Last call for 2.0.6 bugs!


This is in revision control, thanks!


Overscan mode/hint for Logical Renderer

This is in now, thanks!


Thanks. I could modify my test project to be something similar to the old ant example. I was already looking into it, but I wanted to do a bit more testing for a nicer way to include SDL. The android gradle plugin doesn’t quite support all the stuff needed to reference prebuilt native libraries without like I would have wanted.


I know I’m late to this but maybe this one?


Possible fix for:

Added a diff to the ticket


I came across a couple more android build issues. Created a bug report and a patch:


I know I’m late to this but maybe this one?

I think this is a good idea, but not for 2.0.6, since it would break binary compatibility.


Did the bug get deleted? I guess you added it instead to the todo list for 2.1?


Did the bug get deleted? I guess you added it instead to the todo list for 2.1?

It’s still there; when I quoted it, the copy/paste picked up the click counter and added an ‘8’ to the end of the URL in my post (I’ve edited my post to fix that).

(Bugzilla doesn’t allow one to delete bugs, as far as I know, even if they are just spam posts.)



There are some small cmake fixes like

would be cool to get them in.


I hope I’m not too late :slight_smile: 'd be cool if users of my engine wouldn’t have to compile a custom libsdl with this patch applied anymore; it’s rather minimal:

Edit: Looks like the CREATE_FLAGS macro changed quite a bit since then so the patch won’t apply cleanly; just need to add SDL_WINDOW_INPUT_FOCUS to the allowed create flags.


In Android SDL 2.0.5 the function:
no works well at fist time, I need used sometimes for desactivate the accelerometer complety. (Sorry by my English).


In Android SDL 2.0.5 not compile with NDK r14b. I have used the NDK r13b and it´s ok. (Sorry by my English)


You might be experiencing a bug in the clang compiler that is included in r14 NDK. Should be fixed in the latest NDK (r15).


This one is a limited bug, easy to reproduce and easy to confirm fixed. Also, there is no workaround. So it would be great, if you would consider fixing it:

The bug was not present in 1.2 and currently it prevents our game from switching to SDL2.

Sry. It was already (mostly) fixed in 2.0.5


These wrong SDL_TEXTINPUT events on windows are really annoying:


I was hoping to get this one in, if there’s still time!

On that note, what’s the status of the shaped window API? Bugs aside, it seems functional, but it’s not documented anywhere, and the API seems a bit iffy (have to include SDL_shape.h separately, camel-cased enums, functions returning error codes without calling SDL_SetError() etc).

It seems like another candidate for a clean-up for SDL 2.1.


Was wondering if anyone can reproduce this:
It seems serious if it’s not just my system. the bug is, Windows 10 + OpenGL + fullscreen desktop + a desktop resolution that’s not the LCD’s native resolution is broken (no highdpi scaling involved.) edit: appears to be nVidia-only, beacuse it goes away when I switch to a mesa software OpenGL dll.


I think I may have found a bug in SDL_OpenAudio(desired, obtained), whereby the wiki documentation doesn’t seem to match what SDL (both 2.0.5, and 2.x head) is doing. I’ve posted details on it, along with a possible fix, at

If this looks like a bug, any chance it could be considered for 2.0.6+?



Exiting fullscreen-desktop on macOS currently causes the application to freeze; could be reverted (this was an attempt to fix ) until we come up with a better fix?