Re: [gwt] Re: GEP, war/WEB-INF/lib, and source control

2009-04-26 Thread Isaac Truett
Thanks. That is an interesting thread, although it seems to be addressing a separate need: multiple classpath definitions, which I think would be a big win for the GEP. I've been using multiple classpath definitions in my Ant build scripts for years. Usually there's a runtime classpath with JDBC

Re: [gwt] Re: GEP, war/WEB-INF/lib, and source control

2009-04-26 Thread Vitali Lovich
That's pretty much the approach I use. I prefer to keep the war/ directory as minimal as possible in terms of what is in source control. I keep only the web.xml app HTML page. By default the war/ directory gets added to ignore because that is where generated resources are placed. Instead I

Re: [gwt] Re: GEP, war/WEB-INF/lib, and source control

2009-04-25 Thread Allen Firstenberg
There has also been some similar discussion in this thread: http://groups.google.com/group/google-appengine-java/browse_thread/thread/67cb7cdaefc8429f?tvc=2 --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups Google Web

GEP, war/WEB-INF/lib, and source control

2009-04-24 Thread Isaac Truett
When using GEP with GWT and GAE support, is it expected that the jars the plugin automatically places in war/WEB-INF/lib will be committed to the source repository? I expected that they were placed there by one of the builders, so I left them out of my commit. When I check out the project to

Re: GEP, war/WEB-INF/lib, and source control

2009-04-24 Thread Jason Parekh
Hey Isaac, Good question. You can go either route: - If you do not check them in, you'll get a warning in Eclipse. Doing a quick fix (select it and ctrl+1) on that warning will copy the installed SDK's JARs over to your lib directory. - If you do check them in, like you mentioned, there's a

Re: GEP, war/WEB-INF/lib, and source control

2009-04-24 Thread Jason Parekh
Thanks for the suggestion--I'll run it by the team. jason On Fri, Apr 24, 2009 at 1:42 PM, Isaac Truett itru...@gmail.com wrote: Jason, Fair enough. Thanks. What about elevating those to errors so that they interrupt run/debug to point out that there's a problem? Or if always being an