2012/4/5 Jesse Farinacci <jie...@gmail.com> > Greetings, > > On Thu, Apr 5, 2012 at 4:04 AM, Arnaud Héritier <aherit...@gmail.com>wrote: > >> * Yes it is the bad practice to put repos in POMs >> * Yes it many more developers/contributors friendly to add repo in POMs >> (as far as we don't add not controlled repos and our artifacts aren't >> supposed to be reused) >> * Yes it was a bad idea and more generally all this part of projects >> infra description, users/groups settings .... >> > > It's only bad because the location might change. Since it is really just a > CNAME record, I don't think the root cause of why specifying > <{r,pluginR}epositories> applies. > > +1 to coding our own <{r,pluginR}epositories> section for our CNAME record > +1 to coding a m-enforcer-p rule which fails builds for > additional <{r,pluginR}epositories> > > >> As I said to Nicolas & KK at the last hackergarten in Paris I drived a >> contribution for such mass change : https://gist.github.com/2305867 (by >> @jeanhelou) >> Nicolas might probably starts from it. >> > > Please remember that there are close to 200 plugins which are in SVN > still. They would need to be migrated to GitHub before applying this kind > of broad and automatic conversion. >
Indeed, committing such a pom change to github will automatically stop the svn -> github sync, without plugin author to be involved, and this is bad. I'll have to check on pom.xml for <scm> and commit at the right place > > > https://wiki.jenkins-ci.org/display/JENKINS/Moving+from+Subversion+%28svn%29+to+Github > > -Jesse > > -- > There are 10 types of people in this world, those > that can read binary and those that can not. >