Community
Participate
Working Groups
Nightly builds as of Friday 27 May 2016 with bug 476984 completed. In the Type column in the Protocol Message parameters table, for primitive types from the (default language) primitive types library, it is enough to only display their names (no need to for the fully qualified name, and it would probably just be confusing). Note that qualified names should still be displayed even for these primitive types in the content assist proposals (bug 000000). This is for display in the cell only.
(In reply to Christian W. Damus from comment #0) That should be: > Note that qualified names should still be displayed even for these primitive > types in the content assist proposals (bug 494812). This is for display in > the cell only.
(In reply to Christian W. Damus from comment #1) > (In reply to Christian W. Damus from comment #0) > > That should be: > > > Note that qualified names should still be displayed even for these primitive > > types in the content assist proposals (bug 494812). I'm not fully sure that this is needed. I think that it might be enough that primitive types (from the primitive types library) neither are shown with qualified name in the content assist popup, nor in the table itself. That is how the legacy tooling behaves. But I do know that there have been cases where users have selected the wrong primitive type from the wrong primitive types library. How they have done it though is a bit unclear. The legacy tooling have multiple primitive type libraries, which normally should never be used at the same time. And maybe to avoid such issues, the fully qualified name should indeed be shown in the content assist window. But on the other hand, if the user writes an unambiguous name and this completes directly without the content assist window popping up (Bug 494814) the it does not matter that the fully qualified name is displayed or not in the content assist popup window. And with this reasoning, the fully qualified name should *also* be displayed in the table even for primitive types. But I really don't think that will be a good solution either, since in the normal case the library name will be superfluous and will just clutter the table. So maybe the proposal it is best to follow the proposal of Bug 494812, and always display the fully qualified name in the content assist proposal, both for primitive types and user defined types.
Does not gate v1.0
