No problem, I didn't think he was ignoring me, I just saw that he has been active elsewhere on the forum so thought I'd throw a 'bump' in there in case he missed the last one.
We're still working out a way to have a "Known Bug" list that is maintained with QA approval.
Cool, as a few people have now said that would be reassuring, a way to see if our efforts - reporting bugs - were worthwhile / being recorded.
I can say with clarity that @BlackCloakGS and our other GS devs do listen and do work in the community interest.
I don't doubt that for a moment, as I've said elsewhere (and this is, of course, just my own opinion) if an issue does exist it seems to be one of communication, the bug reporting system might be the single most efficient and effective communication channel known to mankind for all I know, but take feedback out of any system and you may as well switch it off, people will stop using it.
Another thing that might help would be if there was a way to flag information on this forum as old or no longer relevant. For example, from what I understand destroying actors use to be a problem so people invented a work around. If a new user happens upon that work around they might think it's still relevant. So it would be cool if it could be flagged as old.
I've worked around a few bugs myself, but I'm VERY close to posting my game. I applied for my Google Play account today and am working on screenshots and vids now. I could have never done this without Game Salad and help from everyone on the forum!!
Another thing that might help would be if there was a way to flag information on this forum as old or no longer relevant. For example, from what I understand destroying actors use to be a problem so people invented a work around. If a new user happens upon that work around they might think it's still relevant. So it would be cool if it could be flagged as old.
Unfortunately this problem is applicable to every forum covering every subject, information will nearly always - in time - become dated / erroneous, it happens everywhere from the BBC News website to GameSalad, NewScientist to National Geographic, I think there should be an onus on the consumers of information (me and you and everyone else) to develop the basic tools to navigate the information they find online, otherwise the whole of the internet would be in a constant never-ending state of repair / revision.
If you interpolate from 0 to 1 you should expect (I'm using only 2 decimal places here):
0.01 0.02 0.03 0.04 0.05 0.06
. . [snip] . .
@Socks I do think that its a float math inconsistency with that one specific issue. As you are only using 2 decimal places, I do not think that "time" is. since its over time, its always changing a value, trying to accurately represent those changing values over time. There is also the time taken to render each frame, which under stress will be visible. This is a whole different subject... (delta time).
Yeah, I agree it would be great if all the info was constantly updated to reflect the latest release / and or latest best practices, but it would be a nightmare to organise, imagine having to trawl through hundreds of thousands of posts checking for erroneous info !!
I do think that its a float math inconsistency with that one specific issue.
What I don't understand then is why interpolate is able to correctly calculate 0.6 to 0.7 and 0.2 to 0.3 but not 0.9 to 1.0 - as they both require the same process / accuracy ?
As you are only using 2 decimal places, I do not think that "time" is.
Your sentence is a little cryptic ? But I think you might be confusing things here, the reference to me only using 2 decimal places was for my post only - it has nothing to do with interpolate - it's just to save me typing out a list of hundreds of numbers. I'm making no reference to GameSalad or the interpolate issue when I say I'm only using 2 decimal places.
There is also the time taken to render each frame, which under stress will be visible.
Again, why then does this issue only (and always) show up where it does - if interpolate placed undue pressure on the processor then why is it able to run for hours, days (years in theory) with zero issues and only trip up in the last 10% of the last whole digit ?
It's a pity no one from GameSalad is able (or willing) to chime in, it would be interesting to hear from the experts.
I am sorry if it feels like we ignore bug reports.
However let me assure you we take bug reports seriously and any that come cross my desk I try to make sure get addressed.
Here are a few bugs that have been in GameSalad for years (literally) - and reported more than once.
Q: Is it possible to say whether GameSalad are aware of these particular bugs ?
Are you aware that the brightness and opacity sliders in the colour selection wheel often crash creator if you try and adjust values using them ?
That the interpolation algorithm has the last part of it's destination value clipped so on things like fading in and out of images and ramping up/down of audio all end up with annoying glitches at the end ?
Importing a sequence of animation frames into creator screws up their order and they have to be manually reordered before being placed in the animate behaviour.
Deleting assets like audio and images often returns in strange results - images are not deleted, the wrong images are deleted, the image panel fails to update so you are not entirely sure whether your action was successful.
The frame rates in the animate behaviour are all 'screwed', they not only don't match the frame rates displayed on the behaviour's slider, but they bear little relation to each other (21fps runs at the same speed as 27fps / 28fps and 29fps both run at 20.1fps / 20fps runs at 16.5 fps . . . etc etc) ?
. . . and so on, there are quite a few bugs like this that have been present for a couple of years or so, like I say they've been reported, and more than once, and exhaustively, and through the correct channels. I understand that you can't address every bug or glitch people might hit, but it would be at least good to know you where aware of them and that we weren't just shooting these reports into a black hole, as I said above I think the real issue is one of communication.
SSS's idea of a known bug list would go someway to giving people the sense that they are not being ignored, even knowing a bug was unlikely to be addressed in the near future would be much better than being left entirely in the dark. False hope can be a cruel mistress
One of my first projects - that relies on reasonably accurate frame rates (or even internally consistent, if not accurate, frame rates) - still lives in hope even though it has now got a fairly thick coat of dust on it.
tatiangMember, Sous Chef, PRO, Senior Sous-ChefPosts: 11,949
@Socks thanks for taking the time to reiterate those bugs. I, too, am hopeful we'll see a known bug list. I think it would cut down on some of the frustration and the frequency of repeated reports on the forums.
@Socks thanks for taking the time to reiterate those bugs. I, too, am hopeful we'll see a known bug list. I think it would cut down on some of the frustration and the frequency of repeated reports on the forums.
Yep, it would save a lot of people the trouble of documenting, testing and reporting bugs that GameSalad are already aware of, it would save GameSalad from having to sift through identical bug reports from different people and would give GameSalad users a sense that GameSalad take their bug reports seriously.
I do think that its a float math inconsistency with that one specific issue.
What I don't understand then is why interpolate is able to correctly calculate 0.6 to 0.7 and 0.2 to 0.3 but not 0.9 to 1.0 - as they both require the same process / accuracy ?
As you are only using 2 decimal places, I do not think that "time" is.
Your sentence is a little cryptic ? But I think you might be confusing things here, the reference to me only using 2 decimal places was for my post only - it has nothing to do with interpolate - it's just to save me typing out a list of hundreds of numbers. I'm making no reference to GameSalad or the interpolate issue when I say I'm only using 2 decimal places.
There is also the time taken to render each frame, which under stress will be visible.
Again, why then does this issue only (and always) show up where it does - if interpolate placed undue pressure on the processor then why is it able to run for hours, days (years in theory) with zero issues and only trip up in the last 10% of the last whole digit ?
It's a pity no one from GameSalad is able (or willing) to chime in, it would be interesting to hear from the experts.
Hey chaps @Socks & @3itg , I've got some more insight on this, not sure if it'll help, but hey... You must be right about it being a floating point issue, because I am having similar trouble with the last 10% clipping, but I am not using interpolate OR timers! I am simply trying to check if A is a divisor of B and I am doing this by...
if ceil(a/b) = (a/b) then a is divisor of b, because the result is integer, no decimals, so it must divide properly, so it must be a divisor...fine...but not fine!....
if the result of the test is too close ie. 5.92543 then GS still treats this as equal to 6! So, I am getting false results in my divisor tests because of this.
So, as you said, the last 10% of ANY floating point, it would seem, is a problem.
Wasn't there a discussion about this in the 10.2 or 10.3 thread? I've been searching, but can't find it again. Can anyone link to it if they know where it is please?
Comments
Bump ?
Are you aware of the bugs I listed in my earlier post in this thread ?
No problem, I didn't think he was ignoring me, I just saw that he has been active elsewhere on the forum so thought I'd throw a 'bump' in there in case he missed the last one. Cool, as a few people have now said that would be reassuring, a way to see if our efforts - reporting bugs - were worthwhile / being recorded. I don't doubt that for a moment, as I've said elsewhere (and this is, of course, just my own opinion) if an issue does exist it seems to be one of communication, the bug reporting system might be the single most efficient and effective communication channel known to mankind for all I know, but take feedback out of any system and you may as well switch it off, people will stop using it.
Your 'known bug' list is an excellent idea.
I've worked around a few bugs myself, but I'm VERY close to posting my game. I applied for my Google Play account today and am working on screenshots and vids now. I could have never done this without Game Salad and help from everyone on the forum!!
THANK YOU!
I guess I agree with you.
I do think that its a float math inconsistency with that one specific issue.
As you are only using 2 decimal places, I do not think that "time" is.
since its over time, its always changing a value, trying to accurately represent those changing values over time.
There is also the time taken to render each frame, which under stress will be visible.
This is a whole different subject... (delta time).
It's a pity no one from GameSalad is able (or willing) to chime in, it would be interesting to hear from the experts.
Bump.
Here are a few bugs that have been in GameSalad for years (literally) - and reported more than once.
Q: Is it possible to say whether GameSalad are aware of these particular bugs ?
Are you aware that the brightness and opacity sliders in the colour selection wheel often crash creator if you try and adjust values using them ?
That the interpolation algorithm has the last part of it's destination value clipped so on things like fading in and out of images and ramping up/down of audio all end up with annoying glitches at the end ?
Importing a sequence of animation frames into creator screws up their order and they have to be manually reordered before being placed in the animate behaviour.
Deleting assets like audio and images often returns in strange results - images are not deleted, the wrong images are deleted, the image panel fails to update so you are not entirely sure whether your action was successful.
The frame rates in the animate behaviour are all 'screwed', they not only don't match the frame rates displayed on the behaviour's slider, but they bear little relation to each other (21fps runs at the same speed as 27fps / 28fps and 29fps both run at 20.1fps / 20fps runs at 16.5 fps . . . etc etc) ?
. . . and so on, there are quite a few bugs like this that have been present for a couple of years or so, like I say they've been reported, and more than once, and exhaustively, and through the correct channels. I understand that you can't address every bug or glitch people might hit, but it would be at least good to know you where aware of them and that we weren't just shooting these reports into a black hole, as I said above I think the real issue is one of communication.
SSS's idea of a known bug list would go someway to giving people the sense that they are not being ignored, even knowing a bug was unlikely to be addressed in the near future would be much better than being left entirely in the dark. False hope can be a cruel mistress
One of my first projects - that relies on reasonably accurate frame rates (or even internally consistent, if not accurate, frame rates) - still lives in hope even though it has now got a fairly thick coat of dust on it.
New to GameSalad? (FAQs) | Tutorials | Templates | Greenleaf Games | Educator & Certified GameSalad User
- Thomas
You must be right about it being a floating point issue, because I am having similar trouble with the last 10% clipping, but I am not using interpolate OR timers! I am simply trying to check if A is a divisor of B and I am doing this by...
if ceil(a/b) = (a/b) then a is divisor of b, because the result is integer, no decimals, so it must divide properly, so it must be a divisor...fine...but not fine!....
if the result of the test is too close ie. 5.92543 then GS still treats this as equal to 6! So, I am getting false results in my divisor tests because of this.
So, as you said, the last 10% of ANY floating point, it would seem, is a problem.
Wasn't there a discussion about this in the 10.2 or 10.3 thread? I've been searching, but can't find it again. Can anyone link to it if they know where it is please?
Thanks
Wayne