For several reasons those changes makes sense. They are being made because making some of those API public.
Only past users of imgui_internal.h with the extra parameters will be affected.
Added asserts for valid flags in various functions to detect _some_ misuses, BUT NOT ALL.
Amend 4448d975 (#456, #2637, #2620, #2891, #3370, #4828, #5108, #5242, #5641)
Rename IsNamedKeyOrModKey() to IsNamedKeyOrMod() for consistency.
Fixed GetKeyName(ImGuiKey_None) from returning "N/A" or "None" depending on IMGUI_DISABLE_OBSOLETE_KEYIO.
See "inputs_keychord_name" in imgui_test_suite.
currently: ImGui::SetNextWindowRefreshPolicy(ImGuiWindowRefreshFlags_TryToAvoidRefresh);
- This is NOT meant to replace frame-wide/app-wide idle mode.
- This is another tool: the idea that a given window could avoid refresh and reuse last frame contents.
- I think it needs to be backed by a careful and smart design overall (refresh policy, load balancing, making it easy and obvious to user).
- It's not there yet, this is currently a toy for experimenting.
My other issues with this:
- It appears to be very simple, but skipping most of Begin() logic will inevitably lead to tricky/confusing bugs. Let's see how it goes.
- I don't like very much that this opens a door to varying inconsistencies
- I don't like very much that it can lead us to situation where the lazy refresh gets disabled in bulk due to some reason (e.g. resizing a dock space) and we get sucked in the temptation to update for idle rather than update for dynamism.
- Simplify and compact some code. Shallow tweaks.
- Add comments.
- Add concave shape demo.
- Remove coarse culling.
- Remove nested types to match coding style and for consistent type nams when translated to other languages.
- Merged ClassifyNode() and ReclassifyNode().
- Extracted ImTriangleIsClockwise().
- Hold copy of points inside nodes instead of pointing to them.
Very highly requested feature (#6962, #5219, #3290, #4627, #5054, #3878, #2881, #1506, #1216, #968).
Also useful for interactive completion/selection popups (#2057, #718)
Based on @kudaba PR. Design for Inputtext V2 should make this obsolete.
There are two issues here - first, this macro uses AT&T specific syntax with $, which is not necessary. Also, some assemblers (nasm) emit different bytes for "int 3" and "int3", so it's better to use "int3" (cd 03 vs cc)
More importantly, GDB has some failing assertion whenever stepping after hitting an "int3" instruction. This makes it practically useless, as is. For some reason, putting a nop afterwards as a workaround is okay.
Related discussions:
https://sourceware.org/bugzilla/show_bug.cgi?id=31194https://lists.sr.ht/~skeeto/public-inbox/%3C2d3d7662a361ddd049f7dc65b94cecdd%40disroot.org%3E
