[Mapbender-dev] Discussion: Character sets (client encoding)

Christoph Baudson christoph.baudson at wheregroup.com
Thu Jul 17 03:15:42 EDT 2008


Arnulf Christl schrieb:
> Christoph Baudson wrote:
>> Arnulf Christl schrieb:
>>> Christoph Baudson wrote:
>>>> This is a proposal to serve UTF-8 encoded HTML pages only. In my 
>>>> opinion there is no need to serve ISO-8859-1 pages.
>>>>
>>>> All character set conversions (when interacting with a database) 
>>>> could be done in PHP. So the architecture would change from
>>>>
>>>> client: ISO/UTF8
>>>> server: UTF8
>>>> database: ISO/UTF8
>>>>
>>>> to
>>>>
>>>> client: ISO/UTF8
>>>> server: UTF8
>>>> database: ISO/UTF8
>>>>
>>>> This will simplify things a lot and would make Mapbender less 
>>>> error-prone. If anyone has any reason why we should still serve ISO 
>>>> pages, please step up. I have no idea why we should.
>>>>
>>>> Thanks,
>>>>
>>>> Christoph
>>>
>>> You confuse me, I cannot see the difference?
>>
>> neither do I.
>>
>> Here's what I meant
>>
>> client: UTF8
>> server: UTF8
>> database: ISO/UTF8
>>
>> Christoph
>>
>> nr: copy and paste for beginners
>
> I am just wondering - what is the reason that we need an ISO database 
> dump at all? Why do we have it?

Legacy. If even you consider dropping it, we really should.

>
> Obviously all who currently have an ISO database would need to convert 
> to UTF8 and this is painful but if we can implement a good migration 
> path this should be manageable.
> Unfortunately the Mapbender architecture is way more complex than just 
> client, server and database. What happens to WMS that serve ISO 
> encoded FeatureInfo results through the Mapbender tunnel. Or if you 
> have one ISO and one UTF8 in the tunnel? WFS with GML that contains 
> ISO encoded text. The full set of encoding trouble is a quite a stack 
> indeed.
> If it is just the three mentioned layers I am all in favor.

In my opinion the server side is responsible for converting incoming 
data to UTF8. I mentioned database only, but of course you are right, 
there is much more.

Anything that works encoding-wise in Mapbender now is based on pure 
chance, so it can't get any worse.

Christoph

>
> +1
>
> Regards, Arnulf.
>
> _______________________________________________
> Mapbender_dev mailing list
> Mapbender_dev at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/mapbender_dev



More information about the Mapbender_dev mailing list