include can cause problems about daemonmode.jl HOT 3 CLOSED

Thank you for the feedback. I have not notice it. I will try to fix it.

from daemonmode.jl.

I think I discovered an easy fix!!
Just add the following at the beginning of your program:

include(arg)=Base.include(@__MODULE__,arg)

and everything seems to work.
[Note this is not function overloading - just appending a method for this situation]

I had a related problem, which I have been trying to fix similarly, but cannot.

When I try to load a saved Neural Network model (built in Flux) using BSON,
I get an undefined reference to NNlib, even if I explicitly load NNlib via 'using NNlib'.
This does not happen in an ordinary sess of Julia

Can you think of any approach to help?

Thanks for any ideas.

BTW this is a great idea and module!

PS I managed to get around the BSON problem by saving/loading with the Serialization
module, and it all now works great!

from daemonmode.jl.

Thank you for your solution. I have already included your proposal in the server DaemonMode, and I have register a new version with the fix. Also, now when there is any error the complete stack is shown.

Now you can use "include" function directly in the scripts that use DaemonMode, you do not need to define your include function, DaemonMode is doing that for you before the file is run.

from daemonmode.jl.