|
|
Posted by David Mark on 10/29/07 12:33
On Oct 29, 7:55 am, Neredbojias <monstersquas...@yahoo.com> wrote:
> Well bust mah britches and call me cheeky, on Mon, 29 Oct 2007 11:34:51
> GMT David Mark scribed:
>
> >> >> You can j/s preload sequentially: ie, not starting the following
> >> >> preload until the previous is finished. I've done that and it
> >> >> works. But a better idea (I think) is just to make a
> >> >> position:absolute; visibility:hidden; div "layer" encompassing all
> >> >> the images which won't show because of the css.
>
> >> > That will mess up the semantics of the page and will look strange
> >> > when style is disabled. For the scriptless approach, it is better
> >> > to use background images.
>
> >> Background images don't always load as one might wish, though. The
>
> > I don't know what you mean by that. But the typical need for
> > preloading is for rollovers and the like, so it isn't a catastrophe if
> > the images fail to preload.
>
> No, of course not. But I seem to remembering seeing in the past
> background images sometimes loading last (-in ie, I think) and that sort
> of goes counter to the whole idea. And neither is it a catastrophe if
> one adds a hidden layer with regular images.
>
> >> styling-disabled is a valid concern, but despite conventional
> >> mythology, styling is necessary nowadays and anyone who disables it
> >> except for testing
>
> > That is only part of it. Some agents don't support style at all.
>
> No, but I'll bet 99%+ used actively on the Web today do. A real
For the sake of argument, say it is 0.1%. Multiply that by the number
of users on the Web today and you get a fairly substantial number.
> consideration might be the useragents of ipods, picture phones and the
> like.
>
> >> is a moron. As for semantics - phffft! Very few pages have correct
>
> > It is not the users' fault if a document is poorly designed.
>
> Who said it was? We are discussing the viability of an additional layer
> for a specific purpose.
You said that those who disabled style were morons. If the document
doesn't work in that case, does it reflect poorly on the user or the
author?
>
> >> semantics, anyway, and a sequential list of links in a "layer" will
>
> > Very few pages are written by competent Web developers.
>
> >> certainly not mess them up if they are correct.
>
> > What list of links? The suggestion was for a "layer" with hidden
> > images. Search engines, screen readers, style-challenged agents, etc.
> > will have no idea what to make of such a thing.
>
> I meant "list of images". Why would search engines have a problem? As
Do you mean a series of images. An HTML list would further complicate
the semantics.
> for screen readers, it may cause some confusion but I don't evaluate this
> anywhere near reason enough to avoid the technique. Unquestionably, it
Why would you want to confuse the sight-impaired for the sake of
faster rollovers?
> will work and work well. Background images may also work, but there is
It will work for users with CSS support and without impaired vision.
> some doubt. And lastly, I'm not one of those who subscribe to the "least
> common denominator approach" to web page creation just to strictly
I don't see what this has to do with an LCD approach.
> satisfy certain concepts which are arguable to begin with.
It is arguable whether the images will load faster as inline or
background images. So why use a technique that may or may not
slightly improve the load time of rollovers, when it will definitely
have other negative implications?
Navigation:
[Reply to this message]
|