Where to draw the line between structure and presentation?

Not many people will argue that a table is a structural element of a graphic interface. Tabular data is meant to be presented as such. A desktop web browser, your $700 phone and links will render tables as is. Why is then relative container position not considered structure, but presentation?

You can't make a two column web page without using CSS (unless, of course, you use tables). Wouldn't it be weird if your desktop applications placed the Add button under the text box when you need it to be next to the box? Does it make sense to browse the web on links and having to scroll down at least three pages to get to the content because the super long left sidebar is rendered before the actual content?

Would adding layout elements like as position and size properties to the structure language of a GUI really make things less orthogonal? I'm not saying you should allow style attributes on GTK+, but it makes no sense that web documents use CSS for layout. You can browse the web without colors or special fonts, though usability is affected when you break the layout the original author had in mind. Not sure whether I'm missing an important design detail or the web is being an unorthodox hacked-together piece of shit as usual.

>>32
You probably mean that the incompatibility between UTF-16, UCS-2, wchar_t and char16_t (which are all meant to be 16 bit encodings, though wchar_t isn't required to actually be 16 bits) is a bloated mess.
UTF-8 is cool, man.

>>34
ASCII doesn't support box drawing characters. ANSI and Unicode do.

>>41
No, the Unicode standard is what's bloated. What the hell do I need a black man kissing white man emoji for?

>>42
Ok, you're right. It all started when they confused 'code point' with 'emoji' (wasn't it called 'emoticon' at some point?). Anyway, emojicons and emotis should be things like :-) and not actual code points. Which they also were at some point.
But hey, who says you have to actually support the whole Unicode space? My boxes with digits in them are enough for me. Just boycott faggy fonts and systems that support them.

>>41
As mentioned above, I meant "ASCII doesn't support box drawing characters". And is

And "ANSI" isn't a character encoding. ASCII and Windows-1252 are both ANSI standards, but neither supports box drawing characters. Those are only available with the Unicode/UTF encodings.

>>44
My default codepage, 850 (Western European, as returned by chcp), supports some box drawing characters and 437 supports even more.

>>42
Isn't the point of a universal text encoding standard to encompass all of humanity's text data? If this is so, why is it bloated?

>>46
Emoji isn't text data.

>>46

text data

What is text data? Letters? Glyphs? Pictograms? Symbolic Pictures? Pictures?
As I already said, they confused emoticons and code points. Where to draw the line anyway?

Personally, I think they should only add things that remain in continuous use in real languages for at least a few years. Remember that latin characters have been in use for hundreds of years and will likely continue to do so. But does anyone really think that even a few people will remember that there is a rollercoaster emoji (U+1F3A2) or that there is pictogram showing a clamp denoting 'compression' (U+1F5DC) in Unicode?