[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Random question: GNUStep site design
From: |
MJ Ray |
Subject: |
Re: Random question: GNUStep site design |
Date: |
Sun, 20 Jul 2003 23:41:20 +0100 |
Yes, the link you gave is the plain. The main other thing that I
think needs more thought is how the site should be structured and how
we should promote GNUstep in general.
On 2003-07-20 21:22:22 +0100 Scott Stevenson <address@hidden> wrote:
[...] Is the revamp something you personally want to do or something
you feel you have to do?
More that I want to.
[...] The problems were simple -- just misnamed selectors/properties.
Oh yeah, that's a fix for a cockup of mine. I rather stupidly merged
the CSS before retemplating, which let to interesting side-effects on
the current site, so I introduced some typos to stop the selectors
matching. Sorry! I forgot about that. I'll create a template dir
and put working copies in there soon. Well-caught on text-color,
though!
MacIE5 is another story. It has this interesting "behavior" that
basically boils down to this:
If these conditions are met for a given CSS box (div container):
1: The box contains block-level display elements
2. The box has an undefined width
There is no undefined width. It is auto.
... the box will not automatically scale to the minimum size
required for its content.
Instead, it will grow to the maximum size allowed by its parent.
It is meant to be the intrinsic width, not the maximum.
Additionally, if a third condition is met:
3. The box is set to positon: absolute
The box boundaries will actually be too big for the canvas, and
trigger the browser's horizontal scrollbar.
Applying these rules to the site results in:
http://createbynumbers.com/downloads/gnustep/macie-before.jpg
I hate IE because we can't fix cockups like that. (I hate Opera for
similar reasons.)
The MacIE folks felt they wre following the W3C specs in at least the
first two conditions, although there have been comments from them
about ambiguiouty in the spec.
The trigger in the case of common.css is the display:block
declaration for the div#menu child elements. I'm not sure what you
want to do about this, if anything. The easiest approach would be to
set some sort of width on div#menu, but that has drawbacks, of
course.
Yes, like having to pray that the font is reasonably close to the one
that I am using when designing, that the longest line is something
like the same number of ems. Not acceptable, IMO.
The current CSS design came about to fix a breakage of IE5/Win.
Wonderful that IE5/Mac is differently broken and it doesn't look
possible to satisfy both while having a fully-working design.
If we take
enough can you licence them under GPL or something suitable (PD
whatever)?
Then, I think we'll give you a link from a credits and copyright
page.
PD is probably the easiest. GPL would be fine too, but in practice
I'm not sure there's much difference when you're dealing with a web
page. It's hard to deteremine what derivative works really are.
Possibly.
Can you subscribe to gnustep-webmasters and discuss this there? I
think others may have views on this.
--
MJR/slef My Opinion Only and possibly not of any group I know.
http://mjr.towers.org.uk/ jabber://address@hidden
Creative copyleft computing services via http://www.ttllp.co.uk/
Thought: Edwin A Abbott wrote about trouble with Windows in 1884