[ckan-dev] Future, flask, breaking things, funding.

[Ross Jones]

Hi Steven,

> On 14 Sep 2015, at 22:11, Steven De Costa <steven.decosta at linkdigital.com.au> wrote:
> 
> 1. Perhaps redev could be bottom up. Start with resources and widen its ability. Crud can then be rebuilt over the top. 
> 2. Carefully consider the longest term possible and how the app may mature in the future.
> 3. Consider interoperability between n+1 platforms via linked open data, again with realtime in mind
> 4. Consider packages further. Could we add new package types that are built on 3.0 thinking and have them co exist with current packages? If so then existing extensions could be modified less dramatically to apply only to v2 packages.

I think this is one of the key points, how much change can we make without breaking the ecosystem of extensions that people have worked on, are in use etc.  It'll be quite hard to break the model in any significant way because we've no idea who the extensions use that model.  There is already some guaranteed breakage because there are things that extensions authors tend to import directly that may not be there in v3 (pylons.config being the first thing that comes to mind).

> 5. Think about migration scenarios. Could a v2 CKAN remain as a dumb web app harvesting from a 3.0? If so, we could priorities workflows around custodians and ETL before end users. 

I think a v3 CKAN would have to support the action API, so harvesting should/could still work as it does today. And including machines as users in a user-needs type approach to v3 would be very welcome.  Some components in various repos have user-stories written, but I don't think these exist for CKAN itself.  It would be awesome to get more of these from outside - As a machine, I want ..., so that ....

> 6. Yes I'm sure others in the steering group would support the work. Just remember they are also just volunteers :)

I do - and thank you for volunteering and engaging so frequently (on here and on twitter and elsewhere). It really is appreciated!

> 7. Yes I'm sure funding could come from the Association, just so long as funding first goes into the association. So, we'd all have a part to play in signing up paying members - happy to take any leads from people on that point :)

It's unfortunate that the Assoc doesn't have access to funds, obviously donating resources is very valuable but I'm not sure that will scale enough - everyone who is 'donated' to CKAN also has other things that are also priorities :( I'm not particularly clued-in as to the story with grants and applications for them, but if there's some other useful way I could help, I will.  I seem to recall the original plan was for some people to be actually employed by the association - is this no longer the plan?

Even if v3 is only the same as the jump from 1.x to 2.x (which would be fine, OKF invested a huge amount in that work and it was definitely worth it) - we still need to find a way to do https://github.com/ckan/ideas-and-roadmap/issues/152

Would also love to see the use-cases doc you mentioned in another email, it'll probably serve as a good starting point for finding user-needs and from there, user stories.

Cheers

Ross