Forums

Everything crashes.

• WGTicon 12,511 Posts

  Fri, Oct 2 2015 3:54 PM

  hi guys,

  if possible, please right click after the crash and submit a detailed report.we are trying to understand the nature of the crash.

  -wgticon

• mkg335 5,491 Posts

  Fri, Oct 2 2015 4:37 PM

  Icon...first, thanks for replying.

  Problem is, when we get "unexpected error" or the white (!) screen, it's too late to submit a bug report.

  Do you know if the programmers have looked at the post from CanineSupervisor on how they can fix the Flash memory leak?  I think that's essential for them as a way to solve the problem.

• JLeary1 416 Posts

  Sat, Oct 3 2015 12:42 AM

  Yes I second that - it is too late when a white screen with a "!" crash comes, by then the right click option is gone..

  Also, whilst memory leaks are an issue, they aren't always the cause of this crash. Why? Firstly, as an example, I was playing a skins game on merion. My approach shot landed about 7 foot above the hole, and when I putted, it missed, started rolling off the green, and then I got the "!" crash. So I cleared my flash and browser caches, and then restarted my browser. Got back into the game, has a chip shot, it again went past the hole in a similar spot to before. Took the putt, missed the hole again, and it gave me another "!" crash. So if it were a memory leak, it occurred after 2 shots only! (remember I cleared all my data).

  Also, several programmers on the adobe forums have reported being able to reliably reproduce what they call the GCOD error (grey circle of death) - aka this "!" error. And one way is to set a vector to a negative value. Either way, this bug is costly. Hope wgt can look into it properly.

• alosso 21,039 Posts

  Sat, Oct 3 2015 1:35 AM

  WGTicon:

  if possible, please right click after the crash and submit a detailed report.

  How would you expect to get it? When Flash crashes, no more Flash pop-up menu is available for reporting.

  @JL: Can you name the hole and the position?

  That might give WGT a chance for two distinctive spots to look at, together with OLY #14 ( link to thread).

• alanti 10,564 Posts

  Sat, Oct 3 2015 2:37 AM

  mkg335:

  Icon...first, thanks for replying.

  Problem is, when we get "unexpected error" or the white (!) screen, it's too late to submit a bug report.

  Do you know if the programmers have looked at the post from CanineSupervisor on how they can fix the Flash memory leak?  I think that's essential for them as a way to solve the problem.

  I would like to congratulate WGT........my last game had 8....yes 8 crashes, or the circle kept spinning, white screen or unexpected error.

  Each time I duly cleared caches etc.

  Why the *** do we have to climb through hoops when we pay to play the game....we must be barking mad.

  So as Mark politely asked, did the programmers look at Lees suggestion, or did that go right over their heads?

  Wake uo before you lose more players,

  For the record, I tried 2 multi games early, both could not finish. IMO it is not worth the frustration in multi player games at the moment, but single player games are no better.

• JLeary1 416 Posts

  Sat, Oct 3 2015 4:42 AM

  alosso:

  @JL: Can you name the hole and the position?

  Merion Hole #12 - pin front (on this hole the pin is either back right or middle front).

  Have a putt from around 6 foot directly top-side (towards the back) from the pin. For me I missed the putt twice, once before clearing the caches and once 2 shots after, and both times I got the GCOD.

  I have literally had this crash 100's of times in the past month or so. It used to crash a lot just as someone's ball was about to land, now it crashes under a zillion other conditions.

• JLeary1 416 Posts

  Sat, Oct 3 2015 4:46 AM

  alanti:

  So as Mark politely asked, did the programmers look at Lees suggestion, or did that go right over their heads?

  Any programmer worth their salt knows about memory leaks and how to avoid them, you guys do also realise that Adobe (makers of flash) have published the fact that even an idle flash run-time will increase memory of at least 5MB/minute, more for larger implementations.

  I'm not saying memory leaks are not a problem, just stated repeatedly here and in other threads that if after 2 shots (after clearing all data) there is an "out-of-memory" exception, then there is something seriously wrong. Also like I have said numerous times here in the forums that many professional adobe flash developers have noted that a reliable way to cause this error (as an example) is to set a vector variable to a negative value.

• FarFoo 2,916 Posts

  Sat, Oct 3 2015 6:09 AM

  WGTicon:

  hi guys,

  if possible, please right click after the crash and submit a detailed report.we are trying to understand the nature of the crash.

  -wgticon

  
  
  Are you saying that after weeks of this you don't know the nature of the problem ?

  The nature of the crashes are that your code breaks flash, almost certainly because of bad code not cleaning up assets in memory locations . ( which would cause the instant crashes and help explain why there is ram build up over time )
  
  Maybe if one of you  wgt people actually play a few games you will see for yourself.... just sayin.

  You cannot right click when your code breaks flash itself.... so NO , you will not get reports.
  
  I'll re post this for you from caninesupervisor. Third time this has been posted for you , please read it.. or pass on to someone who can.
  

  CanineSupervisor:

  Let me see if I can shed some light on this subject for you guys.

  Pay attention WGT, since you willfully disregard helpful emails suggesting how to fix the memory issue. You would never take advice from a former Adobe Computer Scientist, would you? Because, What does he know? Right?

  When any Adobe Flash app. frees a ByteArray, the reference to the array is erased from the domainMemory object, but when the free is done by a worker (a virtual instance of the Flash runtime. Each worker instance controls and provides access to the lifecycle and shared data of a single worker.), then the domainMemory is never informed and you have what is commonly called a Dangling pointer (a pointer that does not point to a valid object of the appropriate type.)

  A worker allows you to execute code "in the background" at the same time that other operations are running in another worker (including the main .swf's worker). In a non-worker context, (some operations), i.e., processing large sets of data in a loop, take so long to execute that they prevent the main application thread from updating the screen quickly enough. This can cause stuttering or freezing the screen and the White / Blue Screens the members are getting.

  Dangling pointers come about during object destruction, when an object that has an incoming reference is deleted or deallocated, without modifying the value of the pointer, so that the pointer still points to the memory location of the deallocated memory.  The system may reallocate the previously freed memory to another process, if the original program then dereferences the (now) dangling pointer, unpredictable behavior may result, as the memory may now contain completely different data. This is especially the case if the program writes data to memory pointed to by a dangling pointer - a silent corruption of unrelated data may result, leading to subtle bugs that will be extremely difficult to find.

  Hint: ByteArray : : Clear ( )... You guys should know the remaining code without me posting it in here. (CLEAR YOUR CACHE, in your own words.)

  The game is overfilling our gas tanks and now fuel is spilling out onto the driveway because someone wasn't standing by the nozzle. That's why we are all getting the dreaded Flash memory ! overload.

  just sayin'....

  Back in my hole, I go.

   

  
  
  .
  
  
  
  

• JLeary1 416 Posts

  Sat, Oct 3 2015 7:24 AM

  FarFoo:

  Are you saying that after weeks of this you don't know the nature of the problem ?

  The nature of the crashes are that your code breaks flash, almost certainly because of bad code not cleaning up assets in memory locations

  The nature of the problem? Yes there are probable memory leaks, but after 2 shots how many destructors have been called and not set their pointers to NULL?

  The GCOD error can and is caused by a myriad of things, only one of which is not setting pointers to NULL when a destructor is called..

  Seeing as so many are hell bent on believing that fixing the memory leaks will solve the GCOD crash (which it won't in most cases), and since no-one is bothering to listen to me, I'll repost for your benefit. Maybe try reading for once:

  I was able to replicate the error consistently.

  Assignment of a negative value to the length property of a vector seems to cause this.

  eg.

  var tLines:Vector.<TextLine> = new Vector.<TextLine>();

  var value:int =  -1;

  tLines.length = value

  [Source: https://forums.adobe.com/message/3113023]

• FarFoo 2,916 Posts

  Sat, Oct 3 2015 8:15 AM

  JLeary1:

  Seeing as so many are hell bent on believing that fixing the memory leaks will solve the GCOD crash (which it won't in most cases), and since no-one is bothering to listen to me, I'll repost for your benefit. Maybe try reading for once:

  
  
  Ironically, I didn't say that fixing memory leaks would solve the problem.
  
  I won't say please re read my post .