On Wed, Dec 07, 2011 at 01:47:04PM -0800, Kalle Korhonen wrote: > There are plenty of other read operations and they should all > then explicitly specify UTF-8, otherwise you get mixed results.
I got the impression from reading <https://issues.apache.org/jira/browse/TAPESTRY-2543> that this was intended to be the current state of affairs. > What if Tapestry, or your own code depends on a library that doesn't > specify the encoding but uses platform default. I'd consider that a bug. What is the point of having the "tapestry.charset" property if it's not truly all-encompassing? > Perhaps a better, more > generic solution is just to document that JVM's default encoding > should be set to -Dfile.encoding=UTF-8. Sounds brittle to me. Few things are more frustrating than environmental dependency problems, because they don't show up in tests or in the development environment. They wait to bite you until deployment time. -- Robert Coie Implementor, Apropos K.K. --------------------------------------------------------------------- To unsubscribe, e-mail: users-unsubscr...@tapestry.apache.org For additional commands, e-mail: users-h...@tapestry.apache.org
