My assumption: Screen readers do not download images.
? A browser downloads images and a screen reader reads from the screen. I think that PPK misunderstood the simple fact that a screen reader is not a browser.
Or am I missing something?
There are more intrusive, and less intrusive IR methods. Although I thought up three of them, I quite intensely dislike the concept of image replacement itself. There are a myriad of ways to ensure that imaginative imagery is reconciled with textous textual textuality, and none of them requires a PhD or any significant investment beyond a handful of idle brain cells. One may counter this statement that in certain circumstances, for certain needs, there is no way out. To such a dictum, I will reply that precisely these demands are weird, to my mind at least. Is is fashionable to criticize various IR techniques, whilst at the same time forgetting that the premise itself deserves the strongest critique.
I disagree with your classification of screen readers as non-browsers. The ability to download images is perpendicular, being a function of some browsers rather than a defining quality. Lynx is a browser. It doesn't touch images.
Screen readers are browsers. The only thing that makes them different is that the way they convey their media; speaking it, rather displaying it.
I think you are wrong there. I couple of weeks ago, I visited Stichting Accessibility and they showed how those things work.
Internet Explorer is launched and the screen reader, already running in the background because it can be used for more than just internet, reads the screen.
A browser interprets a language. Those languages can be HTML, JS, CSS but also other languages like VRML, SVG and so on.
You say that a browser needs his info from the WWW, but a browser can get it's information from another application too in my opinion. A screenreader is a browser, that gets its information from another browser.
When I look in my dictionary, the exact meaning of "browsing" is "grasduinen". So I think that a screenreader browses Internet Explorer.
And maybe Internet Explorer downloads those images (but it can be set off, and I think most blind people do so), but a screenreader doesn't use them. It's like, not having a plug-in. But on the other hand, are plug-ins browsers? They browse within their own known language, and plug-ins like Java and Flash have their own sandbox, just like a browser.
It's just a way of looking at things, and both your opinions are good I think. There's no real definition of a "browser".
"Furthermore, CSS is the way to replace an image"
It has to do with presentation, not with behavior.
Yes, that's a good argument. In fact, I'm glad people start thinking in these terms.
I'm not going to spill the contents of my new column right now, so I have to ask for some more patience.
I think it is a little more complicated that Anne has suggested. I think that screen readers that respect web languages, such as HTML and CSS, should be considered browsers. Those that don't (and by that I mean those applications that simply read any text rendered to the screen, irrespective of source) are not. I don't see how it has anything to do with images. Perhaps all screen readers should be regarded as user agents, however.