You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Lib389 has a legacy Entry interface. It's implemented in a really complex and difficult to understand manner. This interface is deprecated, and we have wanted to remove it for sometime. In wanting to remove it, we have tried to ask people to stop using the _s apis because it adds more techdebt on an interface we don't want.
Developers and QE want to be able to control certain ldap operations though. They want to use the _s apis. But this would make it harder and harder to remove this piece of techdebt.
I have resisted people using the _s ldap apis because I don't want more usage of Entry, but we also can't remove it easily because we would break some legacy tests.
As a middle ground it was suggested that:
We remove the meta layer that turns ldap raw calls into Entry wrapping and unwrapping calls.
We add compat functions for _s_entry that access and does the entry wrap and unwrap
we rework dsldapobject to use the _s calls directly without Entry (IE the dict ldap api).
This way people could use the ldap calls on dirsrv they want for tests, and we move closer to identifying and removing the Entry api.
The text was updated successfully, but these errors were encountered:
Cloned from Pagure issue: https://pagure.io/389-ds-base/issue/50600
Issue Description
Lib389 has a legacy Entry interface. It's implemented in a really complex and difficult to understand manner. This interface is deprecated, and we have wanted to remove it for sometime. In wanting to remove it, we have tried to ask people to stop using the _s apis because it adds more techdebt on an interface we don't want.
Developers and QE want to be able to control certain ldap operations though. They want to use the _s apis. But this would make it harder and harder to remove this piece of techdebt.
I have resisted people using the _s ldap apis because I don't want more usage of Entry, but we also can't remove it easily because we would break some legacy tests.
As a middle ground it was suggested that:
This way people could use the ldap calls on dirsrv they want for tests, and we move closer to identifying and removing the Entry api.
The text was updated successfully, but these errors were encountered: