Aye, one real quick clarification:
I strongly believe that we should support the free alternatives even if they suck more at the moment, possibly enhance them and contribute them back to the larger Hadoop & Cassandra communities.
Using DSE doesn't prohibit us from doing that. DSE has a couple of closed pieces, most importantly CFS, which ties Hadoop and Cassandra together. We'd still be using open source versions of those.
On Sep 25, 2012, at 2:24 PM, Faidon Liambotis faidon@wikimedia.org wrote:
On Tue, Sep 25, 2012 at 10:37:36AM -0400, Andrew Otto wrote:
Okokok. So I'm starting this thread in order to answer my big question: Are we allowed to use DSE even if it is not 100% open source? If the answer is an easy 'no', then our choice is easy: we will use CDH4. Is it possible to get an answer to this question before we go down the road of dedicating time to evaluating and learning DSE?
Based on your choice of words ("allowed"), I think you're looking for some kind of authoritative/executive decision and I'm not sure who can give you that.
As far as my personal opinion goes, I'd hate us to see using so-called "Enterpise" editions and non-free (as in freedom) software. I realize we have a few exceptions already (incl. Oracle Java) but this is a slippery slope to be in. I strongly believe in free & open-source software and its values -- values similar to those of the free culture movement, which we're supposed to support.
I also enjoy the freedoms on a daily basis, incl. studying the code that we use to understand the software and the bugs I'm finding as well as fixing those bugs with no EULA restrictions or binaries to fiddle with. We're also interacting with the communities of said software, giving them feedback and patches and getting "support" for free, without needing friends or special company relations (or paid contracts). Yes, on a *daily* basis, I can think of numerous examples.
I strongly believe that we should support the free alternatives even if they suck more at the moment, possibly enhance them and contribute them back to the larger Hadoop & Cassandra communities.
As a final note and as I mentioned at the meeting yesterday, we currently have no infrastructure to support the use of proprietary software: our repositories (apt, puppet etc.) are public and we're effectively redistributing everything that we put there, which is going to class with a proprietary license. But that's probably a smaller, one-off cost of setting something up for that.
Best, Faidon