Il giorno ven, 17/11/2006 alle 12.41 -0800, Per Bothner ha scritto: Hi Per, Hi Guilhem, > Guilhem Lavaux wrote: > > As a remainder I have made in 2004 a comparison between ICU and > > java.text. I cannot find why I have at the time rejected a ICU binding > > (except it is a huge machinery for so little classes and that > > RuleBasedCollator must have a severe layer to adapt the rules to ICU's > > format). Just back, so I still have to look at the jdk-ICU differences in details, but it seems clear that at least they can be imported and we can base our implementation to these classes. I'm currently reading the 1.6 javadoc to find other clues. It would be great to have all our implementation based on the ICU4J classes, but if this is too work and pain to do with little gain, we can still consider to have only some classes based on this API. I was looking at this to get a fix for the decompose method in Collator, which is absent in Classpath and broken in gcj, as well as complete my DecimalFormat implementation, which has still few issues, but seems to me better then our old code. I can actually just "stole" the bits I need there :) - with given credit where it is due - > Also, before you import large chunks of code that have not been > assigned to the FSF, you should get approval from Stallman. Yes, sure. What I have to do about this? Mark? Thanks, Mario -- Lima Software, SO.PR.IND. s.r.l. http://www.limasoftware.net/ pgp key: http://subkeys.pgp.net/ Please, support open standards: http://opendocumentfellowship.org/petition/ http://www.nosoftwarepatents.com/
Attachment:
signature.asc
Description: Questa =?ISO-8859-1?Q?=E8?= una parte del messaggio firmata digitalmente
