Comparing stylesheets

[David]

I've got a 50k main.css file that's referenced by a load of sites.
Each of these sites also has a site.css file that modifies certain
styles defined in main.css. Changing the colour of borders, boxes etc,
as well as defining certain site specific styles. site.css is small
and easy to understand, and the only bit that can be edited by the
masses.

That's the theory. The practice is that there are now 4 sites going
live
in a week where some muppet who claims to understand css has just
copied
main.css in chunks, renamed it site.css, tweaked some bits and added
more, without deleting the inherited styles (*nice full explanation
below). It all works fine, but it means each site's site.css is now
about 60k, when it's meant to be nice and neat and about 10/20k.

So, to the question:
What tools do people out there use that could help me? Is there any
tool that can compare declarations across multiple stylesheets to
prevent me from sitting down and laboriously going through every style
declaration in site.css to check if it's redundant or not? 4 times
over? I've had a look at Topstyle and it doesn't seem to perform this
function, which I find surprising. Am I being stupid?

diff/sed/awk etc aren't options, btw...(I ain't a unix god)


*Crystal clear explanation:
in main.css there is this statement:

A.backtotop:link, A.backtotop:visited {
float:right;
margin-top:5px;
margin-right:10px;
margin-left:7px;
padding-left:10px;
font-size:70%;
color:#fff;
background:url(arrow_white_up.gif) 0 0.6em no-repeat;
}
A.backtotop:hover {color:#fff;}
In site.css there is this statement:

A.backtotop:link, A.backtotop:visited {
float:right;
margin-top:5px;
margin-right:0;
margin-left:7px;
padding-left:10px;
font-size:70%;
text-decoration:none;
color:#000;
background:url(arrow_black_up.gif) 0 0.6em no-repeat;
}
A.backtotop:hover {text-decoration:underline; }
site.css doesn't need those duplicate declarations, so that the
declaration in site.css *should* read:

A.backtotop:link, A.backtotop:visited {
margin-right:0;
text-decoration:none;
color:#000;
background:url(arrow_black_up.gif) 0 0.6em no-repeat;
}
A.backtotop:hover {text-decoration:underline; }
But, like, there's 240k of this stuff...

[C A Upsdell]

"David" <da********@yahoo.co.uk> wrote in message
news:99**************************@posting.google.c om...

I've got a 50k main.css file that's referenced by a load of sites.
Each of these sites also has a site.css file that modifies certain
styles defined in main.css. Changing the colour of borders, boxes etc,
as well as defining certain site specific styles. site.css is small
and easy to understand, and the only bit that can be edited by the
masses.

That's the theory. The practice is that there are now 4 sites going
live
in a week where some muppet who claims to understand css has just
copied
main.css in chunks, renamed it site.css, tweaked some bits and added
more, without deleting the inherited styles (*nice full explanation
below). It all works fine, but it means each site's site.css is now
about 60k, when it's meant to be nice and neat and about 10/20k.

So, to the question:
What tools do people out there use that could help me? Is there any
tool that can compare declarations across multiple stylesheets to
prevent me from sitting down and laboriously going through every style
declaration in site.css to check if it's redundant or not? 4 times
over? I've had a look at Topstyle and it doesn't seem to perform this
function, which I find surprising. Am I being stupid?

If you made a backup before the muppet showed up, restore the files from the
backup. Otherwise ...

.... if you have Windows, you can try to recover the old files from the
Recycle Bin.

[Lauri Raittila]

David wrote;

I've got a 50k main.css file that's referenced by a load of sites.
Too heavy.
below). It all works fine, but it means each site's site.css is now
about 60k, when it's meant to be nice and neat and about 10/20k.
Worse
So, to the question:
What tools do people out there use that could help me? Is there any
tool that can compare declarations across multiple stylesheets to
prevent me from sitting down and laboriously going through every style
declaration in site.css to check if it's redundant or not? 4 times
over? I've had a look at Topstyle and it doesn't seem to perform this
function, which I find surprising. Am I being stupid?

diff/sed/awk etc aren't options, btw...(I ain't a unix god)
Too bad... Perl?... If you can't find any tool, it might made sence to
pay someone to do such tool, so next time such exsist. If I ahd this
problem, I would certainly make a tool... Someone with knowledge of some
CSS parser would make better tool.
in main.css there is this statement:

A.backtotop:link, A.backtotop:visited {
float:right;
margin-top:5px;
margin-right:10px;
margin-left:7px;
padding-left:10px;
font-size:70%;
color:#fff;
background:url(arrow_white_up.gif) 0 0.6em no-repeat;
}
A.backtotop:hover {color:#fff;}
Easy solution: Back to top links are not good idea, so delete both and
markup related...
But, like, there's 240k of this stuff...

Hm, If I got you right, you have 4 site.css, and you need to get common
stuff out of them to main.css? Or do you have some specific reason to
just remove stuff in main.css from site.csses?

--
Lauri Raittila <http://www.iki.fi/lr> <http://www.iki.fi/zwak/fonts>

[Harlan Messinger]

"Lauri Raittila" <la***@raittila.cjb.net> wrote in message
news:MP************************@news.individual.ne t...

Easy solution: Back to top links are not good idea, so delete both and
markup related...

That's the first time I've seen that opinion. Why aren't they a good idea?

[David]

Lauri Raittila <la***@raittila.cjb.net> wrote in message news:<MP************************@news.individual.n et>...

I've got a 50k main.css file that's referenced by a load of sites.
Too heavy.

Yeah. You're right. It's big. Lots of commonality across sites...we're
actually going to split it up a bit so we've got a

main.css
sitegroup.css
site.css

....sort of layout. Should reduce main to about 30k or so.

Too bad... Perl?... If you can't find any tool, it might made sence to
pay someone to do such tool, so next time such exsist. If I ahd this
problem, I would certainly make a tool... Someone with knowledge of some
CSS parser would make better tool.

I am getting the horrible sinking feeling that it's going to be a perl
jobby. Not *too* difficult (for someone other than me), but I was
really hoping TopStyle had this functionality - I'm quite surprised it
doesn't, given the nature of CSS it would seem natural that at some
point you'd be importing multiple stylesheets into a document and have
the possibility of duplicate styles occurring. Be a dead useful
feature. For me. Right now. Ho hum.

Easy solution: Back to top links are not good idea, so delete both and
markup related...

Interesting. Perhaps not the forum for it (I imagine we'd get justly
flamed for discussing non-css stuff), but if you want to take this off
list I'd be interested to know your reasons for saying back to top
links aren't a good idea...it's something I've never heard before...

Hm, If I got you right, you have 4 site.css, and you need to get common
stuff out of them to main.css? Or do you have some specific reason to
just remove stuff in main.css from site.csses?

Nope. I need to remove declarations in the site.css files that are
identical to declarations in main.css...

[David]

"C A Upsdell" <cupsdell0311XXX@-@[e-mail]> wrote in message news:<hu********************@rogers.com>...

If you made a backup before the muppet showed up, restore the files from the
backup. Otherwise ...

... if you have Windows, you can try to recover the old files from the
Recycle Bin.

Oh, if only it were that simple :)

The muppet has (to be honest) done a pretty extensive amount of OK
work on the site.css files, it's just the inclusion of redundant stuff
that's the problem.

[Mark Tranchant]

David wrote:

Lauri Raittila <la***@raittila.cjb.net> wrote in message news:<MP************************@news.individual.n et>...

Easy solution: Back to top links are not good idea, so delete both and
markup related...

Interesting. Perhaps not the forum for it (I imagine we'd get justly
flamed for discussing non-css stuff), but if you want to take this off
list I'd be interested to know your reasons for saying back to top
links aren't a good idea...it's something I've never heard before...

It's a feature. We don't like features round here. ;-)

--
Mark.
http://tranchant.plus.com/

[Daniel Déchelotte]

"Harlan Messinger" <h.*********@comcast.net> a écrit :

| "Lauri Raittila" <la***@raittila.cjb.net> wrote in message:
|
| > Easy solution: Back to top links are not good idea, so delete both and
| > markup related...
|
| That's the first time I've seen that opinion. Why aren't they a good
| idea?

As a user, have you ever used one of them? Have you ever _needed_ one of
them?

--
Daniel Déchelotte
http://yo.dan.free.fr/

[Alan J. Flavell]

On Tue, 19 Oct 2004, Daniel [ISO-8859-15] Déchelotte wrote:

| "Lauri Raittila" <la***@raittila.cjb.net> wrote in message:
|
| > Easy solution: Back to top links are not good idea, so delete both and
| > markup related...
[...]
As a user, have you ever used one of them? Have you ever _needed_
one of them?

No; but there's some kind of theory (I'd tend rather to think of it as
"superstition", I'm afraid) that they are useful for accessibility.

However, I don't know a browser that doesn't have several different
ways of its own to get to the head of the page[1], so I reckon they're
just a confusing waste of good markup. Those disabled folks that I've
met have shown much better abilities at using the browser's own
built-in tools on every page (rather than letting themselves be misled
by page-specific or site-specific ersatz features) - as compared to
some ordinary folks whose only "disability" was shortage of a clue.

cheers

[1] Notice that I didn't say "back to the head of the page", because a
reader who went to a fragment URL (wibble#whatever) might never have
seen the head of the page before - hence, it can be illogical to speak
about going "back" to it.

[Harlan Messinger]

"Daniel Déchelotte" <ma**********@fr.club-internet.invalid> wrote in message
news:20041019172426.6d459e1e.ma**********@fr.club-internet.invalid...

"Harlan Messinger" <h.*********@comcast.net> a écrit :

| "Lauri Raittila" <la***@raittila.cjb.net> wrote in message:
|
| > Easy solution: Back to top links are not good idea, so delete both and
| > markup related...
|
| That's the first time I've seen that opinion. Why aren't they a good
| idea?

As a user, have you ever used one of them?
Yes, frequently.

Have you ever _needed_ one of them?

I don't need a mouse either--at least for web sites that build pages fully
accessible by keyboard--but I use one.

[Harlan Messinger]

"Alan J. Flavell" <fl*****@ph.gla.ac.uk> wrote in message
news:Pi*****************************@ppepc56.ph.gl a.ac.uk...

On Tue, 19 Oct 2004, Daniel [ISO-8859-15] Déchelotte wrote:

| "Lauri Raittila" <la***@raittila.cjb.net> wrote in message:
|
| > Easy solution: Back to top links are not good idea, so delete both and | > markup related...
[...]

As a user, have you ever used one of them? Have you ever _needed_
one of them?

No; but there's some kind of theory (I'd tend rather to think of it as
"superstition", I'm afraid) that they are useful for accessibility.

I haven't thought of them as useful for accessibility, but for general
usability. Not an important feature, but a convenience.

However, I don't know a browser that doesn't have several different
ways of its own to get to the head of the page[1], so I reckon they're
just a confusing waste of good markup. Those disabled folks that I've
met have shown much better abilities at using the browser's own
built-in tools on every page (rather than letting themselves be misled
by page-specific or site-specific ersatz features) - as compared to
some ordinary folks whose only "disability" was shortage of a clue.

cheers

[1] Notice that I didn't say "back to the head of the page", because a
reader who went to a fragment URL (wibble#whatever) might never have
seen the head of the page before - hence, it can be illogical to speak
about going "back" to it.

True--and that may actually be a good reason to have these links. They alert
the reader to the fact, which may not otherwise have reached his attention,
that the place to which he has been taken is in the middle of a larger
page--which might contain even more useful information worth exploring.

[Darin McGrew]

"Lauri Raittila" <la***@raittila.cjb.net> wrote:

Easy solution: Back to top links are not good idea, so delete both and
markup related...

Harlan Messinger <h.*********@comcast.net> wrote: That's the first time I've seen that opinion. Why aren't they a good idea?

Well, they do replicate a basic browser function (go the top of the current
page) that has been no more than a couple keystrokes (or mouse clicks or
stylus taps) away in any browser I've used. And in general, replicating
basic browser functions isn't a good idea.

But most actually reliably do what they claim to do, which is more than can
be said for most of the links that replicate browser functions (e.g., back,
close window, print).

So they're unnecessary at best. And in the rare case where the author
screws up, they're confusing when they don't actually take the user to the
top of the current document.
--
Darin McGrew, mc****@stanfordalumni.org, http://www.rahul.net/mcgrew/
Web Design Group, da***@htmlhelp.com, http://www.HTMLHelp.com/

"Entering Yosemite National Park: laws of gravity strictly enforced"

[Neal]

On Tue, 19 Oct 2004 13:08:33 -0400, Harlan Messinger
<h.*********@comcast.net> wrote:

True--and that may actually be a good reason to have these links. They
alert
the reader to the fact, which may not otherwise have reached his
attention,
that the place to which he has been taken is in the middle of a larger
page--which might contain even more useful information worth exploring.

If the page is long and has numerous link points throughout, it probably
ought to have a table of contents. "Go to table of contents" is more
meaningful than "Go to top", because the user might assume that they WERE
at the top, so why bother clicking there?

[Harlan Messinger]

"Neal" <ne*****@yahoo.com> wrote in message
news:op**************@news.individual.net...

On Tue, 19 Oct 2004 13:08:33 -0400, Harlan Messinger
<h.*********@comcast.net> wrote:

True--and that may actually be a good reason to have these links. They
alert
the reader to the fact, which may not otherwise have reached his
attention,
that the place to which he has been taken is in the middle of a larger
page--which might contain even more useful information worth exploring.

If the page is long and has numerous link points throughout, it probably
ought to have a table of contents. "Go to table of contents" is more
meaningful than "Go to top", because the user might assume that they WERE
at the top, so why bother clicking there?

If you're a user and you see a link that tells you "Go to top", you know
instantly that you're not already at the top, and it no longer matters if
you didn't know it a moment earlier.

[Neal]

On Tue, 19 Oct 2004 14:18:13 -0400, Harlan Messinger
<h.*********@comcast.net> wrote:

"Neal" <ne*****@yahoo.com> wrote in message
news:op**************@news.individual.net...

If the page is long and has numerous link points throughout, it probably
ought to have a table of contents. "Go to table of contents" is more
meaningful than "Go to top", because the user might assume that they
WERE
at the top, so why bother clicking there?

If you're a user and you see a link that tells you "Go to top", you know
instantly that you're not already at the top, and it no longer matters if
you didn't know it a moment earlier.

What I mean is this:

I click a link and I get this:

---------- top of viewport ----------
This is the information I linked to.

Paragraph paragraph paragraph, I need
to scroll down to get to the bottom.

This is the last paragraph.

_Go to top_

Now, I'm thinking this link will merely bring me to the starting point I
was at up the page a bit, which as far as I knew was the top. In fact, it
will bring me to a region of the page I don't even know existed.

If the link read _Go to table of contents_, well, I know I haven't seen
that. It's worth it to me to click here.

[Harlan Messinger]

"Neal" <ne*****@yahoo.com> wrote in message
news:op**************@news.individual.net...

On Tue, 19 Oct 2004 14:18:13 -0400, Harlan Messinger
<h.*********@comcast.net> wrote:

"Neal" <ne*****@yahoo.com> wrote in message
news:op**************@news.individual.net...

If the page is long and has numerous link points throughout, it probably ought to have a table of contents. "Go to table of contents" is more
meaningful than "Go to top", because the user might assume that they
WERE
at the top, so why bother clicking there?
If you're a user and you see a link that tells you "Go to top", you know
instantly that you're not already at the top, and it no longer matters if you didn't know it a moment earlier.

What I mean is this:

I click a link and I get this:

---------- top of viewport ----------
This is the information I linked to.

Paragraph paragraph paragraph, I need
to scroll down to get to the bottom.

This is the last paragraph.

_Go to top_

Now, I'm thinking this link will merely bring me to the starting point I
was at up the page a bit, which as far as I knew was the top. In fact, it
will bring me to a region of the page I don't even know existed.

I kind of know what you mean. But then, for regular user these links become
*associated* with pages where they might have arrived in the middle. So it
provides useful information. And for newbies who didn't realize that when
they come to a web page, they might not be at the top--well, you gotta learn
*some* time!

If the link read _Go to table of contents_, well, I know I haven't seen
that. It's worth it to me to click here.

That makes sense.

[Lauri Raittila]

David wrote;

Lauri Raittila <la***@raittila.cjb.net> wrote:
(in message news:<MP************************@news.individual.n et>)
[Code confusion about
A.backtotop:link { foo:bar;}
&
A.backtotop:link { bar:foo;}
]

Easy solution: Back to top links are not good idea, so delete both and
markup related...

Interesting. Perhaps not the forum for it (I imagine we'd get justly
flamed for discussing non-css stuff),

So X&fups to ciwah
but if you want to take this off
list I'd be interested to know your reasons for saying back to top
links aren't a good idea...it's something I've never heard before...

Here is what I can think:

1. There is functionality in browser
2. It never is only option
3. It makes keyboard navigation harder (assuming there is other links in
page)
4. It doesn't tell what is on the top - so why would someone follow that
link?
5. Sometimes link don't work, and user may avoid it because. For example,
if there is no anchor at the top of page. Surprisingly often top of
page links are done by just <a href="#">, and so don't work on some
browsers. (Opera 7.6p1, for example)
6. Only tells you that you are not at the top of document, if it is first
thing after anchor. Otherwise, one can think that it is link to back
the anchor, which user arrived to, if we suppose he didn't knew he
wasn't on top in first place.
7. Back to the top is confusing, if user never was on the top.

Can't think any use for them, that would really help.

I do use them sometimes, if availble. Even so often that I have noticed
#5. Also, when I tried to disable scrollbars, noticed #6. #7 is #7
because "back to top" links are so much used.

--
Lauri Raittila <http://www.iki.fi/lr> <http://www.iki.fi/zwak/fonts>

[David]

"Harlan Messinger" <h.*********@comcast.net> wrote in message news:<2t*************@uni-berlin.de>...

As a user, have you ever used one of them?

Yes, frequently.

Yeah, I agree. Use them all the time and find them quite useful.
Especially useful when using lynx/links, screen magnifiers. They're
more or less ubiquitous, have no detrimental effects on the browsing
experience worth mentioning, so you might as well provide them. Lots
of people use them. Rock on back to top links.

[David]

Neal <ne*****@yahoo.com> wrote in message news:<op**************@news.individual.net>...

What I mean is this:

I click a link and I get this:

---------- top of viewport ----------
This is the information I linked to.

Paragraph paragraph paragraph, I need
to scroll down to get to the bottom.

This is the last paragraph.

_Go to top_

Now, I'm thinking this link will merely bring me to the starting point I
was at up the page a bit, which as far as I knew was the top. In fact, it
will bring me to a region of the page I don't even know existed.

Yes. "top of page" would be more descriptive link text.

[Neal]

On 19 Oct 2004 16:43:25 -0700, David <da********@yahoo.co.uk> wrote:

Neal <ne*****@yahoo.com> wrote in message
news:<op**************@news.individual.net>...

_Go to top_

Now, I'm thinking this link will merely bring me to the starting point I
was at up the page a bit, which as far as I knew was the top. In fact,
it
will bring me to a region of the page I don't even know existed.

Yes. "top of page" would be more descriptive link text.

But that still doesn't avoid the problem that as far as I knew, I was
already at the top of the page. Why click it? I just read what was there.

Best to use content-related link text rather than position-related.

[Chris Morris]

da********@yahoo.co.uk (David) writes:

"Harlan Messinger" <h.*********@comcast.net> wrote:

[back to top links]

As a user, have you ever used one of them?

Yes, frequently.

Yeah, I agree. Use them all the time and find them quite useful.
Especially useful when using lynx/links,

Really? In those browsers I find it far quicker to hit home on the
keyboard than to press up/down a few times to get to the back to top
link and then press return.

--
Chris

[Harlan Messinger]

Neal <ne*****@yahoo.com> wrote:

On 19 Oct 2004 16:43:25 -0700, David <da********@yahoo.co.uk> wrote:

Neal <ne*****@yahoo.com> wrote in message
news:<op**************@news.individual.net>...

_Go to top_

Now, I'm thinking this link will merely bring me to the starting point I
was at up the page a bit, which as far as I knew was the top. In fact,
it
will bring me to a region of the page I don't even know existed.

Yes. "top of page" would be more descriptive link text.

But that still doesn't avoid the problem that as far as I knew, I was
already at the top of the page. Why click it? I just read what was there.

These links are especially useful on pages that begin with a large
number down-links. FAQs, for example. They simplify the pattern of
jumping down to one piece of information of interest and then
returning to the list of links.

Best to use content-related link text rather than position-related.

--
Harlan Messinger
Remove the first dot from my e-mail address.
Veuillez ôter le premier point de mon adresse de courriel.

[David]

Chris Morris <c.********@durham.ac.uk> wrote in message news:<87************@dinopsis.dur.ac.uk>...

Yeah, I agree. Use them all the time and find them quite useful.
Especially useful when using lynx/links,

Really? In those browsers I find it far quicker to hit home on the
keyboard than to press up/down a few times to get to the back to top
link and then press return.

You know, I can honestly say that until I read this thread, I had *no
idea* that the home button took you to the top of a page. I've been
building websites since 1996.

<makes sudden mental leap />

....and the end button takes you to the bottom! Excellent.

<repeatedly mashes head into desk />

[Neal]

On Wed, 20 Oct 2004 07:34:00 -0400, Harlan Messinger
<hm*******************@comcast.net> wrote:

Neal <ne*****@yahoo.com> wrote:

Yes. "top of page" would be more descriptive link text.

But that still doesn't avoid the problem that as far as I knew, I was
already at the top of the page. Why click it? I just read what was
there.

These links are especially useful on pages that begin with a large
number down-links. FAQs, for example. They simplify the pattern of
jumping down to one piece of information of interest and then
returning to the list of links.

Yes, but we're not arguing about the same thing. The link is a good idea.
Calling it "top" is not. It's the link text "top" that is ambiguous. As I
said,

Best to use content-related link text rather than position-related.

Name the link for the content it brings you to, not the geographical place
it goes. All goes back to not mentioning the mechanics, semantics, good
link text, etc.

[Harlan Messinger]

"David" <da********@yahoo.co.uk> wrote in message
news:99**************************@posting.google.c om...

Chris Morris <c.********@durham.ac.uk> wrote in message

news:<87************@dinopsis.dur.ac.uk>...

Yeah, I agree. Use them all the time and find them quite useful.
Especially useful when using lynx/links,

Really? In those browsers I find it far quicker to hit home on the
keyboard than to press up/down a few times to get to the back to top
link and then press return.

You know, I can honestly say that until I read this thread, I had *no
idea* that the home button took you to the top of a page. I've been
building websites since 1996.

<makes sudden mental leap />

...and the end button takes you to the bottom! Excellent.

<repeatedly mashes head into desk />

Except that I can't figure out how to get the keys to work consistently in
Firefox. I know, RTFM, but I haven't done that yet.

[Lauri Raittila]

David wrote;

Chris Morris:

Yeah, I agree. Use them all the time and find them quite useful.
Especially useful when using lynx/links,
Really? In those browsers I find it far quicker to hit home on the
keyboard than to press up/down a few times to get to the back to top
link and then press return.

You know, I can honestly say that until I read this thread, I had *no
idea* that the home button took you to the top of a page. I've been
building websites since 1996.

That is your problem - you have been testing in lynx, but not
really using it. Just like Chris, who haven't enabled quick go etc
function to follow link using numbers, instead of using clumsy arrows...
(makes lynx usability better than opera)
--
Lauri Raittila <http://www.iki.fi/lr> <http://www.iki.fi/zwak/fonts>