Open Stack

On Thu, Sep 27, 2018 at 9:06 AM, Doug Hellmann <doug at doughellmann.com> wrote:
> I definitely think having details about the gaps would be a prerequisite
> for approving a goal, but I wonder if that's something 1 person could
> even do alone. Is this an area where a small team is needed?

Maybe, but it does break down along project/API lines for the most
part, only crossing in places like Matt mentioned where compute+volume
interact in server create, etc.

For the purposes of a goal, I think we need to be thinking more about
structural things than specific command changes.  Things like Monty
mentioned elsewhere in the thread about getting all of the exiting
client libs to correctly use an SDK adapter then behaviours converge
and the details of command changes become project-specific.

> We built cliff to be based on plugins to support this sort of work
> distribution, right?

We did, my concerns about splitting the OSC in-repo plugins out is
frankly more around losing control of things like command structure
and consistency, not about the code.  Looking at the loss of
consistency in plugins shows that is a hard thing to maintain across a
distributed set of groups.

>> One thing I don't like about that is we just replace N client libs
>> with N (or more) plugins now and the number of things a user must
>> install doesn't go down.  I would like to hear from anyone who deals
>> with installing OSC if that is still a big deal or should I let go of
>> that worry?
>
> Don't package managers just deal with this? I can pip/yum/apt install
> something and get all of its dependencies, right?

For those using that, yes.  The set of folks interacting with
OpenStack from a Windows desktop is not as large but their experience
is sometimes a painful one...although wheels were just becoming a
thing when I last tried to bundle OSC into a py2exe-style thing so the
pains of things like OpenSSL may be fewer now.

dt

-- 

Dean Troyer
dtroyer at gmail.com