ome / omero-web

Django-based OMERO.web client
https://www.openmicroscopy.org/omero
16 stars 29 forks source link

New Tag feature : Adding colors #483

Open Rdornier opened 1 year ago

Rdornier commented 1 year ago

Having different colors for tags could be a very good option to make the webclient more user-friendly (like in Trello cards). The colors could visually differenciate processes, categories, workflows... For example, blue tags could refer to the biological sample, green ones to the microscope acquisition and orange ones to proessing.

Forum thread : https://forum.image.sc/t/omero-tags-queries/82584

imagesc-bot commented 1 year ago

This issue has been mentioned on Image.sc Forum. There might be relevant details there:

https://forum.image.sc/t/omero-tags-queries/82584/4

will-moore commented 1 year ago

Items of work (if we decide to use LongAnnotations on a Tag to store a colour value)

jo-mueller commented 2 months ago

Is this still being worked on? I'm a bit late to the party, but I think a really straightforward way would be to give every tagset its own color. The latter, for starters, could be randomly chosen, as adding a dialogue for the color setting would probably be a bit more work.

will-moore commented 2 months ago

@jo-mueller it's not being worked on. I don't know if colour on Tagset would satisfy the requirements of most users (especially as many don't use Tagsets)? @Rdornier

What I proposed there was a possible approach to add colour support to Tags (without changing the underlying OMERO model). This has the advantage of not being a breaking change (wouldn't affect other clients or need a server upgrade etc) but still needs wider discussion and agreement before starting to work on it.

cc @knabar

Rdornier commented 2 months ago

I agree with @will-moore ; tagset are not widely used and I personnaly don't use them. I would strongly prefer to assign the color directly to the tag. I cannot really discuss on the implementation way. It will be better if the color can also be chosen at tag creation but if it is too complex to handle, then the proposed solution looks good to me as well

Tom-TBT commented 2 months ago

I side with @jo-mueller :D, I use tagsets a lot and recommend them too. Typically I use them for things like: Condition

or

Genotype

In those cases, I annotate datasets with one tag per category, so to me it makes a lot of sense to use one color for all tags of the same category.

But I don't see an issue having both colors on tags and tagsets (anyway, tagsets are just tags with a special namespace, right?) We can conciliate it like this: If a tag has a color annotation, it's displayed with this color. Else if it has no color annotation, it takes the color annotation of its tagset (so if you move a tag to a tagset, you would have to delete its color annotation to let it take the color from the tagset)

About the random assignment of colors, I'm less in favor. I would rather choose from predefined palettes (also ok to randomly choose from predefined palettes). We could propose two different palettes, one for the tagsets and one for the tags (could make it clear whether the tags are isolated or part of a set):

jo-mueller commented 2 months ago

Hi all,

thanks for the replies on that topic.

Maybe to motivate my idea regarding colors for taggroups a bit more: I basically started today with @zoccoler to create a set of standard tags for people who want to use our OMERO instance. For this we oriented ourselves along the REMBI metadata standards. Essentially, we turned every suggested metadata category into a set of tags for each category. In doing so, we found it quite helpful to organize things in taggroups, simply for the sake of better ordering a potentially large number of tags.

In that respect I don't have a strong opinion about the value of taggroups over just tags - giving all tags in a taggroup the same color would just be a nice feature to see possible redundancies in the used tags. For instance, the tags Confocal and SpinningDisk would (visually) belong in the same tag group. I don't think there is a collision of interests here.

On the implementation side, I think it would be helpful if the individual tag would inherit its color property from the respective taggroup to make the assignment of colors a bit handier. It doesn't prevent you from giving an individual tag an individual color, either.

Rdornier commented 2 months ago

If a tag has a color annotation, it's displayed with this color. Else if it has no color annotation, it takes the color annotation of its tagset (so if you move a tag to a tagset, you would have to delete its color annotation to let it take the color from the tagset)

Ok, it sounds like a good compromise 👍

I would rather choose from predefined palettes

I agree