Open MichaelKetting opened 9 years ago
I believe this was already reported yesterday
Great. Thanks. I did look but looks like for the wrong keywords :) Do you happen to have an issue number?
Okay, it happens if you have 3 project: project 1 contains the type as public. project 2 contains the type as internal (same namespace as the public one). project 3 depends on project 1 and 2 and takes a cref-dependency on the member in the type.
Out of interest, why are explicit crefs such as M:Remotion.Globalization.IResourceManager.TryGetString(string, out string)
not a part of the reference analysis? I can't see a good reason for it not to be.
When I am writing crefs I sometimes use the compiler generated IDs, so I would write M:Remotion.Globalization.IResourceManager.TryGetString(System.String,System.String@)
instead, why are these references not analysed and marked as warnings if they are found to be invalid (either in format or not referencing anything)?
In line 124 of https://svn.re-motion.org/svn/Remotion/trunk/Remotion/ObjectBinding/Web/UI/Controls/BocList.cs I have
<see cref="IResourceManager.TryGetString"/>
. The type "IResourceManager" exists in my solution and is referenced. The method TryGetString also exists and is not ambigous on the type. Still, I receive the following warning:Fully qualifying the reference does not have an effect:
<see cref="Remotion.Globalization.IResourceManager.TryGetString(string, out string)"/>
Setting the reference to hardcoded
<see cref="M:Remotion.Globalization.IResourceManager.TryGetString(string, out string)"/>
works, but then it's no longer part of the reference analysis.I believe there is some weird edgecase in the reference analysis that broke this reference. I haven't yet had time to try to set up an isolated solution containing just the two types involved, so I don't know just how much ambient noise is required to trigger this problem but the repository is open source.