Re: Should we adapt WYSIWYG in CSS?

Fri, 21 Feb 2003 12:25:36 -0500 (EST)  Mikko Rantalainen wrote:

>> 2) It can be done easily. 
>> I think that a reference property is sufficient, if implemented correct
>> Something like this:
>> reference-type:        screen-width  | client-width | none 
>> reference-size:        <pixels>    /*  designers screen width */
>> referenze-minimum: <pixels>    /*  prevents ambigeous small scale*/
>> reference -width:     <mm>        /*  designers screen width in mm */

>Could you explain briefly why do you think that the full screen 
>resolution in pixels should have difference in the rendering?

I am not sure I understand your question.
As pixels is the only unit appliable on many layouts, 
to prevent pixel-based fonts, and the whole layout 
to become too small for normal eyesight, the display needs
to be rescaled according to actual screen resolution, when 
viewed on a high res moinitor. Or I have to scroll to see the 
page in a low res. monitor.

> If I have 
>3000x2000 resolution screen I probably don't want 3000px long lines. 

Depends on the physical size of your screen, 
and the distance it is wiewed from.

>Scaling according to viewport width might make some sense. 

I agree, and changes my proposal to include that.
(On a  second tought, client-width reference seems to be wrong. 
I actually think wiewport-width is what I had in mind.)

>However, not 
>all of us have similar eyesight and some just prefer to use smaller or 
>larger font than the page author. 

I accept that. I don't intend to make the webpage that rigid 
that this can't be changed by the user. I only wants that the 
user don't have  to change screen resolution or buy a new 
monitor to wiew a webpage properly.

>And if I guess correctly, the only 
>reason for the "correct" font/element size would be that the author 
>didn't know how to design fluid layout.

No, that is not the reason. Actually, my proposal can be, 
and will usually be combined with a fluid layout. 
The main reason is that some content are of such nature 
that the designer needs to have strict control on the layout 
to be able to show the content right. User CSS might still be used, 
and override designer CSS. Only contents that needs to be fixed, 
needs to have an  important! property attached.


>What is the intended usage for that "reference-width" property?

To be able to also relate "absolute" length units like mm to 
the screen resolution. As it is now, the relationship is not 
precisely defined.

>If the layout is more important than usability and accessibilty, I'd 
>suggest something like PDF or perhaps SVG instead of anything combined 
>to CSS. Those can display exactly what you want and scale nicely.

CSS and HTML is much more widely used, faster,
more flexible, easer to use and better for many purposes.

>> 3) Who should decide?
>> Should the CSS standard forbid what people wants to use?
>> I think that the webpage author should be more respected. 
>> You must have in sight that the need for a WYSIWYG approach is high.

>I think the webpage author should be more educated. 

You cannot change people's education level.
Remember that most of the web authors is  ordinary people, 
and their profession not being a wellpaid webpage author. 

>The problem is that 
>those webpage authors still author documents for the web just like they 
>author those documents for the paper. Those two medias are totally 
>different and require different logic for layout.

Web is also paper. If I desig a web document, 
for instance an article of web design, 
I might want it to be printed on paper also.

>I hope that the web would be semantic in the future and *I* could decide 
>how *I* want to view *the content*.

I actually agree. I don't see any big conflict here.
The web in future will be both semantic and look in the same time.

>> 4) There is no really drawback
>> The old philosophy can still be used.

>Define "old philosophy". Do you mean that this new WYSIWYG method allows 
>designing layout in the same way we design paper layouts today. IMO, 
>that's not good enough. And if you mean that this new system allows 
>designing fluid layouts in the same way the current specification 
>allows, then we clearly don't need any enchancements.

No, this will not be paper layout. But as in paper layout, 
you can define width of media. And you can control size 
of elements relative to media width.

It is not a fixed layout. 
For instance: If you don't have the desired font, It will still be rendered with the font that UserAgent decide. The result is of course that line breaks etc. will adapt. Or if the user wants to look at it with a larger font, everying will render different. 

But the layout will be approximately the same on 
screen-resolution 800x600 and 1200x800 and on print. 
It will be easier to make pages look good.

>> I also think that such an adaption will not spoil anyting. 
>> It actually makes the content oriented design easier.

>Anything that helps keeping fixed layout is bad (fixed as in the layout 
>stays exactly the same, like in Opera's page zoom. As I understand this 
>proposal the target is to have something like Opera's zoom tweaked so 
>that the fixed content always just fills the viewport or screen). 

That is only one possibilty with my proposal. 
Another is that the user might change the settings by user CSS. 
And the user can still change font size preferences, 
if the author not explicitely have forbidden that with 
a Important! statement.

>Layout 
>must be designed in such way that it doesn't break even though the font 
>size is changed. And one should be able to resize font without scaling 
>all the other dimensions of the page.

Ok, I agree. I don't see any conflict here. 
The only exception is where the layouy needs to be so that 
changes of font-size will ruin the layout and make it unreadable. 
Designer must be able to control those cases that do explicitely.

>> The historical web as an information library only, is outdated.  

>Is it? And do you want that? I want the web to be just one honking big 
>library that I can view with the style I select. The content itself may 
>contain nice pictures and like but the pure information related to those 
>objects should be displayed the way I want, not the way the author likes 
>to view his information.

I agree only partly in the main lines here. 

Problem is that "pictures" also sometimes is constructed by html.
It is the power of DOM, CSS and scripting coming in here and 
changing the whole web to something new.  

Look at Dynamic webpages as a dataprogram, or application. 
This is a new media revolution coming up and alive. 
The possibilities is still unimagined by most people, 
and the use will have great impact on what the web will 
be in the future.

>I understand that not all are ready for such system, but I hope you 
>understand that as the web grows larger and larger we cannot find useful 
>information from it unless we have semantic web. 

As things is today, useful content is often hidden in 
images and that like, makeing it unpossible to search 
for it on the web. 

I also think that this problem might be better adressed 
by developing better ways of marking content of a webpage. 
Maybe something like 
<META name="content-type" content="commercial" > 
would be of help?
A long list of content-types could be standardized. 
Maybe sub types also.

>Compare web pages to 
>email: I hate HTML mail because it allows the sender to decide which 
>font the text is, how big the text is and what color it's rendered with; 

I love HTML mail for the same reasons. 
All those nice christmas letters I have recived. Great!

>I lose the control (not really, my user agent renders HTML mail as plain 
>text). Losing control results to decreased productivity and lost time 
>due to inefficient information transfer to my brain. Uniform look&feel 
>for different content is where we should target to.

Colors and style is also information, it talks to yout heart.

>> 6) Freedom of speech
>> You should not try to control how people use the web.
>> The web has made it easy for everyone to express opinions, 
>> or make their artwork public. 

>I totally support freedom of speech. However, I also think that one 
>shouldn't mix the content and the media. CSS should be the part that 
>describes how that media behaves under different environments. And that 
>behaviour should be better than "just scale the whole page to fit".

Many people wants a behavior of scale to fit. 
It is logical in most situations.

>The stuff I regard as "art" isn't something that should be expressed 
>with some document markup language combined with CSS. Formats like PNG, 
>JPG or SVG sound more like what I've in my mind. (And yes, SVG is a 
>markup language that can be combined with CSS :) Some other possible 
>formats for art include MPEG4 or Ogg.

It is not up to you to decide what others regards as art, 
or how to express it. 

>> 7) Avoid messed up use of  html/css
>> The css standards job is to let the designer be able to
>> do it "right" or "wrong" in a logical, consistent and correct way.

>That's a nice requirement... How would you know once you have achieved 
>that one? How about a short description on this "consistent and correct 
>way"?

The designer should be able to express for what resolution 
his design is intended for, and how he intends it to scale on 
different media-width or wiewport-width.
The user can then override those settings by setting 
user CSS. Thus avoiding doing this by javascript or 
complicated layout that is not part of content or semantic.

>> 8) Adapts better to context philosophy
>> Content is much more than text. Text and style is closely connected. 
>> The word context describes this relationship. 
>> 
>> Do you agree or disagree?
>>Is there other good reasons for?
>> Is there ANY good  reasons AGAIST?

>I agree that we need easier methods for nicer layouts. However, I'm 
>afraid that the proposed method isn't the way it should be done. Most of 
>all, I'm afraid that this proposal would allow an inadvertently author 
>to harm the reader too easily.

I think it will prevent such harm of the reader.

>Perhaps I've just misunderstood your great design. 

(The Idea of a screen-width property is not mine. 
I have just adapted it, and changet the idea somewhat. )

>Could you explain how 
>this could be beneficial without simply scaling everything on the page 
>(like Opera's zoom).


It don' behave like Opera's zoom. It behaves like this: 
You change screen resolution, and the content looks 
approximately the same as it did with the other screen 
resolution. 

If designer have used a reference to wiewport-width the whole 
page will rescale as you resize the window. The designer can 
define a minimum width. If wiewport is smaller, the picture 
stops scrinking, and you have to scroll to se all. 



Gaute Sandvik