W3C home > Mailing lists > Public > www-style@w3.org > July 2005

Re: New layout language.

From: Orion Adrian <orion.adrian@gmail.com>
Date: Wed, 6 Jul 2005 13:56:36 -0400
Message-ID: <abd6c80105070610565e651214@mail.gmail.com>
To: www-style@w3.org

On 7/6/05, Kris@meridian-ds.com <Kris@meridian-ds.com> wrote:
> 
> 
> Orion,
> 
> Most of what you've said is patently absurd.  I've argued with you, I've
> humored you, I've even taken time out of my day to educate you when you
> couldn't (wouldn't) take the time to figure it out for yourself... despite
> the fact that it's obvious.

My point is that it wasn't obvious. You yourself said, you had to
figure some things out. I've spent years with CSS and while I haven't
recently, the system is very complex in how it interacts with itself.
 
> We've both agreed that your system is essentially just a group of
> shorthands, so my 8 elements is equal to your 4 in what it does.  Not to
> mention what it says.

I initially agreed, but it's not. CSS relies on bottom and right to
get layouts which means that CSS won't degrade in some cases where
mine will.

> I see now how your system provides for some margins.  However you're still
> going to need to implement right and bottom margins for certain instances.
> But I agree for the most part your system can handle most margins... most.

Bottom margins are in there. It's all handled quite nicely in fact.
The one part your missing is that the viewport itself is a region with
it's own size.
 
> Again I'll state that this:
> 
> .left {
>   position:absolute;
>   top:10px;
>   left:10px;
>   width:200px;
>   bottom:10px;
>   min-height:400px;
> }
> 
> isn't much different from this:
> 
> .left {
>   left: 0%+10px;
>   top: 0%+10px;
>   width: 0%+200px;
>   height: 400px + 100%; (which again I'll note doesn't have a bottom
> margin)
> }

That's simply taken care of by providing an initial reference point
for the viewport (height, width).

> Additionally, the CSS is more explicit up front which makes it easier for a
> beginner to read and understand.  Which was your whole point initially.

Having taught CSS to beginners and they have a lot of trouble with it.
It's too complex, requires too many constructs and mixes margins,
widths, heights, absolute positioning with static and relative
positioning. It takes quite a bit of time to get it right and
understood. More training is never the answer to dealing with a
complex system.

With pictures, I taught my system in entirety to someone in 3 minutes.
I would doubt teaching the CSS box model would be so simple.

> I understand your point about degrading gracefully since you're only
> measuring from left and top... however css with calc is absolutely capable
> of exactly the same thing.  You've still proposed nothing new.  And my
> notion of min-left and such would handle the same issue without having to
> deal with calc.

And CSS calc() would look a lot like what I have shown but with a
calc() around each equation. It's also received less than a welcome
response by some of the editors.

> Concerning CSS I'll simply point out that what you've
> describe is by no means compelling, and is no different than Absolute
> positioning today with the noted MINOR exceptions.

Given the primer text, even if the syntax were identical there is
major benefit in seeing a split both in CSS and HTML.

> With that said, I'd
> have no hope of seeing it implemented, and I'll quitely point to the fact
> that current CSS standards have been overlooked over there despite that
> fact that I find a number of things QUITE compelling.

What does this tell you? That they're compelling to some, but not to
most people. I can't imagine a usability specialist out there that
would tell you the CSS model is fine as-is, and couldn't use
simplifying.

> "Most layouts in use can be accomplished with either CSS or HTML tables in
> IE. I would imagine to them their job is done."
> 
> The whole point of CSS layout was to get away from tabled layouts to begin
> with.
[snip]
> Thus the entire concept that we
> should continue to use tabled layouts is a crock.  If our goal is usability
> then we should abandon tabled layouts like the horrible mistake they so
> obviously were.

I was trying to offer up an explanation of Microsoft's behavior
beyond, they're an evil company ruled by Satan himself. I have never
said, let's go back to tables.
 
> With that rant out of my system, let's dicuss this point:
> 
> "...the benefit of the system resides not necessarily in it's layout model,
> but in the fact that it separates layout from formatting (a plus)..."
> 
> This can be done with css currently.  You could break up your css into two
> seperate css documents.  One layout, the other formatting.  In addition, if
> this doesn't appeal you can simply seperate them all in one document.  I
> can state .left{} in more than one place.  Once to describe layout, again
> to describe formatting.  That's no different than what you're proposing.
> Let's not confuse your ignorance of CSS with a need for changing the
> system.

That separates file-wise, but it doesn't separate it in source,
concept, syntax or capabilities.

My ignorance of CSS as you put it, isn't ignorance. It's more of,
"gee, look at this very complex system. I think I have a grasp on it.
I can do a lot of layouts with it, but the code isn't very readable.
Isn't not very semetric. Hmm... play around for 30 min. Didn't get
it... must not be possible."

What you're asking of people to use their valuable time to learn a
complex system that doesn't need to be complex. Their time is being
wasted. Let's not do that, ok? The system was invented... the only
complexity in it is what we put into it. It doesn't rely and anything
in the physical world which is what complicates physical objects.

> I'm not even going to address your other points since I don't really see
> their significance.  I don't see how your proposal changes CMS creation in
> any significant way.

Currently I have two options when working on a document. I either put
in all the overhead (banners, ads, navigation) in with the document or
I use a CMS system to merge them at delivery time. If there is a split
I don't have to do either one because the merging is done by the
browser at layout time.

> Final Thoughts:
> 
> I've humored you for quite some time now.  I've taken time out of my
> schedule to prove to you that CSS does what you say it cannot.

For which I appologized. I appologize again, but I don't know if it
will do any good. It does lend credence to my obviousness argument
though. People didn't know for certain it was possible. That should
always be known ahead of time if possible.

> This isn't
> to say that the experience wasn't worth the effort.  In fact, who knows,
> perhaps CSS will benefit from this discussion, but that doesn't excuse your
> ignorance of the system.

How many years do I have to work with CSS before I'm considered enough
of an expert to use it. Why is the system so incredibly complex? It's
not to the people here some of whom have been working on it from day 0
and have designed the system themselves. Just because there are people
here who understand it doesn't mean that's the case everywhere else.

> I don't mind if you truly understand how the
> system works and think it needs tweaking (or tossing out completely,
> although I find that highly unlikely).  I don't mind if you're ignorant and
> could care less. What I do mind is that you're ignorant of the system and
> how it works.

I am ignorant to a lot of things. CSS isn't one of them. I know about
CSS even if my assumption for some of the more complex things wasn't
accurate. I'm willing to bet that most people out there wouldn't know
how to do it either. That to me produces a real problem.

> You won't take the time to learn it yourself, so instead you
> propose that the whole thing should be changed.

This is where the problem is. If I won't learn all there is to know,
I'm somehow harming you? If I knew virtually nothing of the matter,
the result would be the same. It's too complex. It doesn't accomplish
goals it could.

CSS is designed for the masses, right? Or am I wrong? Is there some
desire to see this language limited to those people who can pass our
little intelligence test? If CSS could do what it does, take less time
to learn and take less time to code wouldn't that be worth it?

> That is, in my mind, the height of arrogance.  I don't want to flame the crap out of you.

I'm not being arrogant, I'm being a user. Users aren't going to learn
your system because you say so or that if they don't spend years
mastering it, they're being arrogant. They're going to say, "Look
arrogant <bleep>. I don't know why you think my time isn't as valuable
as yours. Make it simple or I'm going elsewhere."

> You're a
> human being like any of the rest of us, and you deserve some respect.  But
> that respect was granted to you before you started this conversation and
> has slowly degraded up until right now.

Why? Because I said it couldn't be done and you did it? Because I
didn't learn the language to the level you did? Or because I refuse to
learn something that I find complex?

> At this point I truly understand
> that you don't know CSS, you don't care about CSS, all you care is that you
> get your way, and you want us to back you.  Well I'm sorry, but I'm done
> with this conversation.  It was stimulating to be sure, but you refuse to
> listen to reason, and you refuse to learn from your mistakes.  And I'm out.

I appologized? What more do you want from me? When proven wrong I said
"I stand corrected". This, to you, is arrogance? What would be an
appropriate response to being wrong to you? Seriously I want to know.
I do listen to reason, I really do. CSS is capable of a lot more than
I gave it credit for.

And I don't care about my way. I'm a usability specialist who
concentrates on the user. My mental model doesn't matter. Theirs does.
I take no pride in the design put forth. I didn't even specify a
syntax. I just specified my understanding of the mathematics.

-- 

Orion Adrian
Received on Wednesday, 6 July 2005 17:56:46 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 27 April 2009 13:54:39 GMT