Roman Shaposhnik 2013-08-16, 15:48
-Re: Automating Hadoop installation
Konstantin Boudnik 2013-08-16, 18:06
Yup, patches are always welcome!
As for win support - more the merrier! Although I doubt that many people here
have such experience (say my own stops with a C++ compiler on 3.11; and never
touched it since).
Considering a very low interest in packaged stack from win'd crowd , I
personally would be very hesitant to -1 patches that work everywhere yet
breaking some windows functionality. Simply because most of us don't have
windows machines to test it and most of us are working in Unix shops.
So, someone with right skills needs to take the ownership of such support
going forward (if it gonna happen). Apparently, same goes with Mac - I am
trying not to touch it with a 10' pole, but if someone will step up and own
the support in Bigtop it'd be great.
Consider this a call for volunteers ;)
On Fri, Aug 16, 2013 at 08:48AM, Roman Shaposhnik wrote:
> And by the way, we had a few hackathon projects
> for adding Solaris IPS and Homebrew support to
> Bigtop. We just haven't had anybody dedicated
> enough to stick around for helping us maintain
> those bits.
> Other than that -- patches are ALWAYS welcome.
> P.S. apt-get install hadoop works today provided you
> get one of the list files hooked up to your system:
> On Fri, Aug 16, 2013 at 8:39 AM, John Meagher <[EMAIL PROTECTED]> wrote:
> > That sounds like what Bigtop is doing, at least covering the Linux
> > distros. http://bigtop.apache.org/
> > On Fri, Aug 16, 2013 at 11:23 AM, Andrew Pennebaker
> > <[EMAIL PROTECTED]> wrote:
> >> I think it would make Hadoop installation easier if we released standardized
> >> packages.
> >> What if Ubuntu users could simply "apt-get install hadoop" they same way
> >> they "apt-get install apache2"?
> >> Similarly, could we release a Chocolatey package for Windows users? The
> >> easier the installation process, the more people want to use Hadoop.
> >> There's already a Homebrew package for Mac users, so they can "brew install
> >> hadoop"!