> > > > 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? __________________________________________________________________________ OpenStack Development Mailing List (not for usage questions) Unsubscribe: [email protected]?subject:unsubscribe http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
