On 02/08/2013, at 8:14 PM, Luke Daley <luke.da...@gradleware.com> wrote:
> Hi, > > Our handling of mavenLocal() is pretty busted. There are numerous issues for > this. > > The most important is that fact that we don't treat mavenLocal() in that same > way that m2 does; we stop resolution if we find any artifacts (including > POMs) for a dependency in there. Recently I've had lots of people struggling > with this because they have lots of POMs and no jars for artifacts in their > maven local cache. I'm not sure what tooling exactly is populating the local > m2 repo with just POMs, but it seems fairly common. > > The other big problem is that we install unique snapshots into the local > cache, while mvn does not. Even in m3. > > Just wondering what our short - medium term plan on this is. Many people rely > on mavenLocal() daily for sharing development snapshots, regardless of all > the inherent problems with that concept. I can sense some pressure building > around this, and it feels like something we should come up with a solution > for some time soonish. The plan is to fix the issues, then solve the use case in a better way at some point. It's a matter of finding the bandwidth. -- Adam Murdoch Gradle Co-founder http://www.gradle.org VP of Engineering, Gradleware Inc. - Gradle Training, Support, Consulting http://www.gradleware.com
