On 19 March 2012 21:20, Benson Margulies <bimargul...@gmail.com> wrote: > On Mon, Mar 19, 2012 at 5:08 PM, Jos Backus <j...@catnook.com> wrote: >> On Mon, Mar 19, 2012 at 1:39 PM, Jason van Zyl <ja...@tesla.io> wrote: >>> >>> On Mar 19, 2012, at 4:14 PM, Jos Backus wrote: >>> >>>> On Mon, Mar 19, 2012 at 11:46 AM, Jason van Zyl <ja...@tesla.io> wrote: >>>>> I think if some 3rd party wants to provide an RPM have at it. I don't >>>>> think this is something we want to create or support. >>>> [snip] >>>> >>>> Any reason why not, especially when it's easy to do so? It lowers the >>>> bar for users to deploy Maven. >>> >>> You're assuming it's easy to do but as an overall supported aspect of the >>> project nothing is easy. Maybe easy for you, but not for us :-) Generating >>> an RPM is one thing, supporting it and have it undergo the construction >>> that RPM proponents might require like building it offline and running it >>> under our normal gamut of tests is probably not easy. You're making an >>> assumption that it lowers the bar, but I would argue that's for a much >>> smaller segment of the user base then you might think -- I believe Windows >>> users still make up the largest segment. So as I've argued in the past the >>> value to the project overall versus the work to actually support creating >>> an RPM is up for discussion. I don't believe it's worth the effort. >> >> Well, if installing Maven is really as easy as just unpacking a >> tarball, creating an RPM should not be hard. > > The java dependency is an issue. And then ... every RPM of a Java > package I've ever seen has felt the need to take the self-contained > hierarchy of the normal distro and move things to other places. Config > to /etc, logs to /var/log, &cetra. So, right, one of us could probably > come up with a trivial RPM, which would be trivially rejected by all > of the distro packages. > > I could also mention the question of equal rights for debian users, > and don't even get me started on Gentoo. >
Please.... not Gentoo.... the crazy sh1t they do is just plain insane when it comes to Maven... The key thing you need to understand is that any piece of software released by Apache has to have at least 3 x +1 votes from the project's PMC committee... I cannot speak for the other PMCs, I can only speak for my self, but I only vote +1 on things I have actually tested myself... most of the time I try to rebuild the artifact from the source bundle, as that is another thing that the PMC is supposed to verify, and I try to look out for the LICENSE.txt and NOTICE.txt files too... I am not going to be in a position to add testing an RPM build into that mix (being not on an RPM based OS... OK I could spin up a VM, but that is not going to happen with the bandwidth I currently have), certainly not going to be able to test the RPM in a reasonable cross-section of the OSes that would give me a confident +1.... and MSI support I have no time to test, let alone set up the toolchain to support generating the MSI... then I'd need the DEB toolchain and test VMs.... I am not going to veto, but I am -0.9 on adding RPMs into the Maven build, i.e. I have no intention of being in a position to vote +1 on a RPM binary being released by Maven -Stephen > --------------------------------------------------------------------- > To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org > For additional commands, e-mail: dev-h...@maven.apache.org > --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org For additional commands, e-mail: dev-h...@maven.apache.org
