On 19/04/2010, Tim Sneddon <tim.sned...@bigpond.com> wrote: > On 19/04/2010 11:31 PM, sebb wrote: > > > On 19/04/2010, Tim Sneddon<tim.sned...@bigpond.com> > wrote: > > > > > Hi, > > > > > > I am currently in the process of porting Maven (and all the > > > bit and pieces it drags in) to OpenVMS. One of the things it > > > drags in is Commons Exec. So I have done a bit of work to > > > get it running properly (I found the existing support for VMS > > > to be lacking) and now I would like to hopefully have this go > > > back into the source pool. > > > > > > So, I was wondering how I go about getting this stuff passed > > > back into the source pool. The stuff I read mentioned > > > submitting a patch via JIRA. Is that all I have to do? Is > > > there a contribution agreement I need to fill out? > > > > > > > JIRA is normally sufficient for a patch. > > > > Excellent. I'll post it in the morning (quite late, here in > Australia :-). > > > > > > For larger contributions, then a CLA may be required. > > > > http://www.apache.org/licenses/#clas > > > > Thanks. I've filled that out. I'll scan it and send it on. > > > > > > > > > Also, are there many other OpenVMS users out there? > > > > > > > I used to use OpenVMS in my day job (and had used it since VMS 5.x on > > VAX) but I've not used it for a couple of years. > > > > > > > Is there a current maintainer for OpenVMS support? > > > > > > > I did a bit of work on the OpenVMS parts of Exec and fixed a few > > problems a while back. > > > > However I did not fix all the problems; at the time there seemed to be > > some limitations in the VMS JVM implementation that made it very > > difficult to get Exec working as well on VMS as it does elsewhere, > > particularly in regard to process termination. > > > > > > Do you have more information on this? I'd like to investigate it. I > have tweaked it to follow the same logic as the C run-time library > exec*() functions. >
At the time (May 2008), I wrote: "There seems to be a problem with the behaviour of process destroy on VMS - it does not always guarantee that the process is destroyed. I think that is a JVM bug." IIRC this was something to do with launching processes via a DCL script. It should show up in the test cases - either as a failure or as a specific exemption for VMS to avoid the failure. > > > I'd be very happy to assist on this in any way I can. > > > > > > Thanks. > > Tim. > > --------------------------------------------------------------------- > To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org > For additional commands, e-mail: dev-h...@commons.apache.org > > --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org For additional commands, e-mail: dev-h...@commons.apache.org
