Open ctrueden opened 1 year ago
gselzer
commented
1 year ago @ctrueden I assume this is also a problem in Fiji, yeah? napari-imagej doesn't do anything special with search result categories, there's just a single category for each Searcher implementation. Shouldn't we fix it upstream?
ctrueden
commented
1 year ago In practice, it's not a problem in Fiji because you can differentiate them via the icons—hence my linking to #284 here. But yeah, changes on the Java side might be warranted here: e.g. if we were to finish tackling imagej/imagej-legacy#264, it would change the dynamics of ImageJ vs SciJava commands under the hood, and probably we would want corresponding changes to the searchers to account for that. Are you OK with leaving this issue open on the napari-imagej side until it becomes viable for users to tell the difference between original ImageJ commands and SciJava modules? I'll change the title accordingly.
gselzer
commented
1 year ago Are you OK with leaving this issue open on the napari-imagej side until it becomes viable for users to tell the difference between original ImageJ commands and SciJava modules? I'll change the title accordingly.
Yup - I'm 100% on board with filing a PR to add icons to the search results - and if not, we could probably move this issue elsewhere.
One possible solution: the ImageJ command results and SciJava module results could be split to separate tree branches, rather than mixed together under a single "Commands" tree branch. Alternately or in addition, solving #284 would help mitigate the confusion, making this issue less of a concern to me. But I think even with icons it might be nice to separate them out. Or: we could make each branch have an icon, and separate the original ImageJ commands to their own branch, which I think would be clear enough. We can strike a balance between ease of implementation versus reducing ambiguity and user confusion.