|  | Posted by dorayme on 05/03/06 06:43 
In article <nHx5g.3562$cZ3.1766@clgrps13>,"Michael Laplante" <nowhereman@twilightzone.net> wrote:
 
 > "dorayme" <doraymeRidThis@optusnet.com.au> wrote in message
 > news:doraymeRidThis-D55A3D.09231702052006@news-vip.optusnet.com.au...
 > > In article <Z%p5g.2928$Yy5.1479@edtnps89>,
 > > "Michael Laplante" <nowhereman@twilightzone.net> wrote:
 >
 > > It is good advice to be rid of tricky absolute positioning if you
 > > can do without it.
 >
 > Because. . .?
 
 Well, I have just quickly read the 38 posts following this in the
 thread and there are some answers that go into a bit of detail
 about this. But you can surely see that my "if you can do without
 it" and "tricky" in relation absolute positioning just has to add
 up to the advice that Mark Parnell gave you in the first place.
 
 Let me put it another way. Absolute positioning would be the last
 thing that you learn to use well in a css course, it is a
 counter-intuitive tool and gives surprising results. You will
 come to see that sometimes the best course is to avoid it
 altogether to begin with and proceed with great caution after
 that.
 
 Just a note about "relative positioning" that you often contrast
 with "absolute positioning" as if it is the natural contrast in
 every context. It is not the natural contrast. (kchayka, I think,
 hinted at this without - understandably - going into it.) The
 biggest contrast is no overall positioning at all (this is not
 "relative positioning"). Pages can rely on the order of the HTML,
 the instructions about margins and other css, even floats (these
 are taken out of the flow in controlled ways but are not part of
 what is known as "relative" positioning in css.) Relative
 positioning is a technically specific thing to do with offsets.
 
 > > Your blue menu floats to sometimes obscure text or look oddly
 > > misplaced (it looks to have no place!). And it breaks badly when
 > > text is enlarged.
 >
 > You're the first one who has ever mentioned this. Nor can I can replicate
 > this behaviour in any of my three browsers. What browser / version are you
 > using? Do you use a personalized stylesheet?
 
 No personalised sheet, no. I was using a variety of browsers on a
 Mac. If you cannot replicate it at all, I suspect you are not
 getting my meaning. See below.
 
 > "It looks to have no place!" Can you expand on that?
 
 When the window and/or text-size is varied, this menu floats
 about in unexpected ways, sometimes obscuring things that may be
 important underneath. Even when the window is large and the text
 small enough, there is a sense that it it is jammed up against an
 edge, as it if needs some margin of grace outside itself. This
 bit is an aesthetic matter which needs sensing. You have the
 sense of needing grace with margins for printing, now you need to
 imagine grace of a slightly more complex kind. Best I can do old
 chum? All this is, of course, has nothing to do with the
 advisability of having such a scripted object. It could happen to
 any absolutely positioned object!
 
 > What do you mean by "breaks badly?' Lots of pages -- even ones that
 > validate -- break badly when text is enlarged.
 
 I mean that the text in the menu half or wholly disappears too
 easily. You seem to take some comfort or refuge in the principle
 that perfection is rare to see. The point is that it is too far
 from perfect when just 2 or 3 clicks of text enlarging and window
 variations causes trouble. Jonathan Little (to whom, btw, you
 were unnecessarily rude. Yes, he's another mate of mine, so piss
 off!) made this point (naturally, not with my sense of eloquence
 - but he is just a simple artist and keen and knowledgeable
 student of html and css)
 
 We must be careful of this mistake in thinking. Not too many folk
 would disagree with you about even good sites breaking (which
 means roughly doing surprising things that shock idealists) under
 extreme conditions. Basically you need to cater for a reasonable
 variety of viewing or using conditions.
 
 --
 dorayme
  Navigation: [Reply to this message] |