Sorry your browser is not supported!

You are using an outdated browser that does not support modern web technologies, in order to use this site please update to a new browser.

Browsers supported include Chrome, FireFox, Safari, Opera, Internet Explorer 10+ or Microsoft Edge.

AppGameKit Classic Chat / Process terminated with status -1073741819 (0 minutes, 2 seconds)

Author
Message
Phaelax
DBPro Master
21
Years of Service
User Offline
Joined: 16th Apr 2003
Location: Metropia
Posted: 7th May 2013 03:42
That's the error code I got when trying to access an array incorrectly.

I had:


Should have been:



Just seemed like a rather odd error message to get in the console in bright red letters.

"You're all wrong. You're all idiots." ~Fluffy Rabbit
easter bunny
11
Years of Service
User Offline
Joined: 20th Nov 2012
Playing: Dota 2
Posted: 7th May 2013 04:32
I've had similar problems before as well, but I've forgotten how they happened, probably the same reason though

Naphier
13
Years of Service
User Offline
Joined: 2nd Oct 2010
Location: St Petersburg, Florida
Posted: 7th May 2013 05:47
My favorite error!
Happens for a variety of reasons.
UDT Array out of bounds is one reason
Issue 495

Another reason is putting variables other than strings where strings are expected in some commands.

I know I've seen it happen for a couple other reasons, but can't remember. I hope that any new debugging features from AppGameKit fix get us some more error codes!

Phaelax
DBPro Master
21
Years of Service
User Offline
Joined: 16th Apr 2003
Location: Metropia
Posted: 7th May 2013 07:45
I was lucky and managed to find the bug almost immediately. Unlike another typo I had which wasted an hour of my life tonight. Some advice, never hard-code values where you plan to replace them with constants later.

"You're all wrong. You're all idiots." ~Fluffy Rabbit
Hodgey
14
Years of Service
User Offline
Joined: 10th Oct 2009
Location: Australia
Posted: 7th May 2013 10:13
I could be wrong about this but I think the -107.... error is code::block's way of telling you that an abnormal exit occurred. AppGameKit is supposed to catch those errors but should it miss any that causes the program to crash or terminate abnormally, I believe code::blocks prints that number.

MarcoBruti
12
Years of Service
User Offline
Joined: 20th Nov 2011
Location: Caput Mundi
Posted: 7th May 2013 11:13
Quote: "I could be wrong about this but I think the -107.... error is code::block's way of telling you that an abnormal exit occurred. AppGameKit is supposed to catch those errors but should it miss any that causes the program to crash or terminate abnormally, I believe code::blocks prints that number."

I am happy to see the some AppGameKit bugs are so permanently built-in inside the product, that people have lost the memory of the original source of the errors and begin inventing new interpretations.
In fact that error is due to a codeblock issue. Try to add/subtract randomly dummy lines, and the error disappears.
No way to have this issue solved in 1.8???
Hodgey
14
Years of Service
User Offline
Joined: 10th Oct 2009
Location: Australia
Posted: 7th May 2013 13:24 Edited at: 7th May 2013 15:05
Quote: "I am happy to see the some AppGameKit bugs are so permanently built-in inside the product, that people have lost the memory of the original source of the errors and begin inventing new interpretations. "

Firstly, I said I "could be wrong".

Secondly, those were based off my observations of Code::blocks when using AppGameKit and when not. When programming in C++ (no AppGameKit at all) using code blocks, I received that error message (or something very similar) when my app didn't terminate normally. I've also received that error or similar errors when I've had AppGameKit programs crash. I haven't "invented" this theory out of thin air.

I might have misinterpreted when the error occurred (I'm assuming compile time now rather than runtime as I did earlier) but I thought my two cents would have been worth adding; apparently not.

Edit- It's been a long day, sorry if I took this out of proportion.

Cliff Mellangard 3DEGS
Developer
18
Years of Service
User Offline
Joined: 20th Feb 2006
Location: Sweden
Posted: 8th May 2013 22:38
I really want a proper debugger and error codes on the console.

van b and my Project have growed alot and have alot of code.

suddenly so did the app constantly crash.

van and i code differrently and i tracked it down to image id handeling and how agk deletes them on exit etc.

I ended up with Writing my own complete auto id for images and image commands to get rid of these crashes.

To my knowing so whas it how we mixed auto ids and manuall ids and loaded them in an mixed order.

This happens as van! maybe writes an file with all inventory code and i write the core Engine and when we merge these files together so is the ids not created Always in an order of 1-1000 or 10000-10000.

Only an heads up for anyone Writing large Projects with alot of files included.

Android 2.3 , ZTE Skate , 480x800 , 800 mhz , Samsung Galaxy Y , 240x320 , 832 mhz , Sony ericson arc 480x854 , 1 ghz
Android 4.0 , Dmtech 3g 9738B , 1024x768 , 9.7 inches , 1.2 ghz
Phaelax
DBPro Master
21
Years of Service
User Offline
Joined: 16th Apr 2003
Location: Metropia
Posted: 9th May 2013 03:01
Sounds like you guys need to agree on a standard methodology. Perhaps consider building some UML diagrams so when you're working apart you know what the functions take and what they should return.

Honestly, I can't see why anyone would want to manually control object IDs when the system can handle it for you already.

"You're all wrong. You're all idiots." ~Fluffy Rabbit
Cliff Mellangard 3DEGS
Developer
18
Years of Service
User Offline
Joined: 20th Feb 2006
Location: Sweden
Posted: 9th May 2013 10:38 Edited at: 9th May 2013 10:46
Quote: "Sounds like you guys need to agree on a standard methodology."

And thats why i wrote our own auto image load id.

Wherever the command is called so does all images get ids in sequence.

1-1000

Agks own auto id starts at 10000-

And all subimages also grabb a id when created.

We have alot of subimage calls where we replace subimages on 3d objects.
sometimes did i only neaded to simply replace an image with the same id or get a new one.

This is because we sometimes change a parent object or clone Another object.

While using agks auto id so whas we forced to keep an track off the image ids with globals to delete them manually or use the delete all image command.

And agk delete all even deletes all images used by virtual buttons and such.

And if you would replace an image id that agk use internally by mistake so will your app crash when agk wants to use an virtual button or Another agk command that uses images etc.
Or simply display them with no image and crash later.

The image commands are good for smaller Projects and our Project is not small.

And untill all agk issues are solved so have i bin Writing my own workarounds.

There is no sader thing then spending 2-4 Days trying to solve a bug and find that its an agk bug and not your code.

Here is how we now load all images and the sample is for subimages.

IMG_Door = SK_SUBIMAGE__LOAD_( MainTexture , "door1.png" )

As you see do i mimic agks own command but get auto ids starting with 1.

and i can keep simple imag usage checks by simply print(SK_IMAGE__COUNT)

And now when i have replaced all agk image load commands so did we both use auto ids and agk still crashed after that,so probably a good thing i wrote my own temporary auto image id.

The problem with large Projects is that you get alot of media and the simpliest way would be that all starts from 1 for when deleting them.

The app crashes where probably simply that we used an id used by agk itself that made it crash?

Quote: "I can't see why anyone would want to manually control object IDs when the system can handle it for you already."

Or your own function

Ps...
the SK in the command starts is for our Engine name
SpaceKitten.

And here is how simple i Clean all our own media without deleting agk internall media to avoid crashes.

REM // Used to make a clean image media cleanup //////////
Function SK_IMAGE__DELETE_ALL_()
img=0
while doThis=0
inc img
if GetImageExists(img) then DeleteImage(img)
if SK_IMAGE__COUNT=<img
SK_IMAGE__COUNT=0
exit
endif
endwhile
Endfunction

Android 2.3 , ZTE Skate , 480x800 , 800 mhz , Samsung Galaxy Y , 240x320 , 832 mhz , Sony ericson arc 480x854 , 1 ghz
Android 4.0 , Dmtech 3g 9738B , 1024x768 , 9.7 inches , 1.2 ghz
MarcoBruti
12
Years of Service
User Offline
Joined: 20th Nov 2011
Location: Caput Mundi
Posted: 9th May 2013 20:33
2 simple comments:
- mixing auto-id and custom ids it is not a robust way to program. Choose in advance which approach you will use (auto-id or custom ids) and stick to it during all development phases.
- do not assume that AppGameKit uses id from 10000+. It could change in the future and again it is not robust programming.
BTW, in your opinion are there some hopes to have an AppGameKit finished product before the end of this year? Apart my problem with my old laptop with Intel Graphics not compatible with OpenGL 2.0, I think that no serious Tier-1 programmer (as we are) could rely on a product that will never become stable. No serious project can be started, etc. And, as you can see, bugs are corrected at a "snail rate" or even not corrected. Debugger and Editor are still at the early stage, etc. I do not know if this product will have some future apart TGC internal use...
Phaelax
DBPro Master
21
Years of Service
User Offline
Joined: 16th Apr 2003
Location: Metropia
Posted: 9th May 2013 21:06
Quote: "SpaceKitten"

I am intrigued.

"You're all wrong. You're all idiots." ~Fluffy Rabbit
Cliff Mellangard 3DEGS
Developer
18
Years of Service
User Offline
Joined: 20th Feb 2006
Location: Sweden
Posted: 10th May 2013 14:58 Edited at: 10th May 2013 14:59
Quote: "- mixing auto-id and custom ids it is not a robust way to program. Choose in advance which approach you will use (auto-id or custom ids) and stick to it during all development phases."

It where more in the begining when van made alot of new stuff and we hadt talked so much on how to use ids

But my experience is that it whas image ids that caused our crashes.
I have rewritten large parts and the id usage is down to 400-600 per level.
Before my rewrite so did each new enemy frame get a new id each time it whas swapped and not a good thing.
The agk limitations made me optimize my code instead as it whas sloppy

Then are these image ids with the subimages as we animate our enemys by swapping subimages.

There is no animate texture command in agk.

Quote: "- do not assume that AppGameKit uses id from 10000+. It could change in the future and again it is not robust programming."

I belive this is a safety thing by the tgc team as most users that manually set an id starts with 1 or values lower then 10000.

The images that agk uses starts somewhere around 20000 or 30000?

Quote: "BTW, in your opinion are there some hopes to have an AppGameKit finished product before the end of this year? Apart my problem with my old laptop with Intel Graphics not compatible with OpenGL 2.0, I think that no serious Tier-1 programmer (as we are) could rely on a product that will never become stable. No serious project can be started, etc. And, as you can see, bugs are corrected at a "snail rate" or even not corrected. Debugger and Editor are still at the early stage, etc. I do not know if this product will have some future apart TGC internal use..."

I see no problem in agk so far as i have a great comunity to ask when i have problems
But yes tgc would nead Another coder to help out with agk devlopment to speed things up .
But i Think paul is doing a great job so far and there silence does probably mean they are soon to release a new stable version.
I belive they want to suprise us with the next release
Number one on the wish list is some better debugging and error reports.

Quote: "I am intrigued."

We will soon release our first demo and start a proper thread in the showcase
I have overwhelmed van b with art requests so far and nead to wait for the latest art

Android 2.3 , ZTE Skate , 480x800 , 800 mhz , Samsung Galaxy Y , 240x320 , 832 mhz , Sony ericson arc 480x854 , 1 ghz
Android 4.0 , Dmtech 3g 9738B , 1024x768 , 9.7 inches , 1.2 ghz
Phaelax
DBPro Master
21
Years of Service
User Offline
Joined: 16th Apr 2003
Location: Metropia
Posted: 14th May 2013 06:56
I found something else that causes this same error, using another function call as a return value.



"You're all wrong. You're all idiots." ~Fluffy Rabbit
Ancient Lady
Valued Member
20
Years of Service
User Offline
Joined: 17th Mar 2004
Location: Anchorage, Alaska, USA
Posted: 14th May 2013 17:40
I'm willing to bet that Tier 1 doesn't handle using the return value of another function in the 'endfunction' statement.

I suspect you would be happier in Tier 2 where all these things are possible.

The limitation in Tier 1 is that everything must go through an interpreter and if they haven't considered the scenario someone uses, it will cause an error.

Cheers,
Ancient Lady
AGK Community Tester and AppGameKit Master
Phaelax
DBPro Master
21
Years of Service
User Offline
Joined: 16th Apr 2003
Location: Metropia
Posted: 15th May 2013 00:59
I've been considering using Tier 2 and I may switch for my next project.

"You're all wrong. You're all idiots." ~Fluffy Rabbit
Ancient Lady
Valued Member
20
Years of Service
User Offline
Joined: 17th Mar 2004
Location: Anchorage, Alaska, USA
Posted: 15th May 2013 05:17
Tier 2 gives you much more control over things and access to lots of standard C and C++ stuff.

But, the Sync() command only gets user inputs once each app::Loop() cycle. So you can't do simple quick loops that have something displayed and then wait for a user input before processing. You end up doing a variety of state machine type things if you need to wait for specific inputs (like selecting from a menu).

And you do get a good debugger in the Visual Studio IDE (VS Express 2010 works just fine with AGK).

It also allows you to do customizations within your code for the different platforms while still using the same code files (using #ifdef's).

Cheers,
Ancient Lady
AGK Community Tester and AppGameKit Master

Login to post a reply

Server time is: 2024-05-06 18:22:03
Your offset time is: 2024-05-06 18:22:03