On 2/14/19 3:30 PM, Dean Troyer wrote:
On Thu, Feb 14, 2019 at 9:29 AM Lance Bragstad <lbragstad@gmail.com> wrote:
On 1/31/19 9:59 AM, Lance Bragstad wrote: Moving legacy clients to python-openstackclient
Artem has done quite a bit of pre-work here [2], which has been useful in understanding the volume of work required to complete this goal in its entirety. I suggest we look for seams where we can break this into more consumable pieces of work for a given release.
For example, one possible goal would be to work on parity with python-openstackclient and openstacksdk. A follow-on goal would be to move the legacy clients. Alternatively, we could start to move all the project clients logic into python-openstackclient, and then have another goal to implement the common logic gaps into openstacksdk. Arriving at the same place but using different paths. The approach still has to be discussed and proposed. I do think it is apparent that we'll need to break this up, however.
Artem's call for help is still open [0]. Artem, has anyone reached out to you about co-championing the goal? Do you have suggestions for how you'd like to break up the work to make the goal more achievable, especially if you're the only one championing the initiative? I'll outline my thoughts on how to break these down in that etherpad. Fortunately there are a lot of semi-independent parts here depending on how we want to slice the work (ie, do everything for a small number of projects or do one part for all projects).
I am planning to scale back some involvement in StarlingX in 2019 to free up some time for this sort of thing and am willing to co-champion this with Artem. I'm likely to be involved anyway. :)
Awesome - that'll be a huge help, Dean! If you need help getting things proposed as a goal, just let me or JP know.
dt