First of all sorry for this perhaps silly question. But I'm trying to figure out if the <dir="rtl"> attribute really is mandatory when displaying arabic characaters on a web page.
I've searched for a while now on this subject but I'm still puzzled. The W3C website states: "If a document contains right-to-left characters, and if the user agent displays these characters, the user agent must use the bidirectional algorithm." Source: https://www.w3.org/TR/html4/struct/dirlang.html
As far as I understand the above text this means that you will have to use <dir="rtl"> attribute but I can't find out why exactly.
The reason I'd like to know is that we have to create a 'somewhat' arabic version of a page without spending too much time on aligning and mirroring the whole website.
Does somebody know why the <dir="rtl"> attribute is mandatory? Or why it maybe why it is not mandatory? The more I read about it the more it starts messing with my head and I have a hard time finding a clear answer.
Kind regards,
Jan-Willem
This is mandatory because; 1. Its makes bots and users know how what language your page uses. 2. Its would display the text according to how that language is expected to be, rtl is mearnt for arabic and/or hebrews.
I was curious how Imgur was rendering their upvote/downvote arrows:
I assumed they were images, but I found something that I did not expect:
A custom font that contains glyphs for up and down arrows, mapped to the 'o' and 'x' characters, respectively:
Is this method considered acceptable these days? I have never considered using a custom font for something that doesn't semantically map into an alphabet. This approach is not even on my radar of best practices for web design.
I can imagine the reasons for:
Your site uses a standard icon set that can be mapped to single-character codes.
You only need control over foreground/background color for the icons.
You want icons that scale the same as text.
I want to know any specific reasons against using this method.
In particular, I'm looking for answers that address any of the following:
browser/platform compatibility
future maintenance implications
semantics
performance
standards compliance
The only thing I have come up with so far, is that, semantically, it does not make sense to map an upvote icon to the character 'o' and a downvote icon to the character 'x'. And, just to be specific, I'm not talking about keyboard mappings, but rather language mappings, character codes. It seems to me that raster images or SVG are much more preferable alternatives in this case.
I thought of one other possibility: language and encoding compatibility. Would the html lang attribute or character encoding of the page have any effect on the character mappings into the font in the CSS stylesheet (the stylesheet uses 'x' to represent a downvote icon)?
However, I'm certain Imgur has thought all of this through already. So, why am I wrong?
Modern browsers (e.g. IE9 above) support custom fonts.
Even Bootstrap also uses custom fonts for icons, known as Glyphicons! It is a nice way to beautify the websites icons without having to do it from Photoshop as an image which may cause responsive issues.
They are usually used by calling the class name which links to the CSS that call the icons from the font family. Html lang would not have any issues with it.
Many websites use "icon fonts". But yes, assigning language letters to them would be wrong. It would be best to assign an arrow icon to the Unicode character code for a similar arrow. Another option would be to use the Private Use Area of Unicode. In this case, if your font fails to load for any reason, you won't have a good fallback strategy. But if you choose meaningful char codes for your icons, you would.
Many people are in favor of using SVGs over icon fonts. But there are pros and cons to both icon fonts and SVGs. I think that it's great that as web developers, we get to choose among different implementations or solutions to the same problem.
To answer your question, I would say that if done right, there is nothing wrong with using fonts for implementing icons.
As Mike 'Pomax' Kamermans put it:
"Fonts are for encoding vector graphics that are to be used in
typesetting context. That can mean letters, or icons, or emoji"
One big reason is accessibility. There are many browser extensions which swap out a website's font for one that's more legible for people with different visual impairments. If you use fonts for your icons, these will be swapped out too, leaving your user looking at whatever string you placed in for your icons.
I'm wondering if it's a bad idea to use weird characters in my code. I recently tried using them to create little dots to indicate which slide you're on and to change slides easily:
There are tons of these types of characters, and it seems like they could be used in place of icons/images in many cases, they are style-able and scale-able, and screen readers would be able to make sense of them.
But, I don't see anyone doing this, and I've got a feeling this is a bad idea, I just can't decide why. I guess it seems too easy to be true. Could someone tell me why this is or isn't okay? Here are some more examples of the characters i'm talking about:
↖ ↗ ↙ ↘ ㊣ ◎ ○ ● ⊕ ⊙ ○ △ ▲ ☆ ★ ◇ ◆ ■ □ ▽ ▼ § ￥ 〒 ￠ ￡ ※ ♀ ♂ &⁂ ℡ ↂ░ ▣ ▤ ▥ ▦ ▧ ✐✌✍✡✓✔✕✖ ♂ ♀ ♥ ♡ ☜ ☞ ☎ ☏ ⊙ ◎ ☺ ☻ ► ◄ ▧ ▨ ♨ ◐ ◑ ↔ ↕ ♥ ♡ ▪ ▫ ☼ ♦ ▀ ▄ █ ▌ ▐ ░ ▒ ▬ ♦ ◊
PS: I would also welcome general information about these characters, what they're called and stuff (ASCII, Unicode)?
There are three things to deal with:
1. As characters in a sentence/text:
The problem is that some fonts simply do not have them. However since CSS can control font use you probably will not run into this problem. As long as you use a web safe font, and know that that character is available in that font, you should probably be okay.
You can also use an embedded font, though be sure to fall back on a web safe font that contains the character you need as many browser will not support embedded fonts.
However sometimes certain devices will not have multiple fonts to choose from. If that font does not support your character you will run into problems. However depending on what your site does and the audience you are targeting this may not be a problem for you. Not to mention that devices like that are very old, and uncommon.
All in all it was probably not a good idea a handful of years ago, but now you are not likely to have problems as long as you cover all your bases.
It is important however to point out that you should never hard code those characters, instead use HTML entities. Just inserting those characters into your code can lead to unpredictable results. I recently copied some text from Word directly into my code, Word used smart quotes (quote marks that curve inwards properly). They showed up fine in Notepad++, but when I viewed the page I did not get quotes, I got some weird symbol.
I could have either replaced them with normal quotes " or with HTML entities to keep the style “ and ” (“ and ”).
Any Unicode character can be inserted this way (even those without special names).
Wikipedia has a good reference:
http://en.wikipedia.org/wiki/List_of_XML_and_HTML_character_entity_references
2. As UI elements:
While it may be safe to use them in many cases, it is still better to use HTML elements where possible. You could simply style some div elements to be round and filled/not filled for your example.
As far as design goes they are really limiting, finding one that fits with the style of your page can be a hassle, and may mean that you will definitely need to embed a font, which is still only supported by the latest browsers.
Plus many devices do not support heavy font manipulation, and will often display them poorly. It works in the flow of your text, but as a vital part of the UI there can be major problems. Any possible issue one of those characters can bring will be multiplied by the fact that it is part of your UI.
From an artistic stand point they simply limit your abilities too much.
3. What are you doing?
Finaly you need to consider this:
Text is for telling
Image is for showing
HTML is for organizing
CSS is for making things look good while you show them
JavaScript is for functionality
Those characters are text, they are for telling someone something. So ask the question: "What am I doing?" and then use what was designed for that task. If you are telling use them, if you are showing use Image, or CSS.
I've seen this done before (the stars) and I think it's an awesome idea! It's also becoming quite popular to use a font (with #font-face) full of icons, like this one: http://fortawesome.github.com/Font-Awesome/
I can't see any downside to using a font like "font awesome" (only the upsides you mention like scalabilty and the ability to change color with CSS). Perhaps there's a downside to using the special characters you mention but none that I know of.
The problem with using those characters is that not all of them are available in all fonts used by all users, which means your application may look strange, or in the worst case be unusable. That said, it is becoming more common to assume the characters available in certain common fonts (Apple/Microsoft's Arial, Bitstream Vera). You can't even assume that you can download a font, as some users may capture content for offline reading with a service like Instapaper or Read It Later.
There are a number of problems:
Portability: using anything other than the 7-bit ASCII characters in code can make your code less portable, as recipients may use the wrong encoding. You can do a lot to mitigate this (eg. use UTF16 or at least UTF-8 encoded files). Most languages allow you to specify strings in characters using some form of escape notation (eg. "\u1234" in C#), which will avoid the problem, but loses some of the advantages.
Font-dependency: user interface elements that depend on special characters being available in a font may be harder to internationalize, since those glyphs might not be in the font that you want/need to use for a particular audience.
No color, limited choice of art: while font glyphs might seem useful to a coder, they probably look pretty poor to a UI designer.
The question is very broad; it could be split to literally thousands of questions of the type “why shouldn’t I use character ... in HTML documents?” This seems to be what the question is about—not really about code. And it’s about characters, seen as “weird” or “uncommon” or “special” from some perspective, not about character encodings. (None of the characters mentioned are encoded in ASCII. Some are encoded in ISO-8895-1. All are encoded in Unicode.)
The characters are used in HTML documents. There is no general reason against not using them, but loads of specific reasons why some specific characters might not be the best approach in a specific situation.
For example, the “little dots” you mention in your example (probably not dots at all but circles or bullets), when used as control elements as you describe, would mean poor usability and poor accessibility. Making them significantly larger would improve the situation, but this more or less proves that such text characters are not suitable for controls.
Screen readers could make sense of special characters if they used a database of various properties of characters. Well, they don’t, and they often fail to read properly even the most common special characters. Just reading the Unicode name of a character can be cryptic or outright misleading. The proper reading would generally depend on meaning and context.
The main issue, however, is that people do not generally recognize characters in the meanings that you would assign to them. How many people know what the circled plus symbol “⊕” stands for? Maybe 1 out of 1,000, optimistically thinking. It might be all right to use in on a page about advanced mathematics or physics, especially if the notation is defined there. But used in general text, it would be just… a weird character, and people would read different meanings into it, or just get puzzled.
So using special characters just because they look cool isn’t a good idea. Even when there is time and place for a special character, there are technical issues with them. How many fonts do you expect to contain “⊕”? How many of those fonts do you expect Joe Q. Public to have in his computer? In this specific case, you would find the font coverage reasonably good, but you would still have to analyze it and write a longish list of font names in your CSS code to cover most platforms. In the pile of poo case (♨), it would be unrealistic to expect most people to see anything but a symbol for unrepresentable character. Regarding the methods of finding out such things, check out my Guide to using special characters in HTML.
I've run into problems using unusual characters: the tools editor, compiler, interpreter etc.) often complain and report errors. In the end, it wasn't worth the hassle. Darn western hegemony, or homogeneity, or, well, something!
I see some websites use fancy icons. An example is this curly leaf icon at Dive Into HTML5:
❧
I would like to know the formal name of these icons. I've tried searching "html entities" and "html glyphs" in Google but the results always show the boring old arrows and Greek characters.
This is a unicode character.
If you are using UTF-8 encoding - which the example site is - you have a very large range of characters you can use - not just arrows and greek characters.
I'd encourage you to use Firebug or some other developer inspector in your browser. It makes it very easy to reverse engineer things like this.
These don't have any particular name as far as I know, they are simply Unicode glyphs. Here is a table.
I would like to print some kind of ASCII "art" on a web page in pre-tags. These graphics use DOS characters to show a map like old maze games did. I didn't find anything in the HTML special character reference. Is there a way to use these characters in HTML ?
Thanks in advance.
With the right Unicode characters, the old character encodings shouldn't make much odds. The tricky bit may be converting existing ASCII art into Unicode - at which point you need to know the original encoding.
The relevant code charts will be listed on the Unicode "symbols" charts page. In particular, I suspect you'll find the box drawing and block elements charts useful.
You'll need to make sure that your page uses a font which contains the right characters, of course...
As an example, you can render this:
┌┐
└┘
With:
<pre>┌┐
└┘</pre>
Not quite a proper box, but getting there...
You can send them in the <pre> tags, although in XHTML you'll need to encapsulate it in <![CDATA[[]> I think. Be careful though, not all encodings render this correctly. For example, a lot of ASCII art designed for DOS code page 430 (US) fails over here in the UK (830). Eastern Europe suffers especially.
I think the best approach here would be to render images.
EDIT: Oh. You could try , but I'm not sure if that would work.