Open Stack

> > 
> > I think one of the important things is if it fits in to furthering what is
> > "OpenStack", as far as whether it is a service or functionality that is needed
> > and useful for those running an OpenStack cloud. This is one of the parts that
> > may be more on the subjective side. We need to see that adding the new project
> > in question will enhance the use or operation of an OpenStack environment.
> 
> What do you think we can do to be better informed about whether
> something is actually useful, or just appears useful?
> 

This is definitely a tricky part. We need to be willing to get out and make
connections outside of our small group and learn what we can about how things
are used in the real world. This is one of the main reasons I've been involved
in the ops meetups. I want to be able to hear directly from the folks running
OpenStack clouds what their challenges are and how they are addressing those
challenges today. That helps inform later decisions about whether some new
service fits in with what they need, or if it would be something that doesn't
actually fit with what is commonly done.

> > 
> > There is the question about overlap with existing projects. While I think it's
> > true that a new project can come along that meets a need in a better way than
> > an existing solution, I think that bar needs so be raised a lot higher. I
> > personally would much rather see resources joining together on an existing
> > solution than a bunch of resources used to come up with a competing solution.
> > Even with a less than ideal solution, there is a lot that is learned from the
> > process that can be fed into and combined with new ideas to create a better
> > solution than just having a new replacement.
> 
> Where should we draw the line with building something new and using
> tools available from other communities?
> 

Fighting "not invented here" tendencies is always a challenge. There's usually
no clear line with these things from my experience. I think we need to be
willing to take a look at what something is trying to solve, and able to take a
look around and see if there is already something solving it, or doing
something close enough to be easily adapted to fit our specific usage.

Even if there is a potential existing tool available, we also need to evaluate
whether that tools technology (programming language, platform, etc) fit and
whether its community is compatible enough with our. For example, are they
willing to work with outside consumers like us that may have some different
needs than their current user base? Are they an open community and not a vendor
of a proprietary tool?