> Date: Sat, 27 Jul 2013 10:53:09 +0200
> Subject: Re: [DISCUSS] Should the Maven PMC be an example of how we want the
> Maven Community to behave (was Re: svn commit: r1506778 -
> /maven/site/trunk/content/markdown/project-roles.md)
> From: fred.co...@gmail.com
> To: dev@maven.apache.org
>
> I have a bunch of IRL stuff to deal with at the moment, and can't even work
> on my own FOSS or proprietary code, so don't expect a "fork" soon.
>
> When I say "broken", don't be offended, few codebases aren't broken in some
> way :-)
>
> One example of "broken" is my SCM url being munged from this:
>
> scm:git:git:${project.groupId}/${project.artifactId}.git
>
> To this:
>
> scm:git:git:${project.groupId}/${project.artifactId}.git/${project.artifactId}
MG>set up a link#!/bin/sh
if test "${SCM}x" = "x"; then
SCM="LN -s
scm:git:git:${project.groupId}/${project.artifactId}.git/${project.artifactId}
scm:git:git:${project.groupId}/${project.artifactId}.git"
if test -x ${SCM}; then
export SCM
else
echo "Error: Set the SCM environment variable."
exit
fi
fiifndef SCM
SCM=${SCM}
endif
MG>then reference SCM in your maven pom.xml
>
> And all I have to say to that, other than the pain it caused me, is LOL!
> :-p OK, I can add one more thing: F*** SVN. OK, now I'm done. :-)
>
> Then there's the issue where my .ssh/config file was totally ignored unless
> I used the external provider, and so on and so forth.MG>export SSH="ssh -s -F
> /home/fcooke/.ssh/config $1"
>
> Reality check: It's a lot easier and quicker to hack in a fix for me that
> breaks everyone else, than it is to agree on and implement a solution that
> works for everyone in a backward compatible way. Hence fork. Make what I
> need happen now and not sit around waiting for agreement on how to progress
> "properly".
>
> Other than a fork of my very own, I'd be VERY hesitant to trust anyone
> else's forks. For example, I don't trust Debian Maven nor Fedora Maven out
> of unjustified and paranoid fear that their patches to "just use what the
> distro has" for deps might be active when they shouldn't be, etc. Forks
> rarely gain traction, unless the base project is severely lacking in some
> way.
MG>specific examples?
> Fred.MG>Martin
>
> On Sat, Jul 27, 2013 at 9:56 AM, Hervé BOUTEMY <herve.bout...@free.fr>wrote:
>
> > Le vendredi 26 juillet 2013 01:05:44 Fred Cooke a écrit :
> > > About the fork, though, I'm interested. I'm interested in why it's
> > > maintained and how it differs. I'm interested because I intend to fork it
> > > myself to fix things which would break others which I know won't change
> > > upstream (site and release stuff). Perhaps someone already fixed what I
> > > consider broken and others, perhaps, do not consider broken. Email me
> > > privately if need be.
> > >
> > > If the fork is just a branch (or set thereof), and not promoted as a fork
> > > (in a social sense), then is there really any harm? Is it being viewed
> > as a
> > > loss of potential work done? Open source is open for a reason.
> > +1
> >
> > in your case for example:
> >
> > if you make a fork to site plugin to fix it in a personal way, please don't
> > hide it but show us where you fork/branch and explain how you change the
> > code
> > (in a Jira issue?): I'm interested to see your work and understand what
> > you're
> > doing, since I definitely don't understand how you find the plugin broken
> > (and
> > to be honest, every time I read this "broken" sentence, I feel aggressed).
> >
> > IMHO, a personal branch can be an excellent way of working with the
> > community:
> > I'm really interested to see you fix what I can't do because I really don't
> > understand what is "broken".
> >
> > Of course, if the difference between your branch and trunk gets big, it'll
> > be
> > harder to find how to converge later: but we'll see. Explaned with the
> > community from start, and used to interact with the community to show how
> > it
> > can evolve, we can understand and define with consensus that this can be a
> > real
> > friendly fork = a different choice on some fundamental point. Or perhaps it
> > will be the start og m-site-p 4: who knows?
> >
> > Regards,
> >
> > Hervé
> >
> > >
> > > Fred.
> >
> >
> > ---------------------------------------------------------------------
> > To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org
> > For additional commands, e-mail: dev-h...@maven.apache.org
> >
> >
