Re: Vote on Omar Design - Mailing list pgsql-www

From Omar Kilani
Subject Re: Vote on Omar Design
Date
Msg-id 419F481A.8070003@tinysofa.org
Whole thread Raw
In response to Re: Vote on Omar Design  (Alexey Borzov <borz_off@cs.msu.su>)
Responses Re: Vote on Omar Design
List pgsql-www
Alexey,

> Yes, but these changes were made for current wwwdevel design and can be
> ported to your one with zero additional work.

The other way around, however, doesn't. That's the point.

Those changes weren't made for the current wwwdevel design. They were
made for our design, and back ported to the current one. The time wasted
back porting for no good reason is the issue.

>> Me no English?
>
> Sorry if it offends you, but your name is not quite English-sounding to me.

Neither is yours. Yet I never presumed anything about your level of
English competency.

It's a global world and you're on the Internet. People with non-English
sounding names are everywhere, and a lot of them speak and write better
English than those with English sounding names. Please be mindful of
this fact before making assumptions about people.

> BTW, the issue I had with Lukasz' design is that current position within
> the site's navigation structure was not highlighted. I see the same is
> true with your design. Can this feature be added?

It can. But, we're trying to finish the content first. Getting there.

Since the side navigation persists across every category, and the top
element is always the root of the category, it didn't/doesn't seem
necessary.

> "current web-design mantra"? Fixed-width is so 90s! Besides, table-based
> designs are also so 90s: look at new mysql.com done with strictly
> CSS-based approach.

The HTML can be redone at any time by changing one file, thanks to the
CMS. Even then, it's valid XHTML. And, it looks the same in Mozilla,
Opera and IE. Oh, and it looks great in a text browser, too.

The current wwwdevel design uses tables, too. And if fixed width is so
90s, then why is every major company still using it? Even design
companies like Adobe and Macromedia? What about well respected web
development resources like A List Apart (http://www.alistapart.com/)?
Yahoo also used fixed width. As does www.diveintomark.org.

Do you like how mysql.com looks? Do you think their *users* care if
they've used divs or tables? mysql.com doesn't validate. Would you stop
using mysql because of that?

If you resize your browser window to 800x600, mysql.com looks horrible,
as the top navigation elements move onto a new line.

> Joshua pointed us in the direction of Mozilla and RedHat sites. Well,
> Mozilla's one has a variable-width approach.

Let's take a look at mozilla.org. I'm on 1280x1024. The mainContent div
is using 910 pixels of my screen, and it's centered. The side navigation
is using 225 pixels until it meets the text. Therefore, there's 910 -
225 pixels = 685 pixels of usable screen real estate in which I have text.

Now, let's compare this to our design. 765 pixel main content table. 165
pixels until the side navigation meets the text. 765 - 165 = 600 pixels
of usable space for text. I get the exact same results for 1024x768 --
it didn't scale up.

800x600: Mozilla = 732 - 182 = 550 pixels of usable real estate.
800x600: postgres= 765 - 165 = 600 pixels of usable real estate.

Mozilla.org is a very nice looking site. But that nice, variable width
site is valid... HTML 4.01. Setting the validator to XHTML 1.0
Transitional gives 74 errors. Do you care? Will you stop using
Firefox/Thunderbird/Mozilla?

Look at what the semi-variable width design does to the "Firefox 1.0 is
here" div. There's a huge light blue space between the green box and the
Firefox logo. It doesn't look good. The same thing would happen to the
elephant on the front page in our design.

Variable versus fixed width shouldn't hold up progress. Yes, it's
possible to make it variable width, and yes, I understand the geek
factor of variable width: scalability, flexibility, etc. But it does not
take into account usability or aesthetics, which are much more
important. If variable width is a requirement, then the requirement is
flawed.

Everything has pros and cons. When taking our design into account, I
believe the pros far outweigh the cons. That makes it a winner to me.

> As for RedHat... well... the more I look at that the more I see some
> uncanny *similarities* with your design. Care to comment?

Look at mozilla.org, macromedia.com, apple.com, redhat.com, nikon.com,
wacom.com, ford.com, benq.com, fiat.com, blogger.com, gmail.com,
oxygen.com, nbc.com etc.

Rounded design is the style right now. You can even read some
interesting articles about rounded edges being "in vogue." Square edges
are so 90s. ;)

> P.S. I had some questions [1] concerning your language-handling patch

1) Didn't see it in .htaccess. Revert if you like. Though I haven't seen
anyone not have it in their code.

2) I wanted to experiment with many other languages, as per our pgweb,
and wanted an easy way to change things without knowing the control
codes for 9 different languages. Since the internationalisation support
wasn't really 100% anyway, I had a look around at other websites, most
notably WikiPedia, and saw that most people use entities for encoding.
So I did. Revert if you like.

3) It handles them by ignoring them:

+            $accepts_lang = explode(';', $accepts_lang);
+            $accepts_lang = $accepts_lang[0];

Since they're already sorted, and you want the first valid language.
Again, revert if you like. Or add support for q.

Best Regards,
Omar Kilani

pgsql-www by date:

Previous
From: Devrim GUNDUZ
Date:
Subject: Re: New News Entry
Next
From: Alexey Borzov
Date:
Subject: Re: Vote on Omar Design