Unexpected behaviour with keyboard checks in HTML5

[flerpyderp]

I have a game designed for Windows that I've made a HTML5 export version of just to share for testing etc. It appears as if keyboard_check_pressed functions are being treated like keyboard_checks instead.

For example, jumping is activated by a key press, but in the HTML5 version, holding jump causes the player to jump again upon landing.

I made a test project with only a room with one object which adds 1 to a counter each time a key is pressed and draws the value to the screen. Again, it appears to be checking for the key being held and adds 1 per frame. There is a slight delay between the first count, and all the ones that follow it.

Does anyone know why this might be happening? Since I can't find any information about this, I'm beginning to wonder if it's an issue specific to certain keyboards.

EDIT: Have now tested on a different keyboard, and it behaves the same.

 

[Mike]

If it's a bug... Do a very simple example, and file it with YoYo.

 

[chmod777]

Which version are you using? I know this bug did happen in older versions (1.4.9999 being affected as well).

 

[flerpyderp]

If it's a bug... Do a very simple example, and file it with YoYo.

I forgot to mention I am using 1.4.9999.

Which version are you using? I know this bug did happen in older versions (1.4.9999 being affected as well).

Others have reported not having this issue using 1.4.9999. Since a lot of people still use it, I would expect to find reports of other people having the same issue, but I couldn't find any.

 

[Mike]

Well, if it's 1.4 there's nothing you can do except work around it - or upgrade.

 

[rIKmAN]

Others have reported not having this issue using 1.4.9999. Since a lot of people still use it, I would expect to find reports of other people having the same issue, but I couldn't find any.

Just blew the dust off 1.4.9999 and tested this and can confirm I also get the bug where keyboard_check_pressed() acts like keyboard_check() when the key is held down when exporting to HTML5.

Have you tried with the previous version to see if it happens with that too?
Previous version was 1.4.1804 I think without checking?

It won't ever be fixed in 1.4.9999 so that might be your only hope without upgrading to GMS2.

 

[flerpyderp]

Just blew the dust off 1.4.9999 and tested this and can confirm I also get the bug where keyboard_check_pressed() acts like keyboard_check() when the key is held down when exporting to HTML5.

Have you tried with the previous version to see if it happens with that too?
Previous version was 1.4.1804 I think without checking?

It won't ever be fixed in 1.4.9999 so that might be your only hope without upgrading to GMS2.

I appreciate you testing this. I have not tried it with the previous version.

 

[TailBit]

Well, a workaround is to keep track of its last 2 states.

Create

key=0

Step

key = ((key&1)<<1) + keyboard_check(vk_anykey)

Then you can check if it is pressed, held or released with key==1, key&3 and key==2

I have an array that does this check for all my keys, and have scripts for the press/release checks with array position as argument