User Details
- User Since
- Mar 12 2021, 11:34 AM (4 w, 3 d)
Yesterday
@Daniel Bystedt (dbystedt) No problem, I've added the two decimal points for the status bar display and replaced the RNA definition for the angle increment with one designed for rotation/angles which means the knife tool settings and top bar now look like so:
Sat, Apr 10
@Pratik Borhade (PratikPB2123) I did indeed commit locally it would seem, this diff should have all of the correct changes.
Thanks @PatrikPB2123 I considered that however I feel it makes more sense to reset to default if the angle is set out of the bounds 0 to 90 degrees. However, if many feel this should work as you suggest I can indeed add it.
(I'm also noticing now I may have messed up this diff revision? Does it still contain my changes from my old diff? If not I have them saved locally so I can attempt to fix it)
Fri, Apr 9
Thank you so much for the feedback @Campbell Barton (campbellbarton) @Daniel Bystedt (dbystedt) ! I've implemented the following list of changes (copied from your comments):
Tue, Mar 30
For me it only works using the actual number keys along the top of the keyboard, not the numpad keys. This is intentional as such functionality is the same for tools like scale and rotate.
Sorry I realise I was unclear about that.
Mon, Mar 29
Wed, Mar 24
Thu, Mar 18
Sounds great!
Wed, Mar 17
Tue, Mar 16
From looking into it it seems to be caused by the function wm_operator_finished in wm_event_system.c being ran even if the user clicked on the hud window. I would imagine the fix for this would be to block this function from running if a hud window is being clicked? Perhaps by considering clicking anywhere within a hud window to not be a VIEW3D_OT_select operation. I'm not totally sure how to implement this but so perhaps someone more experienced can help out, however I will give it a go.
Sun, Mar 14
This appears to be caused by the LISTBASE_FOREACH loop in ed_undo.c at line 930 not considering the fact that two seperate objects can have the same object data. Because of this it only increases len on the first iteration, as it then updates the id->tag of the object data. Once this tag is updated, len does not change on the second iteration. However the object in the second iteration is obact, causing r_active_index to be set incorrectly.