ok after further investigation I know exactly how this happens:
you click “c” to split/cut.
you click a segment for red (or blue it doesnt matter which colour goes 1st for this bug, whether you do red > blue or blue > red it’s the same thing). Immediately after that 1st segment click the seg ID 0 gets unselected automatically. Hereafter if you select blue (or red if its your 2nd colour) and submit all is fine. Upon completion of the split the segment ID 0 gets automatically reelected.
but if instead you click clear (w/o selecting the 2nd colour and submitting) and/or if you select the 2nd colour and then clear w/o submitting.
voila segment ID 0 does not get automatically reselected and you need to do it manually.
Possible solutions:
→ remove split being able to deselect and/or reselect that option, making it effectively able to change only manually. (idk what purpose it serves to even have it deselected in either split or merge considering ID-less segments cannot be split or merged into/from anything in the 1st place.)
or
→ remove the option entirely from being clickable by humans and remove it from split/merge as we cant do anything w/ unsegmented IDs since again we cant do anything with those, and just make it perma enabled in the background w/o us or split being able to enable/disable it.
i am getting a bug where both segments get deleted after splitting a segment ( not in segment list) when clicking on the segment again in 2d it come back as splitted.
when pasting this link neuroglancer a new tab it looks to work correctly for a short time but soon get the bug again.
i’ve noticed this bug as well (in banc as well as pyr) since a bit before KK’s comment here: [addon] Undelete (BANC) - #9 by Krzysztof_Kruk I think it has something to do w/ neuroglancer’s new version and whatever they changed in there.
I think, this one might’ve been the fault of the Undeleted script (as @Nseraf correctly suggested). I’ve just released a new version and now it’s should work fine.
Hi All, is this bug still happening?
We had some issues with data storage for some datasets on Friday afternoon. I believe it is fixed now as of yesterday. Please let us know!