servoz
commented
11 months ago For the mia_processes bricks, we don't expose the plugs corresponding to the parameters related to the internal machinery and we don't index these parameters in the database.
For other bricks, such as pure nipype processes, they are exposed until now.
I've had a look at what happens when we drop a brick into the pipeline editor, and in fact we only have control over the process in populse_mia once the node has been created in capsul. This makes it impossible to change the exposure of certain plugs in the node, unless we change things in capsul, which I don't think is desirable here. We've always said that pure nipype bricks are offered as they are, so this may make sense.
On the other hand, I'm going to make sure that machinery-related outputs (such as the pyscript.m file) aren't indexed in the database, but that they appear in the history (which could be interesting if we want to look in detail at the history of the parameters that created a data).
There is now an entry in the History tag with the history button for nipype script file.
1/ this script should not be indexed in the project (it's an internal code file that doesn't interest the user, although this can be discussed, by the way I don't remember if we made sure not to index it like we do with the mia_processes bricks?).
2/ but more importantly I think that there shouldn't be the button to access the history for this file.