Bug in long lev
Moderator: Moporators
Bug in long lev
Quite incredibly frustrating bug for me: in FFS99a my Elma freezes weirdly exactly at the same time: 35:49:63. In other long levs this has never happened (haven't tested just to make sure though).
http://kopasite.net/up/d80rxzy6w3n450a/ffs1.png
http://kopasite.net/up/d6g9ciqxt5sng2x/ffs2.png
Pretty much exactly same place bot times too (these runs were ~a week apart)
During this "freeze" chat works normally and all the functions work, the timer just stops and I can't move.
http://kopasite.net/up/d80rxzy6w3n450a/ffs1.png
http://kopasite.net/up/d6g9ciqxt5sng2x/ffs2.png
Pretty much exactly same place bot times too (these runs were ~a week apart)
During this "freeze" chat works normally and all the functions work, the timer just stops and I can't move.
Team MiE - MiE Cup 1
Prestigious member of 14.6x Tutor14 club
Prestigious member of 14.6x Tutor14 club
Re: Bug in long lev
This once happened to Kokso17 at the very end of Tabi004. We were all spying him and already congratulating him on the finish, then suddenly he froze just some seconds from the flower. Markku, Danielj and maybe some other people should remember. Perhaps you could get him to do some testing on this.
Re: Bug in long lev
Have had such freeze in int52 when spying zew. Bike just stops. But when froze i was tabbed out and using wine. Not mid during playing.
Signatür ruined by SveinR - smaller plz :*
Re: Bug in long lev
I used to get this in warm up very often some years ago. After i minimized elma i would come back and bike was freezed, some times in mid air. Got it also while playing, in one of my longest pob levs, after over 30 minutes playing :f It was at the very top of the level, i thought it had to do something with some level limits. But then, i also could get this on warm up, it was a pretty recurring thing for me. Now i dont get that anymore.
Re: Bug in long lev
It happened to me more than one time and its very frustrating . I think i also might have freezed time around 35: ... when i had 999 fps. I found out that when i put 800 fps it was more (over 40 mins i think) till freezing and with lower fps i havent managed to get this bug . But I dont understant almost anything about computers so i cant really offer any cause or solution
Re: Bug in long lev
This is interesting because 35 * 999 = 34965, which is pretty close to 2^15 = 32768. And 32768 is a known limit in many data types, like the classic C++ integer. So it is possible that Elma freezes after it draws 32768 frames due to an overflow. The fact that lower FPS allows you to play more seems to support this theory (it takes more time to draw 32768 frames if there are less frames drawn per second). 35 * 999 gives more than that, but perhaps you actually have a bit less than 999 FPS.
So I guess you can just limit your FPS a little bit for long levs. 32768 / 60 = 546, so I assume something like 500 FPS should allow you to play for an hour at least.
So I guess you can just limit your FPS a little bit for long levs. 32768 / 60 = 546, so I assume something like 500 FPS should allow you to play for an hour at least.
Re: Bug in long lev
Or wait, not 32768 frames because that's minutes... I dunno what I'm talking about.
Re: Bug in long lev
You could still be partially right though. I always play at roundabout 60 and I've never had this issue, so it could be something that only happens on super high fps.
<veezay> antti also gonna get stabbed later this month
<nick-o-matic> niec
My fake plants died because I did not pretend to water them.
<nick-o-matic> niec
My fake plants died because I did not pretend to water them.
Re: Bug in long lev
it's probably eol related bug, not an elma bug.
Re: Bug in long lev
It is caused by line bs.bs$qv+1A52F which modifies elma's variable at eol.exe+9900C.
e: It seems that the function is flawed since it always sets eol.exe+9900C to the same value which tells elma to stay on the same frame.
It is not elma's fault, it is EOL. Placing NOPs at that line will make elma run normally again while staying in the level.
Not sure what that function of EOL does but to fix it, either lower your fps or play shorter levels
The person who made EOL could also fix this but everyone would have to download a new executable.
e: It seems that the function is flawed since it always sets eol.exe+9900C to the same value which tells elma to stay on the same frame.
It is not elma's fault, it is EOL. Placing NOPs at that line will make elma run normally again while staying in the level.
Not sure what that function of EOL does but to fix it, either lower your fps or play shorter levels
The person who made EOL could also fix this but everyone would have to download a new executable.
Canada
Re: Bug in long lev
Sounds like setting a limit on the fps is gonna become a default for ultra-long FFs :/ Possible to figure out what a safe fps limit is for, say, a 60-min long lev?
Team TR
Multi WR in Labyrinth with GRob
Best Internal Total Times, Pipe stats & Pipe archive
World kuski map, World Cup stats
Re: Bug in long lev
these might not be the same for everyone:Pawq wrote:Sounds like setting a limit on the fps is gonna become a default for ultra-long FFs :/ Possible to figure out what a safe fps limit is for, say, a 60-min long lev?
999 fps - 35:47:88
900 fps - 39:44:50
800 fps - 44:38:23
700 fps - 51:06:85
600 fps - 59:37:32
anything lower will be over 1 hour, maybe someone can draw a graph with the data above.
Note that when it freezes, if you wait another big set amount of time (~30 minutes) it will be back to normal although all your inputs will be placed during the freeze.
It seems like a simple overflow bug somewhere, to lazy to find where it is. This would cause the freeze/unfreeze after a huge time.
Canada
Re: Bug in long lev
Sounds like 300fps should be safe then for any length of level. Ty for the feedback!
Team TR
Multi WR in Labyrinth with GRob
Best Internal Total Times, Pipe stats & Pipe archive
World kuski map, World Cup stats
Re: Bug in long lev
Yeah, thanks for solving this, interesting & good to know :>
Team MiE - MiE Cup 1
Prestigious member of 14.6x Tutor14 club
Prestigious member of 14.6x Tutor14 club
Re: Bug in long lev
some different factor must be influencing this also
i have finished over 80min run, and had many 1h+ runs with around 700fps and never had this kind of bug. markku finished a 92min run and i guess he had very high fps also. so its quite odd. and surely many other ppl played for over 35mins on max fps and never heard about that case before
very exceptional stuff
i have finished over 80min run, and had many 1h+ runs with around 700fps and never had this kind of bug. markku finished a 92min run and i guess he had very high fps also. so its quite odd. and surely many other ppl played for over 35mins on max fps and never heard about that case before
very exceptional stuff
It's nice to be important, but it's more important to be nice.
(Devann) u are the master of your own universe. remember that
(Devann) u are the master of your own universe. remember that
Re: Bug in long lev
i belive in cold's theory and the reason for unconsistency is most likely due to random fps drops which in essence means you can drive for longer before freeze
Re: Bug in long lev
I thought the same when I saw jbl's postofta wrote:i belive in cold's theory and the reason for unconsistency is most likely due to random fps drops which in essence means you can drive for longer before freeze
Team TR
Multi WR in Labyrinth with GRob
Best Internal Total Times, Pipe stats & Pipe archive
World kuski map, World Cup stats
Re: Bug in long lev
i guess some variable overflown for fps limiter
i thought nobody uses fps limiter except in internals:)
i thought nobody uses fps limiter except in internals:)
[carebox]