• Hello [name]! Thanks for joining the GMC. Before making any posts in the Tech Support forum, can we suggest you read the forum rules? These are simple guidelines that we ask you to follow so that you can get the best help possible for your issue.

OFFICIAL 2.3.4 Stable Release Thread

Status
Not open for further replies.

RIGmix

Member
Since the 2.3.4.577 update whenever I pull over to the right and to the bottom of a collision block, depending on how stretched it is, I get trough it.
If the block isn't streched i don't get trough it.
In this image you can see the collision (red line) and see how streched the block.
You can see aswell that the player (fox) gets trough the more streched one and stays on top of the other.
Have this happen to anyone else besides me?
View attachment 42894
Also i already updated gamemaker to the 2.3.4.580 version and it ins't working either

collision code
Code:
 if place_meeting(x + hsp,y,o_col_parent) {

            while (!place_meeting(x + sign(hsp), y, o_col_parent)) {

                x = x + sign(hsp);

            }

            hsp = 0;

            walljumped = false;

            canwalk = true;

        }



        x = x + hsp;



        if place_meeting(x, y + vsp,o_col_parent) {

            while (!place_meeting(x, y + sign(vsp), o_col_parent)) {

                y = y + sign(vsp);

                dust = 0;

            }

            if dust = 0 {

                audio_play_sound(a_floor,0,0);

                if !place_meeting(x,y,o_agua) && place_meeting(x, y + 1,o_col_parent)  {

                    instance_create_layer(random_range(bbox_right-3,bbox_right+3), bbox_bottom, "Particles", o_po);

                    instance_create_layer(random_range(x-3,x+3), bbox_bottom, "Particles", o_po);

                    instance_create_layer(random_range(bbox_left-3,bbox_left+3), bbox_bottom, "Particles", o_po);

                    ScreenShake(3,1);

                }

            }

            dust ++;

            vsp = 0;

            walljumped = false

            canwalk = true;

        }
set hsp code
Code:
if (!walljumped && wall_jump_cooldown >= wall_jump_set_cooldown) {

            if (key_right && canwalk = true) {

                if (hsp < walkspd + ehsp && hsp >= ehsp ) hsp += 0.5;

                if (hsp > walkspd + ehsp && (place_meeting(x,y+1,o_col_parent) || place_meeting(x,y+7,o_slope_parent) || (place_meeting(x,y+10,o_slope_parent) && hsp > 7))) hsp -= 0.3;

                if (hsp < ehsp) hsp = ehsp + 1;

            }

            else if (key_left && canwalk = true) {

                if (hsp > -walkspd + ehsp && hsp <= ehsp) hsp -= 0.5;

                if (hsp < -walkspd + ehsp && (place_meeting(x,y+1,o_col_parent) || place_meeting(x,y+7,o_slope_parent) || (place_meeting(x,y+10,o_slope_parent) && hsp > 7))) hsp += 0.3;

                if (hsp > ehsp) hsp = ehsp - 1;

            }

            else if (!key_right && !key_left) {

                if (abs(hsp) > abs(ehsp)){

                    if abs(hsp) > abs(ehsp) + 0.5 {

                        if (hsp > ehsp) hsp -= abs(hsp)/6; else hsp += abs(hsp)/6;

                    } else hsp = ehsp;

                } else hsp = ehsp;
 

kingyo

Member
The snippet shortcuts are broken.
キャプチャ.PNG

As I posted before, Ctrl+T in the room editor does not open the Go To window . Instead it creates a tiled layer.
 
I posted this in the tech support forum, but someone asked me to post here. So I was hoping that today's update would fix the IDE issues from the previous update (2.3.4.577), but that's not the case. I'm still having the same issues as before.

1. If I use the paint bucket tool in the sprite editor, then the sprite vanishes and the entire window is resized and everything is out of place. This is how it looks like after I use the paint bucket tool on a sprite. I have attached it with this post. As you can see everything has vanished.

2. When I drag a tab from the initial workspace to move it to another monitor, the tab moves in the opposite direction, i.e. if I drag the tab to my right monitor, then the tab moves left and vice versa. The same thing happens if I move it up or down. Here is a gif of the bug.


Note that none of these issues happen with 2.3.4.574. Please fix these and other issues. There might be more that I am not currently aware of.

For now, I'm reverting back to 2.3.4.574.
 

Attachments

rIKmAN

Member
I posted this in the tech support forum, but someone asked me to post here. So I was hoping that today's update would fix the IDE issues from the previous update (2.3.4.577), but that's not the case. I'm still having the same issues as before.

1. If I use the paint bucket tool in the sprite editor, then the sprite vanishes and the entire window is resized and everything is out of place. This is how it looks like after I use the paint bucket tool on a sprite. I have attached it with this post. As you can see everything has vanished.

2. When I drag a tab from the initial workspace to move it to another monitor, the tab moves in the opposite direction, i.e. if I drag the tab to my right monitor, then the tab moves left and vice versa. The same thing happens if I move it up or down. Here is a gif of the bug.


Note that none of these issues happen with 2.3.4.574. Please fix these and other issues. There might be more that I am not currently aware of.

For now, I'm reverting back to 2.3.4.574.
Don't forget to file them as tickets too, that's the only way to guarantee that YYG take a look and add them to the bug tracker if required.
You can use the link posted in your other thread, or use Help > Report a Bug in the IDE.
 

Fiskyfisko

Member
Since the 2.3.4.577 update whenever I pull over to the right and to the bottom of a collision block, depending on how stretched it is, I get trough it.
If the block isn't streched i don't get trough it.
In this image you can see the collision (red line) and see how streched the block.
You can see aswell that the player (fox) gets trough the more streched one and stays on top of the other.
Have this happen to anyone else besides me?
Yep same problem. Stretched objects (like 10x and more of the original size) have broken collisions, bottom and right sides seem to be affected.

Going back to older runtime fixes it.
 
Last edited:

bodziozet

Member
Hi currently im having IDE version 2.3.4.580 and runtime 2.3.4.442. My previous problem was fps issues and this version fixed it perfectly but now i have even wore issue. I don't see any Rooms in my editor cannot see Rooms and edit them. And even when i try to create new project I cannot see the Room. I don't know if i should report it because there are some room issues in the headline known issues but they are supposedly fixed? Edit. The room is actually in there but is not visible inside the IDE. I know because when I hover and click the mouse over the area where the Room should be rendered I can move objects inside.RoomIssues.png
 
Last edited:

JadeMonsuta

Member
Supposedly IDE v2.3.4.580 and Runtime v2.3.4.442 were meant to fix a bug regarding projects not loading at all in GMS2.
However, ever since I installed this update none of the assets in my game appear to be listed in the IDE at all, but are still present in the file directory when viewed using windows explorer.
I'm unable to test the game, run the debugger, or even open up the asset browser.

I get the following text in the output, and nothing else.
Verifying views...
No root view detected... automatically creating one
...done
Nobody else on my team is having this problem at the moment it seems, only me, and they're running the same versions I'm running.
Help would be much appreciated here ^^"

Edit:
Changing the underscores in the YYP to spaces seems to have fixed the problem for me, which is weird because this issue hadn't existed before.
 
Last edited:
using 2.3.4.580 on MacOS

Haven't had an issue until today. Now I cannot import sprites or even create sprites.

The sprite editor shows the created or imported content in the layer icon, but it is not reflected in the main sprite window or the view in the sprite overview window.

All previously created sprites are fully viewable and re-importing the same sprites that were used before functions, but anything new breaks and presents a blank image.

V/R,

Revoincubus
 

Speederman

Member
I used to access structs variables this way:
GML:
my_struct.my_var
After updating, it still works on VM targets, but it throws a runtime error on YYC targets (without any further explanation, so I had it hard to find the issue). I had to change every reference like this and now it works fine on both outputs:
Code:
my_struct[$ "my_var"]
Now I'm facing a new issue that is driving me crazy. I have a new operator to create a struct that worked perfectly before the update. Now it still works on VM targets but again crashes on YYC. The runtime error doesn't show any info either and it crashes even if I use try / catch. This is the message I get:

Runtime Error!
Program:
This application has requested the Runtime to terminate it in an unusual way.
Please contact the application's support team for more information.


Has anyone else had this kind of issue?

EDIT: The error didn't show me any information because I was using exception_unhandled_handler() to manage exceptions, although it should have shown a debug message that didn't appear. The problem with the new operator is that it says that the function is not a constructor, but it is indeed. It works on VM, but not on YYC for some weird reason. 🤷‍♂️

EDIT 2: Kinda solved it. I just duplicated the function definition and renamed it (just renaming didn't work). Using the duplicated name works fine now.
 
Last edited:

Flaick

Member
Updated to 580. For me the issue of the double window still remains. When I'm using a second window maybe for editing sprites, the sprite disappers. Also sometimes when I press "5" in the script editor, instead it writes a black little square.
 

Flaick

Member
I confess I'm quite scaried about these updates. I have already 3 projects selling and I'm hoping these issues affect only the IDE and not the output build.
 

Stra

Member
Personally, if there is a release with lots of changes I copy my project and install the new version on the reserve computer and test everything there. I use only the stable version on my primary. Saves headaches.
 

Surlent

Member
The performance of 2.3.4 seems declined very much
In my stg game with about 2000 track bullets on screen,which is cpu heavy,useing 2.3.4 makes FPS drops about 15% compare with 2.3.3. I attached the extrame case bellow.
Reason might be the IDE problem posted here before,or some other problem.
 

Attachments

GamerXP

Member
I didn't find it in changelog, but did the way object's methods are called was changed?



<ref>.<function> before was called with <ref> being "self". Now it's not.

For the code above, Runtime 2.3.3.437 returns "Object", for 2.3.4+ it returns "Struct". This broke a lot of code in my project.
 

chamaeleon

Member
I didn't find it in changelog, but did the way object's methods are called was changed?



<ref>.<function> before was called with <ref> being "self". Now it's not.

For the code above, Runtime 2.3.3.437 returns "Object", for 2.3.4+ it returns "Struct". This broke a lot of code in my project.
You are not declaring customFunc as a method, hence the execution of it will not run in the context of the instance, despite using the dot notation. The dot notation in this case will simply fetch a normal function and execute without any knowledge that it happened to come from an instance (and in your code picks up the name value from its current context which is str). Either of these two ways should give the behavior you expect:
GML:
customFunc = method(id, _customFunc);
or
GML:
customFunc = function() {
    show_debug_message(name);
}
I realize the second way of writing seems not really different from how you did it, but as far as I am aware, this has been how methods have worked from the beginning. You need to declare the method right at the point of defining its name. Just assigning a function defined elsewhere (even if it is in this same event) is not sufficient, hence the need for a method() function to add a method to an instance based on an existing function.
 
Last edited:

GamerXP

Member
Well, I'm fine with either way, but changing how it works all of sudden without even mentioning it in release notes sounds like it was not intentional and should be either reverted, noted or made an option in game settings. I mean, it literally broke my project all of sudden and I had to debug it to find what in the world happened there.
 

chamaeleon

Member
Well, I'm fine with either way, but changing how it works all of sudden without even mentioning it in release notes sounds like it was not intentional and should be either reverted, noted or made an option in game settings. I mean, it literally broke my project all of sudden and I had to debug it to find what in the world happened there.
I'm afraid I don't have an answer for you. All I can say is that the current behavior is what I have experienced the whole time since functions and structs were introduced. Why you would have a different behavior for some earlier release, I don't know.
 
:(
I just upgraded to the newest stable and it won't run. I think I upgraded from 2.3.3 (which worked fine) to 2.3.4. When I double click the GMS2 icon it appears on my taskbar and right when it is about to display something onscreen it quits with no error message. The screen fills with white/gray that would be filled in with the launch menu over time, but it quits before that happens. I've tried reinstall and restarting the computer and none of that changes the behavior. My desktop is running Win 10, and is an Intel i7-2600k.
:(
 
Last edited:

mekaerwin

Member
I updated to the latest version this morning and the copy and paste blur problem still exists.
Does it also blur if you draw a line in the sprite editor (GM's)? I think I have the same problem. I also submitted a bug report. Have they responded to yours? For me, it means no art can be added to the game for anything other then temporary testing.
 

DiMND

Member
Just updated to latest version on my Mac and everything in the IDE is now massive- all of the fonts, icons etc. It looks like I've gone back to 1995. Is there a setting or something I can change to set it back to how it was in the previous version?

IDE v2.3.4.577, Runtime v.2.3.4.440
macOS 11.5.2 (Big Sur)
I have the opposite problem on my Mac. Everything is super small also curious if there is a setting or do I need to downgrade.
 

DiMND

Member
When I run the debugger it won't let me click inside the ide after running my project. I have to force quit and reload. Works fine running the project like normal. I'm running on Mac.
 

Gunner

Member
Is it just me or custom code snippets can no longer be pressed by hotkey? You have to move the cursor to click it.
 

Mool

Member
Guys is it possible, that google play exports arent doable anymore? I get now an error (32bit not allowed).

Before everything worked

google error.PNG
android_options.PNGndk_sdk.PNG
 
Last edited:

O.Stogden

Member
Just to throw my hat into the ring to say I also am struggling with low FPS in the IDE as of 2.3.4.

When opening assets or panning in the workspace, FPS drops below 30 and it stutters.

Looking at task manager, it appears that a single thread gets put under heavy load just having the IDE open.

The last thread of the CPU in this image is the one being used by GM's IDE, GM is constantly using between 3-6% CPU (so 36-72% of a thread), even if nothing is happening within the IDE. The spikes up to 100% are whenever I click an asset to open it up, or hold down the mouse wheel and pan the camera around.

GPU usage doesn't exceed 50% at any point, so it's definitely a CPU issue.

1632051260678.png
 

FrostyCat

Member
Guys is it possible, that google play exports arent doable anymore? I get now an error (32bit not allowed).

Before everything worked
If that was what it used to be, it should have stopped being accepted as early as August 2019.

You checked "Build for x86". That is the codename for 32-bit Intel processors. Right now your Intel-bound binaries are 32-bit only, triggering the error.
 

Smacktard

Member
My numpad commands (home, end, etc.) still not working with latest update. I'd gotten a ticket saying the issue was resolved. It's not.
 

Mool

Member
@FrostyCat Ah yeah you are right x86 is 32bit.

Strange, that i never got any issues with it before. I will test it and let you know.


EDIT: WORKS! Thanks!

Super strange, that it worked before.
 
Last edited:

ScottD

New Member
YYG Staff
IDE v2.3.4.583 Runtime v2.3.4.443 is currently being deployed - It is a small update that should address the issues that some of you have been having with graphical artefacts and performance in the IDE.

As always please do file a ticket if you encounter any issues within the new IDE. Also please do give the beta a go and see if things are better there as this will help us identify issues.

We are aware that Steam is not updated to this release yet, it is a work in progress but may take a while.
EDIT: Steam versions are now live
 
Last edited:

gnysek

Member
small update that should address the issues that some of you have been having with graphical artefacts and performance in the IDE
And that's the spirit! One quick hotfix on first day after weekend, and even more improvements to double-check on beta. I hope that this time more people will check beta (which can be installed side by side, so it doesn't removes 2.3.4), to prevent issues like this in 2.3.5, as it's completely understandable that YYG isn't able to check every possible hardware configuration, and it's good to gather data directly from users. That's also saves time, as stable version is still in version we want.
 
On 2.3.4.583 I still have very low framerates in the IDE compared to 2.3.3.574. It's especially noticeable when dragging the workspace view around but I can notice the lag when typing.
 

Flo

Member
I don't see any update related to keystore generation/Show Key Hash issues.
If I generate a key, it works but the software feezes.
If I import my key and enter the right informations, Show Key Hash freezes the software and it doesn't work.
I tried generating the key via Android Studio but Show Key Hash still freezes GMS.

So I'm stuck and cannot build my APK at all since GMS tells me I don't have a key configured.
 

kingyo

Member
IDE v2.3.4.583 is still laggy. I feel it especially when I'm editing code. Even just moving the cursor causes a lag.
I'm not sure when it started, maybe since 2.3.3, but the game frequently becomes laggy. This is something that was not seen in previous versions.
I would like to revert to the previous version, but I can't because I have rewritten all the functions to fit the new version.
Is the current GMS in a transitional phase where the backend is being changed drastically? I'm waiting for a quick improvement and stability.
 
Last edited:
Does it also blur if you draw a line in the sprite editor (GM's)? I think I have the same problem. I also submitted a bug report. Have they responded to yours? For me, it means no art can be added to the game for anything other then temporary testing.
I reverted to a version from June or something and created a new layer and it unblurred the art. I've never had anything work like that before.
 

RHANITAN

Member
Since the 2.3.4.577 update whenever I pull over to the right and to the bottom of a collision block, depending on how stretched it is, I get trough it.
If the block isn't streched i don't get trough it.
In this image you can see the collision (red line) and see how streched the block.
You can see aswell that the player (fox) gets trough the more streched one and stays on top of the other.
Have this happen to anyone else besides me?
View attachment 42894
Also i already updated gamemaker to the 2.3.4.580 version and it ins't working either
I am having the same issue with collision with any stretched collision objects which is affecting all my hitboxes as well since I create them by stretching a single pixel object to whatever size I need. Hoping this gets fixed as it's pretty much broken my whole collision and combat system. The 2.3.3.437 runtime seems to be the lastest runtime that doesn't have these issues for me.
 

angelwire

Member
I've filed a bug report, just wanted to share in case anyone else was having similar issues.

Using instance_place_list() with a precise collision mask seems to have gotten much slower in the new runtime. (Tested on Windows VM)

I have a game that runs with a huge drop in framerate compared between 2.3.3 and 2.3.4
Using the old runtime or a version not at 2.3.4 the game leave me in that place about 800fps, with 2.3.4 less than 300.
Anyone made some stress test to see the performance ?
Have you isolated what's causing the drop in framerate? Maybe you could use the profiler to see if any collision functions you're using have gotten slower?
 

vdweller

Member
@Dan

I had submitted a ticket regarding the following issue: On my Nvidia RTX 2060 GPU (Acer Nitro Laptop), if I loop a sprite animation by setting its image speed to zero and then every step I do:

image_index+=0.2;

The animation plays only once and then stops looping correctly. However I never got any response about this. Do you have anything newer on this?
 
Status
Not open for further replies.
Top