feelfree to submit the patch, of course with unit test, as long as it follows maven scm architechutre. I will review and commit
Thanks -D On 9/27/07, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote: > > Emmanuel Venisse <[EMAIL PROTECTED]> wrote on 09/27/2007 07:36:48 AM: > > > > > > > 3) Does the provider or individual command implementations <i>have</i> > > > to use a command line framework, or was that just done for ease of > > > implementation? The reason I ask is that our build environment uses > > > Starteam in a manner that exposes a design flaw in their checkout > > > process (I can provide details if anyone asks, but it's not really > > > relevant to the discussion). > > > > We use starteam cli to run command because we can't redistribute > > starteam jars. > > > > Thanks. This is not an issue for us because our local repository will be > accessible only to those developers who have a valid Starteam license > anyway. The sad part is that my implementation cannot be shared back to > Maven :(. > > However, I do have some positive feedback. I have created SCM-347 in JIRA > and placed it there. > > I would also like to promote issue SCM-117. It looks like discussion stopped > on 12/23/2006 and that no further work was done. Implementing this will > really improve the usefulness of stcmd, which has dozens of options (like > the three listed in SCM-347) that you might not want to explicitely support, > because they are used less frequently and/or do not fit into the SCM model. > > As the original reporter OF SCM-117 indicated, the most useful argument by > far is -o (force checkout) because Starteam has real issues with repository > synchronization (SCM-142). This is a Starteam problem and one that can only > be solved by the force checkout option or through the Starteam SDK, which > contains calls to explicitely synchronize. > > > Robert Egan >
