fix local message duplication if joining a group via invite

other fixes

res/example_config.json

@@ -22,13 +22,12 @@
 		}
 	},
 	"ImGuiFonts": {
-		"atlas_extra_text": "🥰💀✌️🌴🐢🐐🍄⚽🍻👑📸😬👀🚨🏡🐦‍🔥🍋‍🟩🍄‍🟫🙂‍↕️🕊️🏆😻🌟🧿🍀🎨🍜",
 		"size": 20,
 		"fonts": {
 			"entries": {
-				"/nix/store/7fjwhgbz16i08xm171arr081bqpivv7k-hack-font-3.003/share/fonts/truetype/Hack-Regular.ttf": true,
+				"/nix/store/ihpjyw2nvm924kf84898v9zqizhbwvn6-hack-font-3.003/share/fonts/truetype/Hack-Regular.ttf": true,
-				"/nix/store/g4hlmhda2xmap333kqnzlsz01k8djnp6-noto-fonts-24.3.1/share/fonts/noto/NotoSans[wdth,wght].ttf": true,
+				"/home/user/Downloads/TwitterColorEmoji-SVGinOT-15.1.0/TwitterColorEmoji-SVGinOT.ttf": true
-				"/nix/store/d7mgcvb59anvaz69cjghbb42616c7xfg-noto-fonts-monochrome-emoji-3.000/share/fonts/noto/NotoEmoji.ttf": true
+
 			}
 		}
 	}

src/bitset_image_loader.cpp

@@ -65,7 +65,7 @@ std::optional<TextureEntry> BitsetImageLoader::haveToTexture(TextureUploaderI& t
 BitsetImageLoader::BitsetImageLoader(void) {
 }
 
-TextureLoaderResult BitsetImageLoader::load(TextureUploaderI& tu, ObjectHandle o, uint32_t w, uint32_t h) {
+TextureLoaderResult BitsetImageLoader::load(TextureUploaderI& tu, ObjectHandle o, uint32_t /*w*/, uint32_t /*h*/) {
 	if (!static_cast<bool>(o)) {
 		std::cerr << "BIL error: trying to load invalid object\n";
 		return {};

src/chat_gui/contact_list.cpp

@@ -156,8 +156,6 @@ bool renderContactBig(
 		ImGui::EndTooltip();
 	}
 
-	ImVec2 post_curser_pos = ImGui::GetCursorPos();
-
 	ImVec2 img_curser {
 		orig_curser_pos.x + ImGui::GetStyle().FramePadding.x,
 		orig_curser_pos.y + ImGui::GetStyle().FramePadding.y

src/frame_streams/audio_stream2.hpp

@@ -24,6 +24,24 @@ struct AudioFrame2 {
 		Span<int16_t> // non owning variant, for direct consumption
 	> buffer;
 
+	AudioFrame2(void) = default;
+	AudioFrame2(const AudioFrame2&) = default;
+	AudioFrame2(AudioFrame2&& other) :
+		sample_rate(other.sample_rate),
+		channels(other.channels),
+		buffer(std::move(other.buffer))
+	{}
+	AudioFrame2(uint32_t sample_rate_, size_t channels_, const std::variant<std::vector<int16_t>, Span<int16_t>>& buffer_) :
+		sample_rate(sample_rate_),
+		channels(channels_),
+		buffer(buffer_)
+	{}
+	AudioFrame2(uint32_t sample_rate_, size_t channels_, std::variant<std::vector<int16_t>, Span<int16_t>>&& buffer_) :
+		sample_rate(sample_rate_),
+		channels(channels_),
+		buffer(std::move(buffer_))
+	{}
+
 	// helpers
 	Span<int16_t> getSpan(void) const {
 		if (std::holds_alternative<std::vector<int16_t>>(buffer)) {

src/frame_streams/locked_frame_stream.hpp

@@ -28,7 +28,7 @@ struct LockedFrameStream2 : public FrameStream2I<FrameType> {
 		FrameType new_frame = std::move(_frames.front());
 		_frames.pop_front();
 
-		return std::move(new_frame);
+		return new_frame;
 	}
 
 	bool push(const FrameType& value) {

src/frame_streams/sdl/sdl_audio2_frame_stream2.cpp

@@ -7,7 +7,7 @@
 #include "../audio_stream_pop_reframer.hpp"
 
 // "thin" wrapper around sdl audio streams
-// we dont needs to get fance, as they already provide everything we need
+// we dont needs to get fancy, as they already provide everything we need
 struct SDLAudio2StreamReader : public AudioFrame2Stream2I {
 	std::unique_ptr<SDL_AudioStream, decltype(&SDL_DestroyAudioStream)> _stream;
 

src/image_loader.cpp

@@ -6,9 +6,14 @@
 #include "./image_scaler.hpp"
 
 ImageLoaderI::ImageResult ImageLoaderI::ImageResult::crop(int32_t c_x, int32_t c_y, int32_t c_w, int32_t c_h) const {
-	// TODO: proper error handling
+	if (
-	assert(c_x+c_w <= width);
+		c_x < 0 || c_y < 0 || c_w < 1 || c_h < 1 ||
-	assert(c_y+c_h <= height);
+		int64_t(c_x) + int64_t(c_w) > int64_t(width) || int64_t(c_y) + int64_t(c_h) > int64_t(height)
+	) {
+		// unreachable
+		assert(false && "invalid image crop");
+		return *this;
+	}
 
 	ImageLoaderI::ImageResult new_image;
 	new_image.width = c_w;
@@ -19,7 +24,7 @@ ImageLoaderI::ImageResult ImageLoaderI::ImageResult::crop(int32_t c_x, int32_t c
 		auto& new_frame = new_image.frames.emplace_back();
 		new_frame.ms = input_frame.ms;
 
-		// TODO: improve this, this is super inefficent
+		// TODO: improve this, this is inefficent
 		for (int64_t y = c_y; y < c_y + c_h; y++) {
 			for (int64_t x = c_x; x < c_x + c_w; x++) {
 				new_frame.data.push_back(input_frame.data.at(y*width*4+x*4+0));

src/image_loader_stb.cpp

@@ -89,8 +89,8 @@ std::vector<uint8_t> ImageEncoderSTBPNG::encodeToMemoryRGBA(const ImageResult& i
 	struct Context {
 		std::vector<uint8_t> new_data;
 	} context;
-	auto write_f = +[](void* context, void* data, int size) -> void {
+	auto write_f = +[](void* context_, void* data, int size) -> void {
-		Context* ctx = reinterpret_cast<Context*>(context);
+		Context* ctx = reinterpret_cast<Context*>(context_);
 		uint8_t* d = reinterpret_cast<uint8_t*>(data);
 		ctx->new_data.insert(ctx->new_data.cend(), d, d + size);
 	};
@@ -126,8 +126,8 @@ std::vector<uint8_t> ImageEncoderSTBJpeg::encodeToMemoryRGBA(const ImageResult&
 	struct Context {
 		std::vector<uint8_t> new_data;
 	} context;
-	auto write_f = +[](void* context, void* data, int size) -> void {
+	auto write_f = +[](void* context_, void* data, int size) -> void {
-		Context* ctx = reinterpret_cast<Context*>(context);
+		Context* ctx = reinterpret_cast<Context*>(context_);
 		uint8_t* d = reinterpret_cast<uint8_t*>(data);
 		ctx->new_data.insert(ctx->new_data.cend(), d, d + size);
 	};

src/main.cpp

@@ -15,11 +15,13 @@
 
 #include "./start_screen.hpp"
 
+#ifdef __ANDROID__
 #include <filesystem>
+#endif
+
 #include <memory>
 #include <iostream>
 #include <string_view>
-#include <thread>
 #include <chrono>
 
 #ifdef TOMATO_BREAKPAD