RE: IRI Templates and Bidi Characters

James M Snell wrote:
> No, not necessarily; the string could have come from another 
> source and may not contain the necessary markup or the 
> developer may have entered the RTL characters using escapes 
> (e.g. \u0xxx); There are a wide range of possibilities. 
> Fortunately, it is exceedingly simple to strip and add the 
> necessary formatting characters.

If the developer entered in the escapes manually, or he otherwise stores and edits the template electronically in a format that cannot preserve the visual ordering, then the visual ordering of the template must not matter much to the user; otherwise, how can he effectively edit the template? And, why does it make sense for the user's editor to display the template differently than the template processor does? That is the part I don't understand. If Abdera prints an error message about an IRI template, and then I go to edit the file where the IRI template is stored, and the contents of that file look different then the error message, then whose fault is it? I would say it is my fault for not storing the template correctly, and Abdera's fault for not implementing the BIDI algorithm faithfully.

- Brian

Received on Monday, 3 December 2007 17:17:36 UTC