[mythtv-users] Swedish characters in database

[Mike Perkins]

Michael T. Dean wrote:
> On 10/09/2007 04:37 PM, Joacim J wrote:
>> On 10/9/07, Michael T. Dean <mtdean at thirdcontact.com> wrote:
>>   
>>> On 10/09/2007 03:50 PM, Joacim J wrote:
>>>     
>>>> I have problems with swedish characters in the database and showing it
>>>> in myth.
>>>>
>>>> The mysql collation is "latin1_swedish_ci" and the database server has
>>>> "UTF-8 Unicode (utf8)"
>>>>
>>>> Should I have collation "utf8_swedish_ci" instead??
>>>>       
>>> No.  And, if you change it now, though things will appear to work, when
>>> you upgrade in the future, not only will Myth break, but all your data
>>> will break.  Changing the encoding is a database time bomb.  Every
>>> post/HOWTO where people suggest this is just setting someone up for a
>>> big problem in the future.
>>>
>>> Why:
>>> http://www.gossamer-threads.com/lists/mythtv/commits/264510#264510
>>> http://www.gossamer-threads.com/lists/mythtv/dev/227559#227559
>>>
>>> So, what you need to do is find the locations where we're missing the
>>> UTF encode/decode in the MythTV source code and submit a patch.
>>>     
>> Just so I understand it. Swedish characters are not encoded/decoded
>> correctly and should be considered as a Myth bug. The data in mysql is
>> (should be) always UTF-8??
>>
>> Is this a correct understanding??
> 
> AIUI, Swedish characters should be handled correctly in many (most?)
> places in Myth.  However, there may be places in Myth where the
> conversion doesn't properly occur.  If you can identify a specific place
> where they're not handled correctly you may have found a bug.  Though
> the best thing to do is post a specific question here to determine
> whether it's a bug, a configuration issue, or broken data (from EIT,
> grabbers, scripts, ...) before reporting it on Trac.
> 
> In other words, a ticket titled, "Swedish characters don't work in
> MythTV," would likely be closed as invalid unless it contained specific
> information to identify a point where we're missing some needed code.
> 
Historically, all this is because, as I understand it, Mysql is made by a 
Swedish company, and that was the default character set that was used by them. 
If this problem had been recognised (and fixed) in the dim and distant 
beginnings of Myth... hindsight is a wonderful thing.

I wonder what will happen in the future as Myth becomes more popular worldwide. 
I know we have a guy in Japan can get that character set OK, but...

Is there any plan, even remote, to make Myth properly Unicode?

Mike Perkins