Allegro.cc - Online Community

Allegro.cc Forums » Programming Questions » [A5] Mouse response is slow

This thread is locked; no one can reply to it. rss feed Print
[A5] Mouse response is slow
adamk kromm
Member #5,432
January 2005

Hi

I'm working on a map editor thing, and keyboard input works great and is responsive, but the mouse input is quite delayed.

Here is the scenario: I have a screen and when I click it puts a white box where the user clicks. The problem being after I click the mouse, there is a ~1 second delay before the box shows up. If I click and drag the mouse, all the boxes show up, but again, there is a 1-3 second delay for all of them to show up. The weird thing is, if I click the mouse and rather than dragging the mouse i use the keyboard to "move" the drawing plane then everything is immediately.

so in short: mouse input slow, keyboard input perfect...

I'm using Allegro 5.0.7 md
with a core i7 quad core
8gb ram
nvidia gtx 660m fully updated.
I'm using opengl 3.0 though i doubt that makes any difference.

Any ideas on how to fix it?

thanks adam.

----------
-Adam Kromm

My Website

Peter Wang
Member #23
April 2000

I couldn't understand what you wrote.

Are you using event queues, and letting the events pile up? Moving the mouse generates a lot more events than you would get with a keyboard.

Timorg
Member #2,028
March 2002

Are you draining the message queue every time? If you are only taking one message out each time, the mouse moving could be causing a flood of messages that have to come through before it gets to processing the mouse click.

____________________________________________________________________________________________
"c is much better than c++ if you don't need OOP simply because it's smaller and requires less load time." - alethiophile
OMG my sides are hurting from laughing so hard... :D

adamk kromm
Member #5,432
January 2005

You are both right.

I was only taking out one message at a time from the event queue, and thus it was backing up.

So speaking of having a backup in the event queue, is there a proper way to handle it? Or is it safe (and practical) to just take the first event off the queue and ignore the rest?

Thanks for the responses.
Adam.

----------
-Adam Kromm

My Website

J-Gamer
Member #12,491
January 2011
avatar

You don't want to miss an event, because then you're even worse off than with polling in A4. So you really have to parse every event in the queue before drawing a frame.

" There are plenty of wonderful ideas in The Bible, but God isn't one of them." - Derezo
"If your body was a business, thought would be like micro-management and emotions would be like macro-management. If you primarily live your life with emotions, then you are prone to error on the details. If you over-think things all the time you tend to lose scope of priorities." - Mark Oates

Arthur Kalliokoski
Second in Command
February 2005
avatar

I have this in my code

#SelectExpand
1 int redraw = 0; 2 while (true) 3 { 4 al_wait_for_event(queue, &event); 5 { 6 switch (event.type) 7 { 8 case ALLEGRO_EVENT_DISPLAY_CLOSE: 9 goto done; 10 11 case ALLEGRO_EVENT_KEY_CHAR: 12 { 13 int i; 14 blah 15 blah 16 break; 17 } 18 19 case whatever: 20 blah 21 blah 22 break; 23 24 case ALLEGRO_EVENT_TIMER: 25 { 26 draw some stuff 27 al_flip_display(); 28 frames++; 29 redraw++; 30 } 31 32 } 33 while( (redraw > 0) && al_is_event_queue_empty(queue)) 34 { 35 redraw--; 36 } 37 } 38done:

They all watch too much MSNBC... they get ideas.

J-Gamer
Member #12,491
January 2011
avatar

@Arthur: that way, you'll be drawing at each timer event, and that'll cause your program to lag behind if the drawing takes too long.

#SelectExpand
1while(!quit) { 2 al_wait_for_event(queue,&event); 3 switch(event.type) 4 { 5 case ALLEGRO_EVENT_DISPLAY_CLOSE: 6 quit = true; 7 break; 8 9 case ALLEGRO_EVENT_KEY_CHAR: 10 blah 11 blah 12 break; 13 14 case whatever: 15 blah 16 blah 17 break; 18 19 case ALLEGRO_EVENT_TIMER: 20 redraw = true; 21 break; 22 23 } 24 if(redraw && !al_is_event_queue_empty(queue)) { 25 //draw some stuff 26 al_flip_display(); 27 redraw = false; 28 } 29}

" There are plenty of wonderful ideas in The Bible, but God isn't one of them." - Derezo
"If your body was a business, thought would be like micro-management and emotions would be like macro-management. If you primarily live your life with emotions, then you are prone to error on the details. If you over-think things all the time you tend to lose scope of priorities." - Mark Oates

Arthur Kalliokoski
Second in Command
February 2005
avatar

Yeah, my event-fu is weak. :-/

They all watch too much MSNBC... they get ideas.

adamk kromm
Member #5,432
January 2005

Here is how I now have it, so that there is not mouse lag, (or any other that i can notice.)

I is set up so that it does logic updates 45x per second, and draws as fast as the computer will let it.

#SelectExpand
1bool running = true; 2 3const int FRAMES_PER_SECOND = 45; 4const double SKIP_FRAMES = 1.0 / FRAMES_PER_SECOND; 5const int MAX_FRAME_SKIP = 15; 6 7double nextFrame = al_get_time(); 8int frameSkip = 0; 9int fpsLogic, fpsDisplay; 10 11long long frames = 1; 12long long framesDisplay = 1; 13double startTime = al_get_time(); 14 15while(running) 16{ 17 frameSkip = 0; 18 19//loop as long as we havent skipped MAX_FRAME_SKIP frames, and we have extra time before the next 20//logic update is needed 21 while(frameSkip < MAX_FRAME_SKIP && al_get_time() > nextFrame) 22 { 23 //do input logic 24 ALLEGRO_EVENT ev; 25 while(al_peek_next_event(eventQueue, &ev)) 26 { 27 al_get_next_event(eventQueue, &ev); 28 if( ev.type == ALLEGRO_EVENT_KEY_DOWN || 29 ev.type == ALLEGRO_EVENT_KEY_UP || 30 ev.type == ALLEGRO_EVENT_MOUSE_AXES || 31 ev.type == ALLEGRO_EVENT_MOUSE_BUTTON_DOWN || 32 ev.type == ALLEGRO_EVENT_MOUSE_BUTTON_UP 33 ) 34 { 35 inputManager->ProcessEvent(ev); 36 } 37 } 38 39 running = HandleInput(); 40 Update(); 41 inputManager->Update(); 42 43 ++frames; 44 ++frameSkip; 45 nextFrame += SKIP_FRAMES; 46 } 47 48 float interp = (float)((al_get_time() + SKIP_FRAMES - nextFrame) / SKIP_FRAMES); 49 50 Render(); 51 ++framesDisplay; 52 53 //calculate frames per second 54 fpsLogic = (int)(frames / (al_get_time() - startTime)); 55 fpsDisplay = (int)(framesDisplay / (al_get_time() - startTime)); 56 57 std::stringstream ss; 58 ss<<"Maze Gaze - Map Editor | FPS (logic): "<<fpsLogic<<" (display): "<<fpsDisplay; 59 al_set_window_title(window, ss.str().c_str()); 60}

----------
-Adam Kromm

My Website

Thomas Fjellstrom
Member #476
June 2000
avatar

I is set up so that it does logic updates 45x per second, and draws as fast as the computer will let it.

What's the point of drawing if nothing has changed?

--
Thomas Fjellstrom - [website] - [email] - [Allegro Wiki] - [Allegro TODO]
"If you can't think of a better solution, don't try to make a better solution." -- weapon_S
"The less evidence we have for what we believe is certain, the more violently we defend beliefs against those who don't agree" -- https://twitter.com/neiltyson/status/592870205409353730

Matthew Leverton
Supreme Loser
January 1999
avatar

Don't you ever write code like this?

int x = 1;
x = 1;
x = 1; // make sure x REALLY is 1

adamk kromm
Member #5,432
January 2005

In this case you are right; in the map editor I have no need to update the display any more than when something changes.

However in the game (where I copied this from) things are a little different. I originally had it update the logic at 25 or 30 times per second. So all movement and animations where only updated at 25-30fps. So if I were to update the display only 25-30 times per second there would be a noticeable jerkiness. Rather than objects smoothly moving across the screen they would appear to skip or jump across the screen.

So by not limiting the display rate, I can calculate an interpolation value (not used in the code I posted because there is no animation nor movement) and interpolate the rendering between the last update and the next expected update. This helps the animations and movements look a lot smoother.

----------
-Adam Kromm

My Website

Thomas Fjellstrom
Member #476
June 2000
avatar

Yeah, I don't have anything against updating at the refresh rate, but any more than that is pretty silly as the display won't show any of it. Interpolating between frames is a pretty good idea, can help smooth things out, even if it does sort of lie about where things are on screen.

Also updating as often as you can wastes power. People on a laptop will be unhappy with you. And people with loud GPUs will be as well.

--
Thomas Fjellstrom - [website] - [email] - [Allegro Wiki] - [Allegro TODO]
"If you can't think of a better solution, don't try to make a better solution." -- weapon_S
"The less evidence we have for what we believe is certain, the more violently we defend beliefs against those who don't agree" -- https://twitter.com/neiltyson/status/592870205409353730

adamk kromm
Member #5,432
January 2005

I guess when I said: "as fast as the computer will let it" I was not very clear. I do have vsync turned on, so it will only redraw the screen at the vsync rate.

----------
-Adam Kromm

My Website

Thomas Fjellstrom
Member #476
June 2000
avatar

Though turning vsync on is more of a hint. Graphics drivers let you disable it system wide or individually per app/game.

I guess though that then its the users fault.

--
Thomas Fjellstrom - [website] - [email] - [Allegro Wiki] - [Allegro TODO]
"If you can't think of a better solution, don't try to make a better solution." -- weapon_S
"The less evidence we have for what we believe is certain, the more violently we defend beliefs against those who don't agree" -- https://twitter.com/neiltyson/status/592870205409353730

adamk kromm
Member #5,432
January 2005

I guess when I said: "as fast as the computer will let it" I was not very clear. I do have vsync turned on, so it will only redraw the screen at the vsync rate.

----------
-Adam Kromm

My Website

Thomas Fjellstrom
Member #476
June 2000
avatar

I guess when I said: "as fast as the computer will let it" I was not very clear. I do have vsync turned on, so it will only redraw the screen at the vsync rate.

Did you mean to double post that?

--
Thomas Fjellstrom - [website] - [email] - [Allegro Wiki] - [Allegro TODO]
"If you can't think of a better solution, don't try to make a better solution." -- weapon_S
"The less evidence we have for what we believe is certain, the more violently we defend beliefs against those who don't agree" -- https://twitter.com/neiltyson/status/592870205409353730

adamk kromm
Member #5,432
January 2005

Woops sorry, having some internet problems here at the university.

----------
-Adam Kromm

My Website

Go to: