http://OakRoadSystems.com/
HTML 4.01 spec: http://www.w3.org/TR/html401/
validator: http://validator.w3.org/
CSS 2 spec: http://www.w3.org/TR/REC-CSS2/
2.1 changes: http://www.w3.org/TR/CSS21/changes.html
validator: http://jigsaw.w3.org/css-validator/
Jul 20 '05 #5

P: n/a
On Tue, 4 May 2004 10:53:53 -0400, Stan Brown <th************@fastmail.fm>
wrote:
Your other option is not to assign colors to links -- but that means
not assigning any colors at all, and few people have that much self
restraint.


I wonder which is ultimately better - changing link colors to give good
contrast to your bg, or leaving link colors as the user is used to seeing
them in the browser and choosing bg colors which afford contrast.

In either case, I think links should generally be blue and visited links
purply-red (as most users see it that way, although not all), but if I
have a background that won't contrast well I have no problem tweaking that
so it's, say, a lighter blue on a darker background. But it's smartest to
let the links be as close to the user's preference as possible.
Jul 20 '05 #6

P: n/a
Neal wrote:
But it's
smartest to let the links be as close to the user's preference as possible.


What is as close to the user's preference as possible? You don't know
the user's preference. Or do you mean browser's default settings?
--
Johannes Koch
In te domine speravi; non confundar in aeternum.
(Te Deum, 4th cent.)
Jul 20 '05 #7

P: n/a
On Tue, 04 May 2004 17:16:23 +0200, Johannes Koch <ko**@w3development.de>
wrote:
Neal wrote:
But it's smartest to let the links be as close to the user's preference
as possible.


What is as close to the user's preference as possible? You don't know
the user's preference. Or do you mean browser's default settings?

I did word that poorly, but I see my meaning was still clear. Yes, you
have divined my meaning :)
Jul 20 '05 #8

P: n/a
DU
Jan Tuxen wrote:
Jakob Nielsen in his most recent Alertbox
(http://www.useit.com/alertbox/20040503.html) tells web authors to
change the color of visited links.

I agree to his purpose: Help users understand where they have been. I
also agree to the background: Too many web authors keep uniform link
colors or their pages, thereby confusing the users.

What I have a hard time agreeing to is his conclusion that web authors
should deliberately change the color of already visited links.

I do not recall having seen a GUI based browser that does not change
the color (or at least the appearance) of visited links by default. In
the browsers that I am currently using (IE6, Moz1.4, Opera7.2),
color/appearance of visited links can even be controlled by the user.

So my question is: Why should I as a web author control the color of
visited links when most browsers do so by default and when users in
most cases are allowed to set their own preferences?


"8. Non-Standard Link Colors
Links to pages that have not been seen by the user are blue; links to
previously seen pages are purple or red. Don't mess with these colors
since the ability to understand what links have been followed is one of
the few navigational aides that is standard in most web browsers.
Consistency is key to teaching users what the link colors mean."
Jakob Nielsen
http://www.useit.com/alertbox/9605.html

If you set
a:link {color:blue;}
a:visited {color:visited;}
and choose a contrasting background color in your page, then I would say
you are following both Nielsen's recommendations.

Blue and purple are default colors in MSIE and Netscape for both
(respective) links. There is no need to change this blue/purple combo
unless your webpage background is too close to blue.

DU
Jul 20 '05 #9

P: n/a
DU
Els wrote:
Jan Tuxen wrote:
Jakob Nielsen in his most recent Alertbox
(http://www.useit.com/alertbox/20040503.html) tells web authors to
change the color of visited links.
I agree to his purpose: Help users understand where they have been. I
also agree to the background: Too many web authors keep uniform link
colors or their pages, thereby confusing the users.

What I have a hard time agreeing to is his conclusion that web authors
should deliberately change the color of already visited links.

I might have misread Nielsen, but I don't think he said anything about
deliberately, other than making sure the colors are different. If web
authors would leave the colours alone, most browsers would show
difference in colour between visited and non-visited links.

Personally I think Nielsen meant to say: make sure you don't make the
visited link colour equal to the non-visited link colour.
Which results indeed, in deliberately changing the colour of visited
links if you changed the colour of any links, I suppose?


This is how I also understood his column.
I do not recall having seen a GUI based browser that does not change
the color (or at least the appearance) of visited links by default. In
the browsers that I am currently using (IE6, Moz1.4, Opera7.2),
color/appearance of visited links can even be controlled by the user.


So my question is: Why should I as a web author control the color of
visited links when most browsers do so by default and when users in
most cases are allowed to set their own preferences?

Because authors happen to change the default colors of the links, and
most users don't bother to set their own preferences? And in some cases,
users aren't allowed to change the settings of their browser (i.e. in
libraries, internet-cafes)


Good point. I entirely agree.

DU
Jul 20 '05 #10

P: n/a
On Tue, 04 May 2004 18:44:13 -0400, DU <dr*******@hotWIPETHISmail.com>
wrote:
If you set
a:link {color:blue;}
a:visited {color:visited;}
then you will discover a new color! ;)
and choose a contrasting background color in your page, then I would say
you are following both Nielsen's recommendations.

Blue and purple are default colors in MSIE and Netscape for both
(respective) links. There is no need to change this blue/purple combo
unless your webpage background is too close to blue.

Jul 20 '05 #11

P: n/a
On Tue, 4 May 2004, DU wrote:
If you set
a:link {color:blue;}
a:visited {color:visited;}
Standard rule: never specify an explicit text colour without
specifying an explicit background colour, and vice versa, at -every-
level of specificity that you use. To say it again: either specify
both, or specify neither.

The W3C CSS checker/validator has a warning for this situation. I'd
say they are right!
and choose a contrasting background color in your page, then I would say
you are following both Nielsen's recommendations.
specifying it on "the page" isn't good enough, since you have no idea
how it's going to cascade with other stylesheet(s) - specifically,
with the user's stylesheet if they apply one.
Blue and purple are default colors in MSIE and Netscape for both
(respective) links.
Right, but we don't know whether the user accepts that. Maybe they
set links to be yellow against a blue background - for whatever reason
- and now you're going to set their link text to blue while leaving
their own choice of background.
There is no need to change this blue/purple combo
unless your webpage background is too close to blue.


Unfortunately, logic says that there is. If you follow the guidelines
(and they seem to me to have an inevitable logic behind them), then
either you leave both colours alone, or you specify both colours. If
you leave them alone, you don't need to know what they are. If you
specify them both, then you have no idea how close they are to the
user's choice of default.

If you do anything other than what that guideline says, then the best
you can claim is "it'll work, for most users, most of the time".
Which may be good enough for some - need I say more?

cheers

Jul 20 '05 #12

P: n/a
*Alan J. Flavell* <fl*****@ph.gla.ac.uk>:
a:link {color:blue;}
a:visited {color:visited;}
Standard rule: never specify an explicit text colour without
specifying an explicit background colour, and vice versa, at -every-
level of specificity that you use.


That doesn't mean

a:link {color:blue; background-color: white;}
a:visited {color: purple; background-color: white;}

was good practice, though.

a[href] {background-color: white;}
a:link {color:blue;}
a:visited {color: purple;}

would be, if attribute selectors, which have the same specifity as
pseudo-classe, would be better supported. For now

a:link, a:visited {background-color: white;}
a:link {color:blue;}
a:visited {color: purple;}

is an working approach where the specifity for each selector is (0, 0, 1, 1)
for CSS 2.1 or (0, 1, 1) for CSS 2.0.
The W3C CSS checker/validator has a warning for this situation. I'd
say they are right!


That lead to thousands of CSS authors writing

background-color: transparent;

just to get rid of that warning.

I don't think that advice is all that important, though. An author should
care for all colors from his author CSSs to work well with each other, e.g.
same specifity for foreground and background. So should do the user for his
user CSSs.
IOW: If all block elements have a white background and black text in the
author CSS, it's not necessary to specify 'background-color' for each inline
element.

--
The Hitchhiker's Guide to the Galaxy:
"He was a dreamer, a thinker, a speculative philosopher,
or, as his wife would have it, an idiot."
Jul 20 '05 #13

P: n/a
On Thu, 6 May 2004, Christoph Paeper wrote:
*Alan J. Flavell* <fl*****@ph.gla.ac.uk>:
a:link {color:blue;}
a:visited {color:visited;}
Standard rule: never specify an explicit text colour without
specifying an explicit background colour, and vice versa, at -every-
level of specificity that you use.


That doesn't mean

a:link {color:blue; background-color: white;}
a:visited {color: purple; background-color: white;}

was good practice, though.


Could you explain why not?
a[href] {background-color: white;}
a:link {color:blue;}
a:visited {color: purple;}

would be,
If one cascaded that with a stylesheet which set a[href] without
setting a:link and/or a:visited, what would happen?
The W3C CSS checker/validator has a warning for this situation. I'd
say they are right!


That lead to thousands of CSS authors writing

background-color: transparent;

just to get rid of that warning.


I know. That's why I proposed to the authors of the W3 CSS checker
that they should produce a warning for this situation also, and I
thought they had agreed it was a reasonable idea. I didn't see it
happen, though.

ISTM that the only reasonable use of "transparent" is in conjunction
with specifying that the text colour is inherited. Doing this
explicitly would have different cascading properties than doing
nothing, so it isn't as redundant as it at first might seem.
I don't think that advice is all that important, though.
Statistically not; but logically, there will occasionally be
situations where the cascade results in text which is the same colour
as its background. If the reader does not know the text is there,
s/he will have no idea that it would be useful to take some overriding
action (e.g to disable CSS or to switch off the author stylesheet). If
everyone follows the guidelines, on the other hand, this occasional
risk is excluded.
An author should care for all colors from his author CSSs to work
well with each other,


Yes, sure; but the author can have no idea how it will cascade with
the user stylesheet, and with (where applicable) next week's corporate
stylesheet from the latest re-branding consultants, and so on...
Jul 20 '05 #14

P: n/a
In message <c7**********@news.eusc.inter.net>, DU
<dr*******@hotWIPETHISmail.com> writes
"8. Non-Standard Link Colors
Links to pages that have not been seen by the user are blue; links to
previously seen pages are purple or red. Don't mess with these colors [...]"
Jakob Nielsen
http://www.useit.com/alertbox/9605.html


And if the background is blue, purple or red?
--
Andy Mabbett
"The Internet is a reflection of our society[ ...]. If we do not like what we
see in that mirror the problem is not to fix the mirror, we have to fix
society." Vint Cerf
Jul 20 '05 #15

P: n/a
Andy Mabbett <us**********@pigsonthewing.org.uk> wrote:
And if the background is blue, purple or red?


Then the choice of background is wrong. Isn't this obvious? The need to
comply with conventional link colors (using values that correspond to
browser defaults or are sufficiently close to them to make it evident
enough what is text, what is an unvisited link, and what is a visited
link, at a glance) implies that the overall color scheme must use a
fairly light background color and a dark text color (typically, black,
though something close to black or e.g. dark brown would normally do,
too).

Attempts to use dark background colors lead to a multitude of serious
problems, not just with link colors but with text legibility too.
There's no really satisfactory font generally available for white on
black text, and then there's the character and word spacing, and line
spacing. If you manage to tune the appearance so that white on black, or
generally light on dark, is tolerably readable, then it will appear as
fairly strange when your color suggestions are overridden. And in CSS,
there is no way to say "if the browser uses white on black, then ...".

Do I need to say why red is a particularly poor choice for background?

--
Yucca, http://www.cs.tut.fi/~jkorpela/
Jul 20 '05 #16

P: n/a
"Alan J. Flavell" <fl*****@ph.gla.ac.uk> wrote in
comp.infosystems.www.authoring.stylesheets:
ISTM that the only reasonable use of "transparent" is in conjunction
with specifying that the text colour is inherited.


Thank you for this reminder.

One of my "get it it some day" tasks has been to change my stylings
of <a> to background-color transparent. Right now my style sheet
http://oakroadsystems.com/screen.css
has about five sets of properties for a:link, a:visited, a:hover,
a:focus, a:active, a[name] because those can occur in different
contexts with different background colors. The five differ only in
background-color, so I was planning to replace them with _one_ set
with background-color:transparent.

But you reminded me that a user style sheet might set text color and
text background-color without setting link colors or backgrounds, in
which case my link colors could clash with (or be invisible on) the
user's background.

Sigh. "This redistribution of wealth is trickier than I thought."

--
Stan Brown, Oak Road Systems, Cortland County, New York, USA
http://OakRoadSystems.com/
HTML 4.01 spec: http://www.w3.org/TR/html401/
validator: http://validator.w3.org/
CSS 2 spec: http://www.w3.org/TR/REC-CSS2/
2.1 changes: http://www.w3.org/TR/CSS21/changes.html
validator: http://jigsaw.w3.org/css-validator/
Jul 20 '05 #17

P: n/a
In message <Xn*****************************@193.229.0.31>, Jukka K.
Korpela <jk******@cs.tut.fi> writes
And if the background is blue, purple or red?


Then the choice of background is wrong. Isn't this obvious? The need to
comply with conventional link colors (using values that correspond to
browser defaults or are sufficiently close to them to make it evident
enough what is text, what is an unvisited link, and what is a visited
link, at a glance) implies that the overall color scheme must use a
fairly light background color and a dark text color (typically, black,
though something close to black or e.g. dark brown would normally do, too).

Attempts to use dark background colors

[...]

Where did I say anything about /dark/ background colours?
--
Andy Mabbett
"The Internet is a reflection of our society[ ...]. If we do not like what we
see in that mirror the problem is not to fix the mirror, we have to fix
society." Vint Cerf
Jul 20 '05 #18

P: n/a
Andy Mabbett <us**********@pigsonthewing.org.uk> wrote:
Where did I say anything about /dark/ background colours?


In the sentence "if the background is blue, purple or red?". They are all
rather dark colours, for a background, especially since you would make
link colours deviate quite a bit from the common defaults to make the
legible against a blue, purple or red background.

--
Yucca, http://www.cs.tut.fi/~jkorpela/
Jul 20 '05 #19

This discussion thread is closed

Replies have been disabled for this discussion.