Will there be any attempt to extract an interface from these or otherwise make common attributes across all types? I am referring to the comment you left in another thread where you might want to request an EmptyDir with specific Throughput performance characteristics.

Assigning to @erictune, feel free to rebalance.

There may be common interfaces across them, but I want to use NFS and iscsi
as drivers for that, rather than just guessing.

On Fri, Feb 20, 2015 at 5:28 AM, Mark Turansky notifications@github.com
wrote:

Will there be any attempt to extract an interface from these or otherwise
make common attributes across all types? I am referring to the comment you
left in another thread where you might want to request an EmptyDir with
specific Throughput performance characteristics.

Reply to this email directly or view it on GitHub
#4653 (comment)
.

LGTM. Not passing tests.

looks like there is a new precommit hook. maybe a rebase will help

rebased and pushed.

On Fri, Feb 20, 2015 at 10:40 AM, Eric Tune notifications@github.com
wrote:

looks like there is a new precommit hook. maybe a rebase will help

Reply to this email directly or view it on GitHub
#4653 (comment)
.

@derekwaynecarr FYI

@thockin LGTM with the exception of renaming SecretType to SecretKind.

@thockin Rereading the comment on #4514 to make sure all feedback was accounted for here, I found this comment #4514 (comment). Tl;dr: need description on fields.

Pushed update with SecretKind. It looks like descriptions are OK now - I don't find any missing.

@thockin One last nit :) @bgrant0607 wanted SecretKind to appear before Data.

Little problem: if we rename Secret.Type to Secret.Kind we now conflict with TypeMeta.Kind, which is embedded in Secret. If we rename it to SecretKind then it is different from a dozen other "Kind" fields, which just happen to not be on API objects.

@bgrant0607 for an opinion on what to do with this.

Annoyingly, this is only a problem in the Go inline structs, not in the json. Taking a look.

Propose to drop the Type/Kind from this series and deal with it on its own? If OK, I'll push a new update.

Also we have AffinityType floating around

And LimitType, FSType, CapabilityType, CauseType.

Kinds include only PodConditionKind and NodeConditionKind.

So I guess it's the Kinds that are odd.

If you want to revert to SecretType, I'm ok with that.

I think this latest push addresses all comments except Kind/Type which we can reconsider on its own.

@thockin LGTM 👍

Thanks for fixing these nits.

LGTM accomplished. Will self-commit when green. Will think on type/kind.