Closed hexylena closed 7 years ago
@erasche You can disable the definitely neat plugin(s), which shouldn't be on by default. The up versus down is just strand info, but it may interfere with other stuff. I added "Neat" plugins to master (though they are not in 2.0.5 I think).
In you go into the apollo-config.groovy file you would comment out this section:
https://github.com/GMOD/Apollo/blob/master/grails-app/conf/Config.groovy#L406-L412
You may want / need to make sure if you do a clean-all that the plugins aren't included.
I updated the JBrowse version as well, but we shouldn't be forcing anything out, though I noticed there were some fixes to Canvas in there (hopefully those are still working).
@erasche If its still over-writing the colors even after you've removed the Neat Plugins, let me know. I'm not seeing those blues / blacks that you are off of master.
Rebuilding my apollo image now, thanks for the tip. That sounds like it's exactly what I need. I haven't made the update to 2.0.5 yet, hopefully will get to that in january.
That change was introduced in https://github.com/GMOD/Apollo/commit/17558740, looks like I haven't checked my apollo-config.groovy against the reference in a long time. Thanks!
Yep, confirmed. Thank you!
I noticed that on more recent builds of apollo (admittedly this hasn't been tested in a while), colors are not rendered quite properly for my canvas features (and completely disabled with HTML features due to style overrides...)
Correct output (as seen in JBrowse by itself):
With a hard-coded color in Apollo:
Hovering over them causes them to go blue and then grey
With the proper color function that works fine in JBrowse:
Here's my trackList.json
Tarball of data directory: https://cpt.tamu.edu/dot/blast-color-issue.tar.bz2
Is there any way we can just... completely disable Apollo styling on our own tracks? Especially for blast results, where the old "single-line-through-the-hit" was very clear. The new has people asking us questions they didn't before, misunderstanding how we're rendering the evidence.