+1 to stronger authN FYI I've quickly tried readthedocs -- see https://readthedocs.org/projects/brooklyn-docs/ and it is nice and free, maybe we should consider it for the website. It is an hosted service well integrated with github which offers full text seatrch out of the box. Many others ASF projects already use them, it's maybe worth a try.
My two cents, Andrea On 5 October 2017 at 12:07, Thomas Bouron <[email protected]> wrote: > I would agree with Geoff on the auth. I think it would be nice to move to > JWT for 1.0. > > I would also point out the Brooklyn website. I'm the one to blame on this, > I redesigned the home page but not the rest (mostly due to the currently > complexity of it) > I'll send a proposal on this today. > > Best. > > On Wed, 4 Oct 2017 at 17:57 Geoff Macartney <[email protected]> > wrote: > > > +1 with observation that there are still some things we might want to do > > before making that 1.0 release. (In re. Richard's comment about "Are we > > feature complete", you may think we are not.) > > > > For example something I think it would be good to do before announcing a > > 1.0 is to get away from basic auth on the UI and REST API and having a > > session token based approach, perhaps based on JWT tokens. (Don't like > > storing the auth credentials for the CLI) > > > > That's one example but you may have your own suggestions - shout out! > > > > Geoff > > > > > > > > On Wed, 4 Oct 2017 at 16:41 Richard Downer <[email protected]> wrote: > > > > > +1 with conditions :-) > > > > > > Yes, I'd like to see Apache Brooklyn 1.0 released. I'd suggest that we > > > would get ASF PR involved to generate a bit of buzz around this, and > > > possibly approach some of our commercial users and friends to chip in > > with > > > some PR too. > > > > > > But going to 1.0 (and generating some media buzz around it) brings > > > responsibilities. If the media buzz brings in new users, we want to > make > > > sure our usability is as high as possible - this covers the app itself, > > as > > > well as the website and the documentation. > > > > > > So to answer your question with some more questions... > > > > > > Are we feature-complete? > > > > > > Do we have great usability? Good user stories for all our types of > users? > > > > > > Are we happy to freeze our API in its current state? > > > > > > Are we happy to accept a stricter deprecation policy going forward? > > > > > > Is there nothing that we want to deprecate before 1.0? (That would > imply > > > another 0.x release cycle) > > > > > > Have we removed every deprecated thing that we can? (If something is > > > deprecated but cannot be removed, why?) > > > > > > Do we have great documentation? > > > > > > Do we have a great website? > > > > > > Does our documentation and examples reflect the "best" (not deprecated, > > > outdated or suboptimal) way of doing things? > > > > > > Will the blueprints in the wider community be compatible with the > > proposed > > > 1.0 release or do they need updating? (We will need to work with those > > > blueprint owners to get the blueprints updated.) > > > > > > Are we prepared to personally get involved in a media and visibility > > push? > > > (e.g. using our own Twitter, networks, getting more people involved in > > > managing the official Apache Brooklyn social media channels, etc.) > > > > > > Are we prepared - in the event of a successful media blitz - to handle > > more > > > users coming to this list, IRC, Twitter etc. looking for help? > > > > > > > > > If we can answer "yes" to all of these questions, then we are ready to > > > release 1.0. If any of these questions is answered "no" or "maybe" then > > we > > > should wait, or consider making a 0.13.0 release first. > > > > > > Richard. > > > > > > > > > On 3 October 2017 at 16:33, Duncan Godwin <[email protected]> > wrote: > > > > > > > Hi All, > > > > > > > > Following the release last week of Apache Brooklyn 0.12.0 I propose > > that > > > we > > > > make the next version of Apache Brooklyn 1.0.0. > > > > > > > > Apache Brooklyn is robust, stable, feature rich and being used in > > > > production by multiple enterprises. Our deprecation policy means we > > > haven't > > > > treated it like a 0.x release in a very long time. > > > > > > > > With 0.12.0, we did the last major thing needed before a 1.0 release: > > we > > > > switched to Karaf as the primary distribution and we deprecated the > > > > "classic mode". What does everyone think? > > > > > > > > Many thanks > > > > > > > > Duncan > > > > > > > > > > -- > > Thomas Bouron • Senior Software Engineer @ Cloudsoft Corporation • > https://cloudsoft.io/ > Github: https://github.com/tbouron > Twitter: https://twitter.com/eltibouron >
