Re: Conflicts and JDT

2006-06-09 Thread Shankar Unni
Yoav Shapira wrote: There's also another reason: if we use JDT then we don't require users to have the complete JDK, only a JRE. This is a big deal (perhaps even more important than the technical / performance benefits) from a licensing perspective, especially for people who repackage / redistr

Re: Conflicts and JDT

2006-06-08 Thread Costin Manolache
I got it, so we have to use jdt. What about using a separate classloader to load JDT ( child or sibling of server classloader for example ), so any jdt in webapps won't be visible ? Well - if you are comfortable maintaining JDT changes, I don't have a problem, it just doesn't feel right. For sma

Re: Conflicts and JDT

2006-06-08 Thread Yoav Shapira
Hi, My understanding was that the main reason is that it's better / faster / less leaky than javac - is this correct ? Is it still true for JDK1.5 ? Could we just use javac as default - and in case of 'conflicts' ? There's also another reason: if we use JDT then we don't require users to have

Re: Conflicts and JDT

2006-06-08 Thread Remy Maucherat
Costin Manolache wrote: Could you summarise what conflicts and why do we have to use JDT ? As usual: version conflicts with people who would use JDT in their webapps. My understanding was that the main reason is that it's better / faster / less leaky than javac - is this correct ? Is it still

Re: Conflicts and JDT

2006-06-08 Thread Costin Manolache
Could you summarise what conflicts and why do we have to use JDT ? My understanding was that the main reason is that it's better / faster / less leaky than javac - is this correct ? Is it still true for JDK1.5 ? Could we just use javac as default - and in case of 'conflicts' ? Package rename kin