Editing Incubator escapee wiki:Help desk/Archive 2 (section)

== Tamil Wikipedia ? Unable to create user in Tamil script ==

I am trying to create a user "&#2992;&#3006;&#2972;&#2970;&#3015;&#2965;&#2992;&#2985;&#3021; &#2980;&#3008;&#2986;&#2965;&#3021;" in the Tamil Wikipedia. It gives an error "&#2984;&#3008;&#2969;&#3021;&#2965;&#2995;&#3021; &#2965;&#3018;&#2975;&#3009;&#2980;&#3021;&#2980; &#2986;&#2991;&#2985;&#2992;&#3021; &#2986;&#3014;&#2991;&#2992;&#3021; &#2970;&#3014;&#2994;&#3021;&#2994;&#3009;&#2986;&#2975;&#3007; &#2951;&#2994;&#3021;&#2994;&#3016;.". All software I am using are UTF-8 based.

I am asking here because I found no Help desk in the Tamil Wikipedia. The username is my English username in Tamil script. The error message means that the username I gave is not valid. [[User:Rajasekaran Deepak|Rajasekaran Deepak]] 11:38, 2004 Jun 12 (UTC)

: I have created a user "&#2980;&#3008;&#2986;&#2965;&#3021;". Could someone change the username to "&#2992;&#3006;&#2972;&#2970;&#3015;&#2965;&#2992;&#2985;&#3021; &#2980;&#3008;&#2986;&#2965;&#3021;"? Thanks in advance. [[User:Rajasekaran Deepak|Rajasekaran Deepak]] 12:02, 2004 Jun 12 (UTC)

::Perhaps your username is too long. The trouble is that tamil script takes quite many bytes per character. [[User:Sverdrup|&#9999; Sverdrup]] 12:05, 12 Jun 2004 (UTC)

::: What is the limit for the length of usernames? [[User:Rajasekaran Deepak|Rajasekaran Deepak]] 12:33, 2004 Jun 12 (UTC)

::::I asked our devs, and I was told it is 255 single-byte characters. If tamil script is encoded at 4 bytes per character, this means you can use at least 50 tamil characters, which is much more than what you used. [[User:Sverdrup|&#9999; Sverdrup]] 16:33, 12 Jun 2004 (UTC)

:::::The limit is 32 bytes not 255 bytes. This limit was instituted in response to people trolling on en with excessively long names. It should really be 32 characters, but unfortunately that's not trivial in our PHP environment. -- [[User:Tim Starling|Tim Starling]] 16:38, Jun 12, 2004 (UTC)

::::::My suggestions, in order of decreasing elegance:
::::::# Make the limit 32*4 bytes in UTF-32 '''or'''
::::::# Use UTF-8 uniformly, and divide the number of bit-8-set characters by 4 while finding the length '''or'''
::::::# Increase the limit to 64 bytes (in the current setup)
::::::[[User:Rajasekaran Deepak|Rajasekaran Deepak]] 04:30, 2004 Jun 14 (UTC)