Will the database convert to "utf-8" after upgrade to VB5 ?

Collapse
X
 
  • Time
  • Show
Clear All
new posts
  • sadiq6210
    Senior Member
    • Jun 2004
    • 379
    • 3.6.x

    Will the database convert to "utf-8" after upgrade to VB5 ?

    I noticed VB5 is using "utf-8"



    But this forum is using "iso-8859-1"



    My question is
    When you upgrade this forum (vbulletin.com/forum) to VB5, will the database convert automatically during the upgrading from "iso-8859-1" to "utf-8" ?
    Signature test :)
  • Hartmut
    Senior Member
    • Nov 2007
    • 2870
    • 4.2.x

    #2
    Originally posted by sadiq6210
    When you upgrade this forum (vbulletin.com/forum) to VB5, will the database convert automatically during the upgrading from "iso-8859-1" to "utf-8" ?
    No, it doesn't.
    No private support, only PM me when I ask for it. Support in the forums only.

    Comment

    • iCafe
      New Member
      • Apr 2005
      • 29
      • 4.1.x

      #3
      Originally posted by sadiq6210
      I noticed VB5 is using "utf-8"
      Originally posted by sadiq6210



      But this forum is using "iso-8859-1"



      My question is
      When you upgrade this forum (vbulletin.com/forum) to VB5, will the database convert automatically during the upgrading from "iso-8859-1" to "utf-8" ?


      it encoding of HTML Character Set not Character Set of database

      Comment

      • manebe
        Member
        • Jan 2011
        • 49

        #4
        Database staying on ISO-8859-1 and the whole board on UTF-8,

        I heard it could cause some issue with special characters, isn't it ?

        Comment

        • sadiq6210
          Senior Member
          • Jun 2004
          • 379
          • 3.6.x

          #5
          So the database for this forum will remaining the same after upgrade to VB5? "iso-8859-1"

          Originally posted by manebe
          I heard it could cause some issue with special characters, isn't it ?
          Because of that I asked this question
          I wish it will not cause any issue with special characters

          Also, it will be great if vbulletin can provide a tool to convert the database to utf-8
          Signature test :)

          Comment

          • manebe
            Member
            • Jan 2011
            • 49

            #6
            Originally posted by sadiq6210
            So the database for this forum will remaining the same after upgrade to VB5?
            Originally posted by sadiq6210
            "iso-8859-1"


            Because of that I asked this question
            I wish it will not cause any issue with special characters

            Also, it will be great if vbulletin can provide a tool to convert the database to utf-8


            Any update about this question?

            Different encoding (database and board), thread IDs being changed, no styles and plugin compatibility

            Not so easy to upgrade from vB4 to vB5

            Comment

            • Marco Mamdouh
              Senior Member
              • Oct 2008
              • 423

              #7
              Originally posted by sadiq6210
              So the database for this forum will remaining the same after upgrade to VB5? "iso-8859-1"
              vBulletin upgrader doesn't convert your database during the upgrade, It's only changing the required tables and fields and not the database collation or character set.
              Marco Mamdouh Fahem.
              OUZMind Lead Developers
              vBFoster Lead Developers

              Comment

              • Wayne Luke
                vBulletin Technical Support Lead
                • Aug 2000
                • 73981

                #8
                There is a little confusion about databases and UTF-8...

                1. MySQL 4.1 and higher can store data as UTF-8. This is based on the Character Set defined. Character set determines how the data is presented in queries. Collation determines the sort order of characters. Many MySQL Servers default to Latin1 though in newer versions of MySQL that was changed to UTF-8. Newer installs of MySQL should use UTF-8.
                2. vBulletin 3 and 4 can talk to MySQL in UTF-8 as long as you're using the MySQLi classes provided.
                3. The character encoding of the HTML output has nothing to do with MySQL and how it stores data. It tells the browsers how to interpret the characters and display them. Ideally it should match MySQL's character set. So if you're using a Latin1 database, it should be ISO-8859-1. If you're using UTF8 then it should be UTF-8.


                The issue with vBulletin and UTF-8 lies a lot in the programming and how to interpret characters and years of use. Some characters have to be interpreted by ICONV and are turned into HTML entities. Others are passed through. There are also some PHP Functions that don't take UTF-8 into account. This is why when people try to convert to UTF-8 they get problems with characters. These are often problems we cannot solve. It is only a real problem in older forums and when you move servers. When you move servers, the character set can cause problems due to interpretation. Especially if they do not match character sets on both servers.

                On new installs with the MySQL character set to UTF8 and the collation set to utf_general_ci or one of the other UTF8 collations, using MySQLi and telling vBulletin to speak in UTF-8 in the /core/config.php file then there will be very few problems in the system. Same for vBulletin 3 and vBulletin 4. Searching on the other hand brings a new set of issues with certain languages, predominantly Asian, which don't have stop characters like Middle Eastern and European languages. This is on new systems.

                On old systems, the issue is converting data stored as HTML entities back into actual characters during conversion. Or dealing with sites that have multiple languages and specify different character encodings for the language. The same entity can be different characters under different HTML encodings.
                Translations provided by Google.

                Wayne Luke
                The Rabid Badger - a vBulletin Cloud demonstration site.
                vBulletin 5 API

                Comment

                • manebe
                  Member
                  • Jan 2011
                  • 49

                  #9
                  Originally posted by Wayne Luke
                  Ideally it should match MySQL's character set. So if you're using a Latin1 database, it should be ISO-8859-1. If you're using UTF8 then it should be UTF-8.
                  Wayne,

                  That is exactly the problem for those who need special characters.

                  Upgrading to vB5 the database remains on ISO-8859-1, whilst page encoding will change to UTF-8.
                  That is not ideal, as you said.

                  How do we solve this, should we convert database from ISO to UTF-8 before upgrading?

                  Thanks

                  Comment

                  • Merjawy
                    Senior Member
                    • Sep 2002
                    • 2613

                    #10
                    One of my sites vB4, uses UTF-8 (charset and database) and each time I upgrade I create a new database and convert collation to UTF-8 before I restore.
                    I restore the database and make sure all is ok, then I upgrade...And I uncomment the UTF-8 line in my config.php... I never had any issues.
                    To be or not to be... Where the hell is the question????
                    My psychiatrist told me I was crazy and I said I want a second opinion. He said okay, you're ugly too

                    Live vBulletin 4.2.0 Multilingual * Alpha/Beta vB 4 - vB 5 Tier 1A
                    CentOS 6.2 - Apache:2.2.15(Apache2Handler) - PHP:5.3.3 - MySQL:5.1.61
                    Xampp/Win-XP - Apache v2.2.21(Apache2Handler) - PHP:5.3.8 - MySQL:5.5.16

                    Comment

                    • manebe
                      Member
                      • Jan 2011
                      • 49

                      #11
                      Originally posted by Merjawy
                      One of my sites vB4, uses UTF-8 (charset and database) and each time I upgrade I create a new database and convert collation to UTF-8 before I restore.
                      I restore the database and make sure all is ok, then I upgrade...And I uncomment the UTF-8 line in my config.php... I never had any issues.
                      Merjawy, thanks for the answer.

                      Any official answer on this issue ?

                      Comment

                      • Wayne Luke
                        vBulletin Technical Support Lead
                        • Aug 2000
                        • 73981

                        #12
                        Originally posted by manebe
                        Any official answer on this issue ?
                        You might want to re-read my post in this thread...

                        However we don't support converting older databases because it doesn't work for everyone.

                        For MySQL databases, there is no such thing as ISO-8859-1 Encoding.
                        Translations provided by Google.

                        Wayne Luke
                        The Rabid Badger - a vBulletin Cloud demonstration site.
                        vBulletin 5 API

                        Comment

                        • manebe
                          Member
                          • Jan 2011
                          • 49

                          #13
                          Originally posted by Wayne Luke
                          However we don't support converting older databases because it doesn't work for everyone.
                          Not really a good news for us managing boards other than english languages.

                          Originally posted by Wayne Luke
                          For MySQL databases, there is no such thing as ISO-8859-1 Encoding.
                          My database is on latin1-

                          Originally posted by Wayne Luke
                          You might want to re-read my post in this thread...
                          Wayne, you look like a great guy and you are used to give well detailed answers.
                          Please just let me say that my question still keeps unanswered.

                          I am asking vBulletin how do I upgrade my forum form vB4.2 to vB5 being my database on latin1- ?
                          Should I convert my database to UTF-8 before upgrading ?

                          Originally posted by Wayne Luke
                          Ideally it should match MySQL's character set. So if you're using a Latin1 database, it should be ISO-8859-1. If you're using UTF8 then it should be UTF-8.
                          Thanks
                          Last edited by manebe; Mon 17 Sep '12, 7:06am.

                          Comment

                          • Wayne Luke
                            vBulletin Technical Support Lead
                            • Aug 2000
                            • 73981

                            #14
                            Originally posted by manebe
                            My database is on latin1-
                            Which is not equal to ISO-8859-1. ISO-8859-1 is the encoding predominantly for English. Latin1 is a lot more than English.

                            I am asking vBulletin how do I upgrade my forum form vB4.2 to vB5 being my database on latin1- ?
                            Shoul I convert my database to UTF-8 before upgrading ?
                            We don't recommend ever changing the encoding of an active database. It can cause problems. Merjawy's example is the exception, not the rule.

                            You can attempt to follow his example but we will not certify the outcome.
                            Translations provided by Google.

                            Wayne Luke
                            The Rabid Badger - a vBulletin Cloud demonstration site.
                            vBulletin 5 API

                            Comment

                            • manebe
                              Member
                              • Jan 2011
                              • 49

                              #15
                              Originally posted by Wayne Luke
                              We don't recommend ever changing the encoding of an active database. It can cause problems. Merjawy's example is the exception, not the rule.

                              You can attempt to follow his example but we will not certify the outcome.

                              Should I keep database on latin-1 and board on UTF-8 ?

                              That is not ideal as we all know, could cause issue with special charcaters rendering.

                              Let me remember that my database is not so old,
                              it has been created 1 year ago in a fresh installation of vB 4 !

                              Comment

                              Related Topics

                              Collapse

                              Working...