"’" showing on page instead of " ' " - encoding

’ is showing on my page instead of '.
I have the Content-Type set to UTF-8 in both my <head> tag and my HTTP headers:
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8" />
In addition, my browser is set to Unicode (UTF-8):
So what's the problem, and how can I fix it?


So what's the problem,
It's a ’ (RIGHT SINGLE QUOTATION MARK - U+2019) character which is being decoded as CP-1252 instead of UTF-8. If you check the encodings table, then you see that this character is in UTF-8 composed of bytes 0xE2, 0x80 and 0x99. If you check the CP-1252 code page layout, then you'll see that each of those bytes stand for the individual characters â, € and ™.
and how can I fix it?
Use UTF-8 instead of CP-1252 to read, write, store, and display the characters.
I have the Content-Type set to UTF-8 in both my <head> tag and my HTTP headers:
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8" />
This only instructs the client which encoding to use to interpret and display the characters. This doesn't instruct your own program which encoding to use to read, write, store, and display the characters in. The exact answer depends on the server side platform / database / programming language used. Do note that the one set in HTTP response header has precedence over the HTML meta tag. The HTML meta tag would only be used when the page is opened from local disk file system instead of from HTTP.
In addition, my browser is set to Unicode (UTF-8):
This only forces the client which encoding to use to interpret and display the characters. But the actual problem is that you're already sending ’ (encoded in UTF-8) to the client instead of ’. The client is correctly displaying ’ using the UTF-8 encoding. If the client was misinstructed to use, for example ISO-8859-1, you would likely have seen ââ¬â¢ instead.
I am using ASP.NET 2.0 with a database.
This is most likely where your problem lies. You need to verify with an independent database tool what the data looks like.
If the ’ character is there, then you aren't connecting to the database correctly. You need to tell the database connector to use UTF-8.
If your database contains ’, then it's your database that's messed up. Most probably the tables aren't configured to use UTF-8. Instead, they use the database's default encoding, which varies depending on the configuration. If this is your issue, then usually just altering the table to use UTF-8 is sufficient. If your database doesn't support that, you'll need to recreate the tables. It is good practice to set the encoding of the table when you create it.
You're most likely using SQL Server, but here is some MySQL code (copied from this article):
CREATE DATABASE db_name CHARACTER SET utf8;
CREATE TABLE tbl_name (...) CHARACTER SET utf8;
If your table is however already UTF-8, then you need to take a step back. Who or what put the data there. That's where the problem is. One example would be HTML form submitted values which are incorrectly encoded/decoded.
Here are some more links to learn more about the problem:
The Absolute Minimum Every Software Developer Absolutely, Positively Must Know About Unicode and Character Sets (No Excuses!), from our own Joel.
Unicode - How to get the characters right?, with more concise and practical information, solutions are targeted on Java environments.
How to setup your PHP site to use UTF8, targeted on PHP environments.


Ensure the browser and editor are using UTF-8 encoding instead of ISO-8859-1/Windows-1252.
Or use ’.


’ (Unicode codepoint U+2019 RIGHT SINGLE QUOTATION MARK) is encoded in UTF-8 as bytes:
0xE2 0x80 0x99.
’ (Unicode codepoints U+00E2 U+20AC U+2122) is encoded in UTF-8 as bytes:
0xC3 0xA2 0xE2 0x82 0xAC 0xE2 0x84 0xA2.
These are the bytes your browser is actually receiving in order to produce ’ when processed as UTF-8.
That means that your source data is going through two charset conversions before being sent to the browser:
The source ’ character (U+2019) is first encoded as UTF-8 bytes:
0xE2 0x80 0x99
those individual bytes were then being mis-interpreted and decoded to Unicode codepoints U+00E2 U+20AC U+2122 by one of the Windows-125X charsets (1252, 1254, 1256, and 1258 all map 0xE2 0x80 0x99 to U+00E2 U+20AC U+2122), and then those codepoints are being encoded as UTF-8 bytes:
0xE2 -> U+00E2 -> 0xC3 0xA2
0x80 -> U+20AC -> 0xE2 0x82 0xAC
0x99 -> U+2122 -> 0xE2 0x84 0xA2
You need to find where the extra conversion in step 2 is being performed and remove it.


This sometimes happens when a string is converted from Windows-1252 to UTF-8 twice.
We had this in a Zend/PHP/MySQL application where characters like that were appearing in the database, probably due to the MySQL connection not specifying the correct character set. We had to:
Ensure Zend and PHP were communicating with the database in UTF-8 (was not by default)
Repair the broken characters with several SQL queries like this...
UPDATE MyTable SET
MyField1 = CONVERT(CAST(CONVERT(MyField1 USING latin1) AS BINARY) USING utf8),
MyField2 = CONVERT(CAST(CONVERT(MyField2 USING latin1) AS BINARY) USING utf8);
Do this for as many tables/columns as necessary.
You can also fix some of these strings in PHP if necessary. Note that because characters have been encoded twice, we actually need to do a reverse conversion from UTF-8 back to Windows-1252, which confused me at first.
mb_convert_encoding('’', 'Windows-1252', 'UTF-8'); // returns ’


I have some documents where … was showing as … and ê was showing as ê. This is how it got there (python code):
# Adam edits original file using windows-1252
windows = '\x85\xea'
# that is HORIZONTAL ELLIPSIS, LATIN SMALL LETTER E WITH CIRCUMFLEX
# Beth reads it correctly as windows-1252 and writes it as utf-8
utf8 = windows.decode("windows-1252").encode("utf-8")
print(utf8)
# Charlie reads it *incorrectly* as windows-1252 writes a twingled utf-8 version
twingled = utf8.decode("windows-1252").encode("utf-8")
print(twingled)
# detwingle by reading as utf-8 and writing as windows-1252 (it's really utf-8)
detwingled = twingled.decode("utf-8").encode("windows-1252")
assert utf8==detwingled
To fix the problem, I used python code like this:
with open("dirty.html","rb") as f:
dt = f.read()
ct = dt.decode("utf8").encode("windows-1252")
with open("clean.html","wb") as g:
g.write(ct)
(Because someone had inserted the twingled version into a correct UTF-8 document, I actually had to extract only the twingled part, detwingle it and insert it back in. I used BeautifulSoup for this.)
It is far more likely that you have a Charlie in content creation than that the web server configuration is wrong. You can also force your web browser to twingle the page by selecting windows-1252 encoding for a utf-8 document. Your web browser cannot detwingle the document that Charlie saved.
Note: the same problem can happen with any other single-byte code page (e.g. latin-1) instead of windows-1252.


You have a mismatch in your character encoding; your string is encoded in one encoding (UTF-8) and whatever is interpreting this page is using another (say ASCII).
Always specify your encoding in your http headers and make sure this matches your framework's definition of encoding.
Sample http header:
Content-Type text/html; charset=utf-8
Setting encoding in asp.net
<configuration>
<system.web>
<globalization
fileEncoding="utf-8"
requestEncoding="utf-8"
responseEncoding="utf-8"
culture="en-US"
uiCulture="de-DE"
/>
</system.web>
</configuration>
Setting encoding in jsp


If your content type is already UTF8 , then it is likely the data is already arriving in the wrong encoding. If you are getting the data from a database, make sure the database connection uses UTF-8.
If this is data from a file, make sure the file is encoded correctly as UTF-8. You can usually set this in the "Save as..." Dialog of the editor of your choice.
If the data is already broken when you view it in the source file, chances are that it used to be a UTF-8 file but was saved in the wrong encoding somewhere along the way.


If someone gets this error on WordPress website, you need to change wp-config db charset:
define('DB_CHARSET', 'utf8mb4_unicode_ci');
instead of:
define('DB_CHARSET', 'utf8mb4');


If the other answers haven't helped, you might want to check whether your database is actually storing the mojibake characters. I was viewing the text in utf-8, but I was still seeing the mojibake and it turned out that, due to a database upgrade, the text had been permanently "mojibaked".
In this case, one option is to "fix" the text with Python's ftfy package (or JavaScript verion here).


You must have copy/paste text from Word Document. Word document use Smart Quotes. You can replace it with Special Character (’) or simply type in your HTML editor (').
I'm sure this will solve your problem.


In DBeaver (or other editors) the script file you're working can prompt to save as UTF8 and that will change the char:
–
into
–
or
–


The same thing happened to me with the '–' character (long minus sign).
I used this simple replace so resolve it:
htmlText = htmlText.Replace('–', '-');

Related

What does this decode to, and is it UTF? Игорќ

I have received this in a name field (so it should be a person's name)
Игорќ
What could that decode to? Is it UTF-8? What language does that translate to? Russian?
If you can give me a hint or maybe links to websites that explain what meaningful letters I should get out of that would be helpful, thank you :)

This typically is UTF-8 interpreted as some single-byte Windows encoding.
String s = "Игорќ"; // Source encoding UTF-8
byte[] b = s.getBytes("Cp1252");
System.out.println("" + new String(b, StandardCharsets.UTF_8));
// Игорќ
The data might easily get corrupted. Above I got some results with Windows-1252 (MS Windows Latin-1). The java source must be compiled with encoding UTF-8 to accept those chars.

Since you already pasted the original code into a UTF-8 encoded site as Stack Overflow your code is now corrupt data perfectly encoded as UTF-8. If you want to ask yourself anything about the data encoding you need to use an hexadecimal editor or a similar tool on the original raw bytes.
In any case, if you do this:
Open a text file in some single-byte encoding (possibly the ANSI code page used by your copy of Windows, I used Windows-1252)
Paste the Игорќ gibberish and save the file
Reload the file as UTF-8
... you get this:
Игорќ
So it's probably valid UTF-8 incorrectly decoded.

Encoding problems in ASP when using English and Chinese characters

I am having problems with encoding Chinese in an ASP site. The file formats are:
translations.txt - UTF-8 (to store my translations)
test.asp - UTF-8 - (to render the page)
test.asp is reading translations.txt that contains the following data:
Help|ZH|帮助
Home|ZH|首页
The test.asp splits on the pipe delimiter and if the user contains a cookie with ZH, it will display this translation, else it will just revert back to the Key value.
Now, I have tried the following things, which have not worked:
Add a meta tag
<meta http-equiv="Content-Type" content="text/html;charset=utf-8"/>
Set the Response.CharSet = "UTF-8"
Set the Response.ContentType = "text/html"
Set the Session.CodePage (and Response) to both 65001 (UTF-8)
I have confirmed that the text in translations.txt is definitely in UTF-8 and has no byte order mark
The browser is picking up that the page is Unicode UTF-8, but the page is displaying gobbledegook.
The Scripting.OpenTextFile(<file>,<create>,<iomode>,<encoding>) method returns the same incorrect text regardless of the Encoding parameter.
Here is a sample of what I want to be displayed in China (ZH):
首页
帮助
But the following is displayed:
首页
帮助
This occurs all tested browsers - Google Chrome, IE 7/8, and Firefox 4. The font definitely has a Chinese branch of glyphs. Also, I do have Eastern languages installed.
--
I have tried pasting in the original value into the HTML, which did work (but note this is a hard coded value).
首页
首页
However, this is odd.
首页 --(in hex)--> E9 A6 96 E9 A1 --(as chars)--> 首页
Any ideas what I am missing?

In order to read the UTF-8 file, you'll probably need to use the ADODB.Stream object. I don't claim to be an expert on character encoding, but this test worked for me:
test.txt (saved as UTF-8 without BOM):
首页
帮助
test.vbs
Option Explicit
Const adTypeText = 2
Const adReadLine = -2
Dim stream : Set stream = CreateObject("ADODB.Stream")
stream.Open
stream.Type = adTypeText
stream.Charset = "UTF-8"
stream.LoadFromFile "test.txt"
Do Until stream.EOS
WScript.Echo stream.ReadText(adReadLine)
Loop
stream.Close

Whatever part of the process is reading the translations.txt file does not seem to understand that the file is in UTF-8. It looks like it is reading it in as some other encoding. You should specify encoding in whatever process is opening and reading that file. This will be different from the encoding of your web page.
Inserting the byte order mark at the beginning of that file may also be a solution.

Scripting.OpenTextFile does not understand UTF-8 at all. It can only read the current OEM encoding or Unicode. As you can see from the number of bytes being used for some character sets UTF-8 is quite inefficient. I would recommend Unicode for this sort of data.
You should save the file as Unicode (in Windows parlance) and then open with:
Dim stream : Set stream = Scripting.OpenTextFile(yourFilePath, 1, false, -1)

Just use the script below at the top of your page
Response.CodePage=65001
Response.CharSet="UTF-8"

How did SourceForge maim this Unicode character?

A little encoding puzzle for you.
A comment on a SourceForge tracker item contains the character U+2014, EM DASH, which is rendered by the web interface as — like it should.
In the XML export, however, it shows up as:
—
Decoding the entities, that results in these code points:
U+00E2 U+20AC U+201D
I.e. the characters —. The XML should have been —, the decimal representation of 0x2014, so this is probably a bug in the SF.net exporter.
Now I'm looking to reverse the process, but I can't find a way to get the above output from this Unicode character, no matter what erroneous encoding/decoding sequence I try. Any idea what happened here and how to reverse the process?

The the XML output is incorrectly been encoded using CP1252. To revert this, convert — to bytes using CP1252 encoding and then convert those bytes back to string/char using UTF-8 encoding.
Java based evidence:
String s = "—";
System.out.println(new String(s.getBytes("CP1252"), "UTF-8")); // —
Note that this assumes that the stdout console uses by itself UTF-8 to display the character.

In .Net, Encoding.UTF8.GetString(Encoding.GetEncoding(1252).GetBytes("—")) returns —.
SourceForge converted it to UTF8, interpreted the each of the bytes as characters in CP1252, then saved the characters as three separate entities using the actual Unicode codepoints for those characters.

IWebBrowser: How to specify the encoding when loading html from a stream?

Using the concepts from the sample code provided by Microsoft for loading HTML content into an IWebBrowser from an IStream using the web browser's IPersistStreamInit interface:
pseudocode:
void LoadWebBrowserFromStream(IWebBrowser webBrowser, IStream stream)
{
IPersistStreamInit persist = webBrowser.Document as IPersistStreamInit;
persist.Load(stream);
}
How can one specify the encoding of the html inside the IStream? The IStream will contain a series of bytes, but the problem is what do those bytes represent? They could, for example, contain bytes where:
each byte represents a character from the current Windows code-page (e.g. 1252)
each byte could represent a character from the ISO-8859-1 character set
the bytes could represent UTF-8 encoded characters
every 2 bytes could represent a character, using UTF-16 encoding
In my particular case, i am providing the IWebBrowser an IStream that contains a series of double-bytes characters (UTF-16), but the browser (incorrectly) believes that UTF-8 encoding is in effect. This results in garbled characters.
Workaround solution
While the question asks how to specify the encoding, in my particular case, with only UTF-16 encoding, there's a simple workaround. Adding the 0xFEFF Byte Order Mark (BOM) indicates that the text is UTF-16 unicode. ie then uses the proper encoding and shows the text properly.
Of course that wouldn't work if the text were encoded, for example with:
UCS-2
UCS-4
ISO-10646-UCS-2
UNICODE-1-1-UTF-8
UNICODE-2-0-UTF-16
UNICODE-2-0-UTF-8
US-ASCII
ISO-8859-1
ISO-8859-2
ISO-8859-3
ISO-8859-4
ISO-8859-5
ISO-8859-6
ISO-8859-7
ISO-8859-8
ISO-8859-9
WINDOWS-1250
WINDOWS-1251
WINDOWS-1252
WINDOWS-1253
WINDOWS-1254
WINDOWS-1255
WINDOWS-1256
WINDOWS-1257
WINDOWS-1258

IE's document supports IPersistMoniker loading too. IE uses URL monikers for downloading. You can replace the url moniker created by CreateURLMonikerEx with your own moniker. A few details about URL moniker's implementation can be find here. See if you can get IHTTPNegotiate from the binding context when your BindToStroage implemetation is called.

What's the difference between UTF-8 and UTF-8 with BOM?

What's different between UTF-8 and UTF-8 with BOM? Which is better?

The UTF-8 BOM is a sequence of bytes at the start of a text stream (0xEF, 0xBB, 0xBF) that allows the reader to more reliably guess a file as being encoded in UTF-8.
Normally, the BOM is used to signal the endianness of an encoding, but since endianness is irrelevant to UTF-8, the BOM is unnecessary.
According to the Unicode standard, the BOM for UTF-8 files is not recommended:
2.6 Encoding Schemes
... Use of a BOM is neither required nor recommended for UTF-8, but may be encountered in contexts where UTF-8 data is converted from other encoding forms that use a BOM or where the BOM is used as a UTF-8 signature. See the “Byte Order Mark” subsection in Section 16.8, Specials, for more information.

The other excellent answers already answered that:
There is no official difference between UTF-8 and BOM-ed UTF-8
A BOM-ed UTF-8 string will start with the three following bytes. EF BB BF
Those bytes, if present, must be ignored when extracting the string from the file/stream.
But, as additional information to this, the BOM for UTF-8 could be a good way to "smell" if a string was encoded in UTF-8... Or it could be a legitimate string in any other encoding...
For example, the data [EF BB BF 41 42 43] could either be:
The legitimate ISO-8859-1 string "ABC"
The legitimate UTF-8 string "ABC"
So while it can be cool to recognize the encoding of a file content by looking at the first bytes, you should not rely on this, as show by the example above
Encodings should be known, not divined.

There are at least three problems with putting a BOM in UTF-8 encoded files.
Files that hold no text are no longer empty because they always contain the BOM.
Files that hold text within the ASCII subset of UTF-8 are no longer themselves ASCII because the BOM is not ASCII, which makes some existing tools break down, and it can be impossible for users to replace such legacy tools.
It is not possible to concatenate several files together because each file now has a BOM at the beginning.
And, as others have mentioned, it is neither sufficient nor necessary to have a BOM to detect that something is UTF-8:
It is not sufficient because an arbitrary byte sequence can happen to start with the exact sequence that constitutes the BOM.
It is not necessary because you can just read the bytes as if they were UTF-8; if that succeeds, it is, by definition, valid UTF-8.

Here are examples of the BOM usage that actually cause real problems and yet many people don't know about it.
BOM breaks scripts
Shell scripts, Perl scripts, Python scripts, Ruby scripts, Node.js scripts or any other executable that needs to be run by an interpreter - all start with a shebang line which looks like one of those:
#!/bin/sh
#!/usr/bin/python
#!/usr/local/bin/perl
#!/usr/bin/env node
It tells the system which interpreter needs to be run when invoking such a script. If the script is encoded in UTF-8, one may be tempted to include a BOM at the beginning. But actually the "#!" characters are not just characters. They are in fact a magic number that happens to be composed out of two ASCII characters. If you put something (like a BOM) before those characters, then the file will look like it had a different magic number and that can lead to problems.
See Wikipedia, article: Shebang, section: Magic number:
The shebang characters are represented by the same two bytes in
extended ASCII encodings, including UTF-8, which is commonly used for
scripts and other text files on current Unix-like systems. However,
UTF-8 files may begin with the optional byte order mark (BOM); if the
"exec" function specifically detects the bytes 0x23 and 0x21, then the
presence of the BOM (0xEF 0xBB 0xBF) before the shebang will prevent
the script interpreter from being executed. Some authorities recommend
against using the byte order mark in POSIX (Unix-like) scripts,[14]
for this reason and for wider interoperability and philosophical
concerns. Additionally, a byte order mark is not necessary in UTF-8,
as that encoding does not have endianness issues; it serves only to
identify the encoding as UTF-8. [emphasis added]
BOM is illegal in JSON
See RFC 7159, Section 8.1:
Implementations MUST NOT add a byte order mark to the beginning of a JSON text.
BOM is redundant in JSON
Not only it is illegal in JSON, it is also not needed to determine the character encoding because there are more reliable ways to unambiguously determine both the character encoding and endianness used in any JSON stream (see this answer for details).
BOM breaks JSON parsers
Not only it is illegal in JSON and not needed, it actually breaks all software that determine the encoding using the method presented in RFC 4627:
Determining the encoding and endianness of JSON, examining the first four bytes for the NUL byte:
00 00 00 xx - UTF-32BE
00 xx 00 xx - UTF-16BE
xx 00 00 00 - UTF-32LE
xx 00 xx 00 - UTF-16LE
xx xx xx xx - UTF-8
Now, if the file starts with BOM it will look like this:
00 00 FE FF - UTF-32BE
FE FF 00 xx - UTF-16BE
FF FE 00 00 - UTF-32LE
FF FE xx 00 - UTF-16LE
EF BB BF xx - UTF-8
Note that:
UTF-32BE doesn't start with three NULs, so it won't be recognized
UTF-32LE the first byte is not followed by three NULs, so it won't be recognized
UTF-16BE has only one NUL in the first four bytes, so it won't be recognized
UTF-16LE has only one NUL in the first four bytes, so it won't be recognized
Depending on the implementation, all of those may be interpreted incorrectly as UTF-8 and then misinterpreted or rejected as invalid UTF-8, or not recognized at all.
Additionally, if the implementation tests for valid JSON as I recommend, it will reject even the input that is indeed encoded as UTF-8, because it doesn't start with an ASCII character < 128 as it should according to the RFC.
Other data formats
BOM in JSON is not needed, is illegal and breaks software that works correctly according to the RFC. It should be a nobrainer to just not use it then and yet, there are always people who insist on breaking JSON by using BOMs, comments, different quoting rules or different data types. Of course anyone is free to use things like BOMs or anything else if you need it - just don't call it JSON then.
For other data formats than JSON, take a look at how it really looks like. If the only encodings are UTF-* and the first character must be an ASCII character lower than 128 then you already have all the information needed to determine both the encoding and the endianness of your data. Adding BOMs even as an optional feature would only make it more complicated and error prone.
Other uses of BOM
As for the uses outside of JSON or scripts, I think there are already very good answers here. I wanted to add more detailed info specifically about scripting and serialization, because it is an example of BOM characters causing real problems.

What's different between UTF-8 and UTF-8 without BOM?
Short answer: In UTF-8, a BOM is encoded as the bytes EF BB BF at the beginning of the file.
Long answer:
Originally, it was expected that Unicode would be encoded in UTF-16/UCS-2. The BOM was designed for this encoding form. When you have 2-byte code units, it's necessary to indicate which order those two bytes are in, and a common convention for doing this is to include the character U+FEFF as a "Byte Order Mark" at the beginning of the data. The character U+FFFE is permanently unassigned so that its presence can be used to detect the wrong byte order.
UTF-8 has the same byte order regardless of platform endianness, so a byte order mark isn't needed. However, it may occur (as the byte sequence EF BB FF) in data that was converted to UTF-8 from UTF-16, or as a "signature" to indicate that the data is UTF-8.
Which is better?
Without. As Martin Cote answered, the Unicode standard does not recommend it. It causes problems with non-BOM-aware software.
A better way to detect whether a file is UTF-8 is to perform a validity check. UTF-8 has strict rules about what byte sequences are valid, so the probability of a false positive is negligible. If a byte sequence looks like UTF-8, it probably is.

UTF-8 with BOM is better identified. I have reached this conclusion the hard way. I am working on a project where one of the results is a CSV file, including Unicode characters.
If the CSV file is saved without a BOM, Excel thinks it's ANSI and shows gibberish. Once you add "EF BB BF" at the front (for example, by re-saving it using Notepad with UTF-8; or Notepad++ with UTF-8 with BOM), Excel opens it fine.
Prepending the BOM character to Unicode text files is recommended by RFC 3629: "UTF-8, a transformation format of ISO 10646", November 2003
at https://www.rfc-editor.org/rfc/rfc3629 (this last info found at: http://www.herongyang.com/Unicode/Notepad-Byte-Order-Mark-BOM-FEFF-EFBBBF.html)

BOM tends to boom (no pun intended (sic)) somewhere, someplace. And when it booms (for example, doesn't get recognized by browsers, editors, etc.), it shows up as the weird characters  at the start of the document (for example, HTML file, JSON response, RSS, etc.) and causes the kind of embarrassments like the recent encoding issue experienced during the talk of Obama on Twitter.
It's very annoying when it shows up at places hard to debug or when testing is neglected. So it's best to avoid it unless you must use it.

Question: What's different between UTF-8 and UTF-8 without a BOM? Which is better?
Here are some excerpts from the Wikipedia article on the byte order mark (BOM) that I believe offer a solid answer to this question.
On the meaning of the BOM and UTF-8:
The Unicode Standard permits the BOM in UTF-8, but does not require
or recommend its use. Byte order has no meaning in UTF-8, so its
only use in UTF-8 is to signal at the start that the text stream is
encoded in UTF-8.
Argument for NOT using a BOM:
The primary motivation for not using a BOM is backwards-compatibility
with software that is not Unicode-aware... Another motivation for not
using a BOM is to encourage UTF-8 as the "default" encoding.
Argument FOR using a BOM:
The argument for using a BOM is that without it, heuristic analysis is
required to determine what character encoding a file is using.
Historically such analysis, to distinguish various 8-bit encodings, is
complicated, error-prone, and sometimes slow. A number of libraries
are available to ease the task, such as Mozilla Universal Charset
Detector and International Components for Unicode.
Programmers mistakenly assume that detection of UTF-8 is equally
difficult (it is not because of the vast majority of byte sequences
are invalid UTF-8, while the encodings these libraries are trying to
distinguish allow all possible byte sequences). Therefore not all
Unicode-aware programs perform such an analysis and instead rely on
the BOM.
In particular, Microsoft compilers and interpreters, and many
pieces of software on Microsoft Windows such as Notepad will not
correctly read UTF-8 text unless it has only ASCII characters or it
starts with the BOM, and will add a BOM to the start when saving text
as UTF-8. Google Docs will add a BOM when a Microsoft Word document is
downloaded as a plain text file.
On which is better, WITH or WITHOUT the BOM:
The IETF recommends that if a protocol either (a) always uses UTF-8,
or (b) has some other way to indicate what encoding is being used,
then it “SHOULD forbid use of U+FEFF as a signature.”
My Conclusion:
Use the BOM only if compatibility with a software application is absolutely essential.
Also note that while the referenced Wikipedia article indicates that many Microsoft applications rely on the BOM to correctly detect UTF-8, this is not the case for all Microsoft applications. For example, as pointed out by #barlop, when using the Windows Command Prompt with UTF-8†, commands such type and more do not expect the BOM to be present. If the BOM is present, it can be problematic as it is for other applications.
† The chcp command offers support for UTF-8 (without the BOM) via code page 65001.

This question already has a million-and-one answers and many of them are quite good, but I wanted to try and clarify when a BOM should or should not be used.
As mentioned, any use of the UTF BOM (Byte Order Mark) in determining whether a string is UTF-8 or not is educated guesswork. If there is proper metadata available (like charset="utf-8"), then you already know what you're supposed to be using, but otherwise you'll need to test and make some assumptions. This involves checking whether the file a string comes from begins with the hexadecimal byte code, EF BB BF.
If a byte code corresponding to the UTF-8 BOM is found, the probability is high enough to assume it's UTF-8 and you can go from there. When forced to make this guess, however, additional error checking while reading would still be a good idea in case something comes up garbled. You should only assume a BOM is not UTF-8 (i.e. latin-1 or ANSI) if the input definitely shouldn't be UTF-8 based on its source. If there is no BOM, however, you can simply determine whether it's supposed to be UTF-8 by validating against the encoding.
Why is a BOM not recommended?
Non-Unicode-aware or poorly compliant software may assume it's latin-1 or ANSI and won't strip the BOM from the string, which can obviously cause issues.
It's not really needed (just check if the contents are compliant and always use UTF-8 as the fallback when no compliant encoding can be found)
When should you encode with a BOM?
If you're unable to record the metadata in any other way (through a charset tag or file system meta), and the programs being used like BOMs, you should encode with a BOM. This is especially true on Windows where anything without a BOM is generally assumed to be using a legacy code page. The BOM tells programs like Office that, yes, the text in this file is Unicode; here's the encoding used.
When it comes down to it, the only files I ever really have problems with are CSV. Depending on the program, it either must, or must not have a BOM. For example, if you're using Excel 2007+ on Windows, it must be encoded with a BOM if you want to open it smoothly and not have to resort to importing the data.

UTF-8 without BOM has no BOM, which doesn't make it any better than UTF-8 with BOM, except when the consumer of the file needs to know (or would benefit from knowing) whether the file is UTF-8-encoded or not.
The BOM is usually useful to determine the endianness of the encoding, which is not required for most use cases.
Also, the BOM can be unnecessary noise/pain for those consumers that don't know or care about it, and can result in user confusion.

It should be noted that for some files you must not have the BOM even on Windows. Examples are SQL*plus or VBScript files. In case such files contains a BOM you get an error when you try to execute them.

Quoted at the bottom of the Wikipedia page on BOM: http://en.wikipedia.org/wiki/Byte-order_mark#cite_note-2
"Use of a BOM is neither required nor recommended for UTF-8, but may be encountered in contexts where UTF-8 data is converted from other encoding forms that use a BOM or where the BOM is used as a UTF-8 signature"

UTF-8 with BOM only helps if the file actually contains some non-ASCII characters. If it is included and there aren't any, then it will possibly break older applications that would have otherwise interpreted the file as plain ASCII. These applications will definitely fail when they come across a non ASCII character, so in my opinion the BOM should only be added when the file can, and should, no longer be interpreted as plain ASCII.
I want to make it clear that I prefer to not have the BOM at all. Add it in if some old rubbish breaks without it, and replacing that legacy application is not feasible.
Don't make anything expect a BOM for UTF-8.

I look at this from a different perspective. I think UTF-8 with BOM is better as it provides more information about the file. I use UTF-8 without BOM only if I face problems.
I am using multiple languages (even Cyrillic) on my pages for a long time and when the files are saved without BOM and I re-open them for editing with an editor (as cherouvim also noted), some characters are corrupted.
Note that Windows' classic Notepad automatically saves files with a BOM when you try to save a newly created file with UTF-8 encoding.
I personally save server side scripting files (.asp, .ini, .aspx) with BOM and .html files without BOM.

When you want to display information encoded in UTF-8 you may not face problems. Declare for example an HTML document as UTF-8 and you will have everything displayed in your browser that is contained in the body of the document.
But this is not the case when we have text, CSV and XML files, either on Windows or Linux.
For example, a text file in Windows or Linux, one of the easiest things imaginable, it is not (usually) UTF-8.
Save it as XML and declare it as UTF-8:
<?xml version="1.0" encoding="UTF-8"?>
It will not display (it will not be be read) correctly, even if it's declared as UTF-8.
I had a string of data containing French letters, that needed to be saved as XML for syndication. Without creating a UTF-8 file from the very beginning (changing options in IDE and "Create New File") or adding the BOM at the beginning of the file
$file="\xEF\xBB\xBF".$string;
I was not able to save the French letters in an XML file.

One practical difference is that if you write a shell script for Mac OS X and save it as plain UTF-8, you will get the response:
#!/bin/bash: No such file or directory
in response to the shebang line specifying which shell you wish to use:
#!/bin/bash
If you save as UTF-8, no BOM (say in BBEdit) all will be well.

The Unicode Byte Order Mark (BOM) FAQ provides a concise answer:
Q: How I should deal with BOMs?
A: Here are some guidelines to follow:
A particular protocol (e.g. Microsoft conventions for .txt files) may require use of the BOM on certain Unicode data streams, such as
files. When you need to conform to such a protocol, use a BOM.
Some protocols allow optional BOMs in the case of untagged text. In those cases,
Where a text data stream is known to be plain text, but of unknown encoding, BOM can be used as a signature. If there is no BOM,
the encoding could be anything.
Where a text data stream is known to be plain Unicode text (but not which endian), then BOM can be used as a signature. If there
is no BOM, the text should be interpreted as big-endian.
Some byte oriented protocols expect ASCII characters at the beginning of a file. If UTF-8 is used with these protocols, use of the
BOM as encoding form signature should be avoided.
Where the precise type of the data stream is known (e.g. Unicode big-endian or Unicode little-endian), the BOM should not be used. In
particular, whenever a data stream is declared to be UTF-16BE,
UTF-16LE, UTF-32BE or UTF-32LE a BOM must not be used.

From http://en.wikipedia.org/wiki/Byte-order_mark:
The byte order mark (BOM) is a Unicode
character used to signal the
endianness (byte order) of a text file
or stream. Its code point is U+FEFF.
BOM use is optional, and, if used,
should appear at the start of the text
stream. Beyond its specific use as a
byte-order indicator, the BOM
character may also indicate which of
the several Unicode representations
the text is encoded in.
Always using a BOM in your file will ensure that it always opens correctly in an editor which supports UTF-8 and BOM.
My real problem with the absence of BOM is the following. Suppose we've got a file which contains:
abc
Without BOM this opens as ANSI in most editors. So another user of this file opens it and appends some native characters, for example:
abg-αβγ
Oops... Now the file is still in ANSI and guess what, "αβγ" does not occupy 6 bytes, but 3. This is not UTF-8 and this causes other problems later on in the development chain.

As mentioned above, UTF-8 with BOM may cause problems with non-BOM-aware (or compatible) software. I once edited HTML files encoded as UTF-8 + BOM with the Mozilla-based KompoZer, as a client required that WYSIWYG program.
Invariably the layout would get destroyed when saving. It took my some time to fiddle my way around this. These files then worked well in Firefox, but showed a CSS quirk in Internet Explorer destroying the layout, again. After fiddling with the linked CSS files for hours to no avail I discovered that Internet Explorer didn't like the BOMfed HTML file. Never again.
Also, I just found this in Wikipedia:
The shebang characters are represented by the same two bytes in extended ASCII encodings, including UTF-8, which is commonly used for scripts and other text files on current Unix-like systems. However, UTF-8 files may begin with the optional byte order mark (BOM); if the "exec" function specifically detects the bytes 0x23 0x21, then the presence of the BOM (0xEF 0xBB 0xBF) before the shebang will prevent the script interpreter from being executed. Some authorities recommend against using the byte order mark in POSIX (Unix-like) scripts,[15] for this reason and for wider interoperability and philosophical concerns

Here is my experience with Visual Studio, Sourcetree and Bitbucket pull requests, which has been giving me some problems:
So it turns out BOM with a signature will include a red dot character on each file when reviewing a pull request (it can be quite annoying).
If you hover on it, it will show a character like "ufeff", but it turns out Sourcetree does not show these types of bytemarks, so it will most likely end up in your pull requests, which should be ok because that's how Visual Studio 2017 encodes new files now, so maybe Bitbucket should ignore this or make it show in another way, more info here:
Red dot marker BitBucket diff view

I save a autohotkey file with utf-8, the chinese characters become strang.
With utf-8 BOM, works fine.
AutoHotkey will not automatically recognize a UTF-8 file unless it begins with a byte order mark.
https://www.autohotkey.com/docs/FAQ.htm#nonascii

UTF with a BOM is better if you use UTF-8 in HTML files and if you use Serbian Cyrillic, Serbian Latin, German, Hungarian or some exotic language on the same page.
That is my opinion (30 years of computing and IT industry).