Thank you, tic, my wiki now looks perfect in the browser, even though I'm using unicode now!
My way to nearly perfect satisfaction: I made the dump on a unix-machine, moved the .sql-file to my w2k-pc and used recode to convert into unicode. After conversion, the .sql-file showed broken special-characters in my windows-notepad - I ignored this and loaded the dump into the database. After running the mediawiki-config again, setting Language to "de - Deutsch - Unicode", the wiki looks like it should.
Still a bit confused: What I do not understand is the fact, phpmyadmin shows broken special-characters in the database. --> This seams to me, like there are (at least) to views to the database - one using the mediawiki-script and another one using the phpmyadmin-script.
Perfect confusion: Strange enough (see above), I'm running another German wiki on the same unix-machine under the same user, using another database. This second wiki is also set to use "de - Deutsch - Unicode" via mediawiki-config. There are no problems, browsing through wiki-pages with special-characters - and also phpmyadmin shows pretty Äs,Ös etc. in the database.
Conclusion: I'm happy, that the front-end for my wiki-users looks perfect and therefore thankful for the tics help. But I'm also curious, why things in the database look, like they look. Maybe, I have a problem in one of the databases, which mediawiki compensates at present, but this could cause troubles in the future - my belly says...
Niki
-----Ursprüngliche Nachricht----- Von: mediawiki-l-bounces@Wikimedia.org [mailto:mediawiki-l-bounces@Wikimedia.org] Im Auftrag von tic@tictric.net Gesendet: Dienstag, 12. Oktober 2004 00:05 An: MediaWiki announcements and site admin list Betreff: Re: [Mediawiki-l] converting from latin1 to unicode?
Am Mon, 11 Oct 2004 15:11:31 +0200 hat Nikolai Neumayer neum@ai.wu-wien.ac.at geschrieben:
Hello,
I'm running a German wiki configured using latin1. Trying to use unicode instead of latin1, all special characters like ÄÖÜ etc. are broken in the browser (and also in the database, if I take a look into the tables via phpmyadmin). This effect can be undone by running the setup again configuring latin1. Not being a specialist in such things it
seems to me
to
You just make a dump of your database like
$ man mysqldump $ mysqldump -u user -p --add-drop-tables database $ cp database.sql database.sql.backup
Then run that database.sql file through a converter. I use recode and it's pretty powerful. If you use debian it's just an
$ apt-get install recode
$ man recode $ recode -l $ recode l1..u8 database.sql
This should be fine normaly. Then load it back into the database.
$ man mysql $ mysql -u user -p database < database.sql
In any case make sure you read the according manpages before you use the commands to get the syntax right. I often make mistakes :-)
Of course you can use phpmyadmin for the dumping and loading but I think it's a bit long-winded for this task.
Unix is cool :-)
--tic
PS: Greetings to the city I was born in :-)
be more "modern" to use unicode. Is this correct, or should I just leave latin1. If it would make sense to convert to unicode, how can I do this? What are the consequences of using the one or the other?
Sorry for this questions of one mainly just beeing a "user" but who also does some things on a server. Any hint, even in the form of hyperlinks to further reading would be appreciated.
Nikolai
MediaWiki-l mailing list MediaWiki-l@Wikimedia.org http://mail.wikipedia.org/mailman/listinfo/mediawiki-l
_______________________________________________ MediaWiki-l mailing list MediaWiki-l@Wikimedia.org http://mail.wikipedia.org/mailman/listinfo/mediawiki-l