Since a single global LCP transcoder is used by XMLString (and another by DOMString) it would appear that LCP transcoders, anyway, must be re-entrant.
Can somebody backup my assertion?
Yes, I always assumed this (and usually, the LCP transcoder is just a wrapper for some OS-specific API, with no fields defined in the class definition).
And does anybody claim that transcoders adhering to the virtual transcoding API would for any reason also need to be re-entrant?
I would say that the transcoding interface does not require its implementation to be re-entrant; only the implementation for the local code page must. If some code asks for a specific transcoder, it is supposed to access it in a thread-safe way.
BTW, that's only my opinion...
Alberto
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
