Finish eliminating plugin usage in 'doctl sls functions' + testing improvements #1252
Conversation
|
This will stay as a draft until PR #1232 is merged since that PR has higher priority. Once 1232 is merged, this PR will be rebased and marked for review. |
joshuaauerbachwatson
force-pushed
the
function-list-no-plugin
branch
from
9d8bb61
to
c7e7be5
Compare
This completes the elimination of plugin usage in doctl sls fn and the functions.go source file.
joshuaauerbachwatson
force-pushed
the
function-list-no-plugin
branch
from
c7e7be5
to
8a26142
Compare
I believe this happened in merge conflict resolution during the recent rebase.
| return []whisk.Action{}, err | ||
| } | ||
| if limit == 0 { | ||
| limit = 30 |
There was a problem hiding this comment.
This was the default before the changeover. Maintaining it to avoid surprising new behaviors.
Affects what happens when a failure occurs in the middle of deleting functions and triggers together.
| nameSort, _ := c.Doit.GetBool(c.NS, flagNameSort) | ||
| nameName, _ := c.Doit.GetBool(c.NS, flagNameName) |
There was a problem hiding this comment.
This is not something new to this PR, but I wonder if we should hide one of these flags as they have the same purpose?
There was a problem hiding this comment.
The Apache OpenWhisk CLI (wsk) just uses --name-sort, short form -n. When Adobe issued their CLI (aio), it appears that they introduced the "alias" flag --name and they moved the -n shortcut to it. By itself, this flag is indeed kind of confusing (seems like it should take a value and allow you to specify a name, somehow).
When I incorporated parts of aio into nim I retained the alias behavior. When I created the doctl delegation path I again retained the behavior. In this PR, in recoding the logic in golang I again retained the behavior (copying is easier than thinking). I think it's appropriate to go back to what wsk does, which retains --name-sort and -n but not --name. This is technically a breaking change but I think it's extremely unlikely to matter to anyone.
@andrewsomething do you agree with that plan? If so, I'm inclined to just do it (by tacking the change onto this PR). What do you think?
There was a problem hiding this comment.
Ach. Since it is a behavior change, and technically breaking, I decided not to do it as part of this PR so I'm proceeding to merge. We can think about making the proposed simplification in the future.
This change covers internal improvements not intended to be visible to typical
doctlusers.doctl sls fn listdoctl sls statusdoctl sls connectpermitting connection to serverless clusters that are not in production (and hence unknown to the serverless portal). This is needed to support internal testing and debugging use cases for which we currently use thenimCLI (which should be phased out).