another new eol question mark topic

Feature requests and ideas for the new unofficial versions of Elma and general talk related to those.

Moderator: Moporators

User avatar
jonsykkel
Kuski
Posts: 916
Joined: 24 Nov 2009, 20:53
Team: BAP
Contact:

another new eol question mark topic

Post by jonsykkel » 22 Jul 2019, 08:35

is there any intrest for new eol client and server

bigest problem is:
vat to do about buggy fysics engine? fix it or let it be bugy forever
fixing would mean constant high fps and no bugbounces/pops/vsync gaking

advantages of fixing it:
- you can drive a good time and there is no question about whether this bounce or that pop was valid or sach nonsense, its just a good time
- you can make anti xiit that actualy works vhere fysics change a tiny bit (unperceptible but engouh to prevent saveloading or replaying moves) based on seeds received from the server (maxdamantus idea)
- you dont have to wonder if you have been höyling with the wrong fps for years and al that time was wasted
- you dont have to wonder if randomnes in fysics fuked ur ride
- you can focus on playing the game instead of mesing with fps numbers
- you can play without huge input lag and horror framerate al the time
- the fysics would be the same across computers/platforms (they arent now, example: zamppe elma doesnt do same thing as bene elma and my elma at unlimited fps. it also acts difrent at limited fps)

disadvantages:
- would have to start new wr table (it would be like another across->elma transition)
- many ppl might not like it (not true elma bla bla) and im dont vant to split the gomunity



no point in talking about new features etc yet, i think best way to do this is just start with making a solid minimal vorking eol first
also this is not a guarante it will be finished but i think i will try (might end up like smibu eol, its hardik to keep motivation up over time for big projekt like this)
Image
Image
Prestigious member of 16.6x IL101 club

User avatar
Zweq
35mins club
Posts: 3953
Joined: 28 Nov 2002, 15:54
Location: suo mesta

Re: another new eol question mark topic

Post by Zweq » 22 Jul 2019, 10:08

if the alternative is no new patches and the slow death of EOL in 10 years, then I'm all in for fixed physics :pig: :thumbsup:
Image

User avatar
FinMan
36mins club
Posts: 1930
Joined: 13 Feb 2007, 11:14
Team: dat
Location: Jyväskylä, Finland

Re: another new eol question mark topic

Post by FinMan » 22 Jul 2019, 10:22

im agree with zamppe, also would like it a bit more because that would also allow us update bugs osv that we have in the game currently and new features could be implemented without being a complete genious god
... .. Pipe levels ... . ...Höylä levels .. ...
"Those who are heartless once cared too much."

User avatar
bene
Hot kuski
Posts: 881
Joined: 18 Aug 2003, 23:33
Team: dat
Location: Sweden
Contact:

Re: another new eol question mark topic

Post by bene » 22 Jul 2019, 11:00

Me want okeol.
Image
Image
Image
Signatür ruined by SveinR - smaller plz :*

User avatar
milagros
Cheatless
Posts: 4446
Joined: 19 May 2002, 17:05

Re: another new eol question mark topic

Post by milagros » 22 Jul 2019, 20:49

we discussed it 8 years ago and nobody cared
[carebox]

User avatar
jonsykkel
Kuski
Posts: 916
Joined: 24 Nov 2009, 20:53
Team: BAP
Contact:

Re: another new eol question mark topic

Post by jonsykkel » 23 Jul 2019, 13:00

milagros wrote:
22 Jul 2019, 20:49
we discussed it 8 years ago and nobody cared
nobody cared about the question of fixing it or nat, or nobody cared about the problems with the old physics (and dident want it fixed)?



the question is: new eol with new fysics and real antixiit or new eol with old buggy fysics with worse antixiit
both are posible alternatives
need ppl to say vat they think about this
Image
Image
Prestigious member of 16.6x IL101 club

User avatar
milagros
Cheatless
Posts: 4446
Joined: 19 May 2002, 17:05

Re: another new eol question mark topic

Post by milagros » 23 Jul 2019, 13:09

jonsykkel wrote:
23 Jul 2019, 13:00
nobody cared about the question of fixing it or nat, or nobody cared about the problems with the old physics (and dident want it fixed)?
ppl i asked were unhappy about new wrs and some styles becoming impsy (31 etc)
[carebox]

ile
Kuski
Posts: 133
Joined: 11 Feb 2014, 16:56

Re: another new eol question mark topic

Post by ile » 23 Jul 2019, 19:33

jonsykkel wrote:
22 Jul 2019, 08:35
is there any intrest for new eol client and server

bigest problem is:
fixing would mean constant high fps
no ty cant vhel pop. stupid idea sry :(

User avatar
ArZeNiK
39mins club
Posts: 486
Joined: 30 Jul 2016, 09:18
Team: Ferrari

Re: another new eol question mark topic

Post by ArZeNiK » 24 Jul 2019, 08:14

jonsykkel wrote:
22 Jul 2019, 08:35
is there any intrest for new eol client and server

bigest problem is:
vat to do about buggy fysics engine? fix it or let it be bugy forever
fixing would mean constant high fps and no bugbounces/pops/vsync gaking

advantages of fixing it:
- you can drive a good time and there is no question about whether this bounce or that pop was valid or sach nonsense, its just a good time
- you can make anti xiit that actualy works vhere fysics change a tiny bit (unperceptible but engouh to prevent saveloading or replaying moves) based on seeds received from the server (maxdamantus idea)
- you dont have to wonder if you have been höyling with the wrong fps for years and al that time was wasted
- you dont have to wonder if randomnes in fysics fuked ur ride
- you can focus on playing the game instead of mesing with fps numbers
- you can play without huge input lag and horror framerate al the time
- the fysics would be the same across computers/platforms (they arent now, example: zamppe elma doesnt do same thing as bene elma and my elma at unlimited fps. it also acts difrent at limited fps)

disadvantages:
- would have to start new wr table (it would be like another across->elma transition)
- many ppl might not like it (not true elma bla bla) and im dont vant to split the gomunity



no point in talking about new features etc yet, i think best way to do this is just start with making a solid minimal vorking eol first
also this is not a guarante it will be finished but i think i will try (might end up like smibu eol, its hardik to keep motivation up over time for big projekt like this)
looking at the vast number of requests in the "new eol features" or what topic, a new eol client would be surely needed if someone like you, had the time, dedication and effort to put into it.
as for the physics...a controversial topic to say the least (at least from my perspective). don't wanna sound dumb but a high percentage of levels now have WRs that are highly dependent on FPS tuning (take 01, 02, 31 as examples - and then again i'm no 35tt man so i have no idea of how much FPS-changing is required for such times). my suggestion would be as follows: just like on speedrun.com there are games with runs categorised under versions (e.g. mario odyssey) , we could divide the WR tables into 2: a table with times driven in "old" EOL (with pops, possibly buggy enough bounces that wouldn't be poss in "new" EOL, stretches, and so on) , and a table with times driven in "new" EOL, which has the aforementioned limitations
i'm not too familiar with the nature of the community overall but i dont think this would split it up more than it is right now
i do agree that it's in vain to talk beyond such a minimalistic approach of barely a new engine yet.
contrary to popular opinion (at least what i read so far), i think smibuelma turned out really good for what it was destined to be. combine the profoundness of eol2 with the fabulousness of all your projects from the past, and some hint of dedication and spare time = a beautiful project and possibly a successor to EOL
if you ever start doing this, i believe in you

ps: new server is surely also needed imo
I like turtles
Image

User avatar
zebra
Kuski
Posts: 947
Joined: 23 Sep 2003, 15:35
Team: TAP
Location: Finland
Contact:

Re: another new eol question mark topic

Post by zebra » 24 Jul 2019, 10:42

> fixing would mean constant high fps and no bugbounces/pops/vsync gaking

Sounds very good! I guess nobody really likes to play with low fps? Or trying to guess the correct fps?

Pops and bounces are bugs and might scare newcomers away. I nearly quitted elma when I saw warmup wheelpop wr. It's just so illogical... I would probably also remove bounces as well (not only bugbounces) because they were originally bugs as well. But I understand there are some people who like bouncing now when they are accustomed to them.

By the way, were comes the arbitrary 30 fps limit? Why can't we play with 29 fps?
A winner of 4 GAA's (mc2 included), winner of mkup206, and a proud member of team TAP.
Play uni levels: http://koti.mbnet.fi/zebra/uni.html
Homepage: http://koti.mbnet.fi/zebra/elma.html

User avatar
Ruben
Kuski
Posts: 800
Joined: 31 Dec 2012, 18:07
Location: Oslo, Norway

Re: another new eol question mark topic

Post by Ruben » 24 Jul 2019, 11:34

I'll chip in my two cents.

I think getting rid of bounces is a bad idea. They're fun, and altogether a very learnable and reproducable part of playing. Bugbounces I could do without. It's hilarious 1/1000 times when it's advantageous and you get an impossible time in a battle, but the rest of the time they just ruin runs. Maybe just the strength of bounces could do with some... refinement. Imo if we get rid of bounces we might as well get rid of alovolt (and supervolt) while we're at it.

I wouldn't mind getting rid of pops, but in this case it might be worth getting the opinion of people who actually use them.
My fake plants died because I did not pretend to water them.

User avatar
jonsykkel
Kuski
Posts: 916
Joined: 24 Nov 2009, 20:53
Team: BAP
Contact:

Re: another new eol question mark topic

Post by jonsykkel » 24 Jul 2019, 11:41

ArZeNiK wrote:
24 Jul 2019, 08:14
my suggestion would be as follows: just like on speedrun.com there are games with runs categorised under versions (e.g. mario odyssey) , we could divide the WR tables into 2: a table with times driven in "old" EOL (with pops, possibly buggy enough bounces that wouldn't be poss in "new" EOL, stretches, and so on) , and a table with times driven in "new" EOL, which has the aforementioned limitations
yes, old wr table wouldnt be "reset" or anything, there would just be a new table for new game
ArZeNiK wrote:
24 Jul 2019, 08:14
ps: new server is surely also needed imo
plan is to rewrite everything including new server
zebra wrote:
24 Jul 2019, 10:42
Pops and bounces are bugs and might scare newcomers away. I nearly quitted elma when I saw warmup wheelpop wr. It's just so illogical... I would probably also remove bounces as well (not only bugbounces) because they were originally bugs as well. But I understand there are some people who like bouncing now when they are accustomed to them.
i agre, i also been thinking it might be something that prevents new players from wanting to gggget good and go for wrs etc since its not enoguh to just play the game, also have to have al kinds of knowlege about fps tuning and exploiting glitches in the physics to be "competitive". i guess some people like that kind of stuff (dont get me vrong im also think its intresting to see wats posible with original physics), but bugpops and dependency on fps is nat vat makes elma a good game imo. im convinced it would just be a beter game if it was only about presing gas brake left right alo turn. simpel and elegant but impsy to master, like all good games that are played for 1000 years (chess)

dont agree about bounces, i think they are very different since they are posible to do with any framerate, they "make sense" intuitively (maybe), are not insanely dificult to master and prety much everyone agres they make the game more interesting even though they werent exacly intended by balazs himself
zebra wrote:
24 Jul 2019, 10:42
By the way, were comes the arbitrary 30 fps limit? Why can't we play with 29 fps?
there is no 30fps limit, it only exists in eolconf.exe. there is a point where the inner physics loop starts executing more than once if the delta time from last frame gets too large, effectively limiting the max timestep used in the physics calculations to prevent the bike from exploding. but this starts happening at 80fps and below

Code: Select all

main loop
  get delta time
  sample input
  phys loop
    calc physics
  end phys loop
  render
end main loop

80.1 - 1000 fps: 1 iteration of phys loop
39.3 - 80.0 fps: 2 iterations
26.7 - 39.2 fps: 3 iterations
19.9 - 26.6 fps: 4 iterations 
16.1 - 19.8 fps: 5 iterations
...
this pattern continues forever
but the lower framerates are probably not very interesting/useful since the max timestep (maximally fuked physics) is already achieved at around 79fps, so one wouldnt expect huge differences in the way the physics behaves below that.
also since the input is only sampled in the main loop it becomes unplayable due to low resolution and input lag
Image
Image
Prestigious member of 16.6x IL101 club

User avatar
Ruben
Kuski
Posts: 800
Joined: 31 Dec 2012, 18:07
Location: Oslo, Norway

Re: another new eol question mark topic

Post by Ruben » 24 Jul 2019, 11:46

Looking at that code... Is it just me, or is basing the physics on the framerate an ingeniously bad idea?
My fake plants died because I did not pretend to water them.

User avatar
jonsykkel
Kuski
Posts: 916
Joined: 24 Nov 2009, 20:53
Team: BAP
Contact:

Re: another new eol question mark topic

Post by jonsykkel » 24 Jul 2019, 11:56

Ruben wrote:
24 Jul 2019, 11:46
Looking at that code... Is it just me, or is basing the physics on the framerate an ingeniously bad idea?
i gues this was written in 90s and probably not many ppl knowed how to do things back then, and balazs probably didnt expect ppl to pley the game for 20years at this kind of scientific level where it becums a problem
Image
Image
Prestigious member of 16.6x IL101 club

User avatar
kuchitsu
Kuski
Posts: 1263
Joined: 13 Aug 2010, 20:31

Re: another new eol question mark topic

Post by kuchitsu » 24 Jul 2019, 12:09

I would love to see elma without tuning and pops. I think they take the game too far into some esoteric dimension that only a select few can understand and appreciate. I don't even want to make super short battles anymore because someone will win with an unnatural looking poo, and these used to be my favorite.

User avatar
zebra
Kuski
Posts: 947
Joined: 23 Sep 2003, 15:35
Team: TAP
Location: Finland
Contact:

Re: another new eol question mark topic

Post by zebra » 24 Jul 2019, 12:16

Thanks for explanation, jonsykkel. So it seems that the lower limit is actually 80.1 fps and it's just stupid to play with lower fps?

I like the constant fps idea also because it would make automatic levels possible (the bike acts same way every time when you are not pressing anything). Am i right?
A winner of 4 GAA's (mc2 included), winner of mkup206, and a proud member of team TAP.
Play uni levels: http://koti.mbnet.fi/zebra/uni.html
Homepage: http://koti.mbnet.fi/zebra/elma.html

User avatar
jonsykkel
Kuski
Posts: 916
Joined: 24 Nov 2009, 20:53
Team: BAP
Contact:

Re: another new eol question mark topic

Post by jonsykkel » 24 Jul 2019, 12:48

zebra wrote:
24 Jul 2019, 12:16
Thanks for explanation, jonsykkel. So it seems that the lower limit is actually 80.1 fps and it's just stupid to play with lower fps?
nat quite, it does behave slightly differently, below 80fps the first iteration is always at max timestep but the timestep for the second iteration depends on the framerate, close to 80 its low (behaves like high fps) and when you approach 39.3 close to max. so you have some interesting values like ~79 where you have alternating high and low timesteps that cud be good for pops since they require fuked physics but also braking for a very short period. cud be there are also more of these interesting values for other reasons too, hard to predict. but its unlikely to get much more interesting below ~38 at least
http://kopasite.net/up/x5tj897r267t7n0/ ... uences.txt
(fps: timestep 1st iter, 2nd iter, ...)

also theres another reason for using diffrent fps numbers, if you are hoyling for specific times there are "good" fps numbers where certain times are wider or narrower targets (due to there not being exactly a multiple of 100 frames in a second so theres no 1:1 mapping from every frame to every .xx time), makeing them more or less likely to hit
zebra wrote:
24 Jul 2019, 12:16
I like the constant fps idea also because it would make automatic levels possible (the bike acts same way every time when you are not pressing anything). Am i right?
ye, but if i implement the antixiit idea im talking about in the first post, it might make them impossible again (depending on how its done, could still be done so that when not pressing anything it behaves 100% identically every run)
Image
Image
Prestigious member of 16.6x IL101 club

User avatar
Hosp
38mins club
Posts: 1904
Joined: 30 Aug 2009, 20:55
Team: MiE
Location: Uppsala, Sweden.
Contact:

Re: another new eol question mark topic

Post by Hosp » 24 Jul 2019, 12:55

for me different ways to play the game with different FPS is part of the magic, finding what works best for certain maps osv too bad when can't reach max fps tho and stuck on some shitty 100 fps laptop so cant do certain tricks
also if remove bounces = elma dies, is 90% of what makes the game fun
new features osv very needed but if has to be fixed fps for that to happen i am not up for it elma is fine as it is but sad that nothing can be added to it (or removed, like apple battles XD)
Image
tomat

User avatar
Kopaka
39mins club
Posts: 6469
Joined: 23 May 2002, 13:59
Team: LAME
Location: In a northern danish city beating YOUR record.
Contact:

Re: another new eol question mark topic

Post by Kopaka » 24 Jul 2019, 13:23

I wouldn't mind constant high fps, as a player I don't care about going for those minescule improvements you can do with tinkering with fps, but probably most are different from me in that regard. New WR table is a worthy sacrifice for "fixed" physics imo, could even edit a couple ints so there's no vsync holes for example.

My questions:
what are your reasons for not continuing on smibu's work instead of something completely from scratch?
have you made any thoughts regarding closed, semi closed, open source?
any thoughts on the legal issues?

User avatar
jonsykkel
Kuski
Posts: 916
Joined: 24 Nov 2009, 20:53
Team: BAP
Contact:

Re: another new eol question mark topic

Post by jonsykkel » 24 Jul 2019, 14:15

Hosp wrote:
24 Jul 2019, 12:55
for me different ways to play the game with different FPS is part of the magic, finding what works best for certain maps osv too bad when can't reach max fps tho and stuck on some shitty 100 fps laptop so cant do certain tricks
also if remove bounces = elma dies, is 90% of what makes the game fun
new features osv very needed but if has to be fixed fps for that to happen i am not up for it elma is fine as it is but sad that nothing can be added to it (or removed, like apple battles XD)
you are not up for it means you would quit the game forever if it was like that?
im would never remove bounces, only fix bugbounces, pops and fps dependency
maby shud be some poll at sum point where if 75%+ or something want new physics it will be done

Kopaka wrote:
24 Jul 2019, 13:23
My questions:
what are your reasons for not continuing on smibu's work instead of something completely from scratch?
have you made any thoughts regarding closed, semi closed, open source?
any thoughts on the legal issues?
why not continue smibus vork:
biggest reason is that its just impsy to work with other ppls code, it would be biger effort to figure it out than to write my own. i also have a decent renderer from okesl that i would use as a starting point
i have too many disagreements about how it has been done so far, so i would have to change many things anyway. nothing wrong with it but its just different from what i have in mind
c++ is horror language, and i need quick build times

about closed/open source thing, it depends on the physics question. if we keep the old physics, that means inferior antixiit which is basicaly just gambling on noone reverse engineering either the network protocol or the game exe, so in that case some of the source has to be closed. random seed antixiit would allow for the source to be open since you cant really trick it in any way i can think of

legal issues: buying the rights for the game seems impsy so far, one possibility i have been thinking about is to try to enter some kind of agreement with balazs where the game is sold under the elasto mania name at 5 dolars or something (maybe put it on steam also, cud be an ez way to advertise it) and balazs receives 50% of the revenue or something, rest goes to server costs
alternatively if balazs isnt interested, it would just continue like todays eol i guess
Image
Image
Prestigious member of 16.6x IL101 club

User avatar
pawq
38mins club
Posts: 6196
Joined: 24 Aug 2008, 19:56
Team: TR
Location: Southampton, UK

Re: another new eol question mark topic

Post by pawq » 24 Jul 2019, 14:58

jonsykkel wrote:
22 Jul 2019, 08:35
- you can make anti xiit that actualy works vhere fysics change a tiny bit (unperceptible but engouh to prevent saveloading or replaying moves) based on seeds received from the server (maxdamantus idea)
Could you pls explain how this doesn't reintroduce randomness into gameplay but at the same time prevents using marcos to auto-play levs? What variables would be randomised osv


Personally I'm pretty torn about teh fixed physics. The gakked physics is one of the reasons why the game kept evolving so much over the last 20 years...

User avatar
jonsykkel
Kuski
Posts: 916
Joined: 24 Nov 2009, 20:53
Team: BAP
Contact:

Re: another new eol question mark topic

Post by jonsykkel » 24 Jul 2019, 15:34

pawq wrote:
24 Jul 2019, 14:58
Could you pls explain how this doesn't reintroduce randomness into gameplay but at the same time prevents using marcos to auto-play levs? What variables would be randomised osv
maby i didnt word it very clearly, it does ofc reintroduce randomness into the gameplay, but it should be tuned carefully so that its not enough to fuk up your start in a level and that its not noticable in any way even for pros (the randomness currently in eol is very high)

i dont know what would be randomized, anything really that would be just enough to prevent autoplaying. maybe a good way is to change many variables but by very small amounts, would have to experiment and see what works

maybe it turns out its not even possible to do this without adding too much randomness, i dont think thats the case but im dont know havent tested it yet
pawq wrote:
24 Jul 2019, 14:58
Personally I'm pretty torn about teh fixed physics. The gakked physics is one of the reasons why the game kept evolving so much over the last 20 years...
cud be true, at least for last couple of years but i think most "good" evolution still comes from players finding new styles and getting better at pleying

duno if i think evolution in a direction where getting good times becomes more about finding the right fps numbers and whelpop spots is very good

if imagine an alternate universe where elma never had these problems and someone asked if we should introduce some new glitches to make the game more interesting, i think noone would say yes to that
Image
Image
Prestigious member of 16.6x IL101 club

User avatar
pawq
38mins club
Posts: 6196
Joined: 24 Aug 2008, 19:56
Team: TR
Location: Southampton, UK

Re: another new eol question mark topic

Post by pawq » 24 Jul 2019, 18:02

jonsykkel wrote:
24 Jul 2019, 15:34
duno if i think evolution in a direction where getting good times becomes more about finding the right fps numbers and whelpop spots is very good
Yes, I don't think it is, but I still feel attached to all the strange gak :< stupid hooman brain

jonsykkel wrote:
24 Jul 2019, 15:34
if imagine an alternate universe where elma never had these problems and someone asked if we should introduce some new glitches to make the game more interesting, i think noone would say yes to that
Ye, I very much doubt that, but sadly we're attached to this stupid gak now so it's not the same scenario :<



Regarding the antixiit, I think doing some testing on that should maybe be the first step in this, because if for whatever reason it turns out impsy or to reintroduce enough randomness to become annoying, then the whole discussion (regarding fixing fisiks) may be meaningless. I'd be happy to help with some testing at least.

User avatar
jonsykkel
Kuski
Posts: 916
Joined: 24 Nov 2009, 20:53
Team: BAP
Contact:

Re: another new eol question mark topic

Post by jonsykkel » 24 Jul 2019, 19:00

pawq wrote:
24 Jul 2019, 18:02
Regarding the antixiit, I think doing some testing on that should maybe be the first step in this, because if for whatever reason it turns out impsy or to reintroduce enough randomness to become annoying, then the whole discussion (regarding fixing fisiks) may be meaningless. I'd be happy to help with some testing at least.
ye, geting client/server comunication with verification of every run vorking properly would be max priority thing to begin with for sure
Image
Image
Prestigious member of 16.6x IL101 club

User avatar
Hosp
38mins club
Posts: 1904
Joined: 30 Aug 2009, 20:55
Team: MiE
Location: Uppsala, Sweden.
Contact:

Re: another new eol question mark topic

Post by Hosp » 25 Jul 2019, 03:33

jonsykkel wrote:
24 Jul 2019, 14:15
Hosp wrote:
24 Jul 2019, 12:55
for me different ways to play the game with different FPS is part of the magic, finding what works best for certain maps osv too bad when can't reach max fps tho and stuck on some shitty 100 fps laptop so cant do certain tricks
also if remove bounces = elma dies, is 90% of what makes the game fun
new features osv very needed but if has to be fixed fps for that to happen i am not up for it elma is fine as it is but sad that nothing can be added to it (or removed, like apple battles XD)
you are not up for it means you would quit the game forever if it was like that?
im would never remove bounces, only fix bugbounces, pops and fps dependency
maby shud be some poll at sum point where if 75%+ or something want new physics it will be done


Nah I wouldn't quit I don't think. I dunno I think this would be a nice thing for the game in many ways, but also possibly a bad thing for the gameplay since it would change a lot. But what I know for sure is that it would be exciting and the options to add new features to the game, like maybe a recsource in game osv would be sick.
Image
tomat

User avatar
Labs
38mins club
Posts: 1038
Joined: 2 May 2005, 14:20
Team: SPEED
Location: Hungary
Contact:

Re: another new eol question mark topic

Post by Labs » 26 Jul 2019, 13:17

How you make the game doing constant high fps? If one using some pentium 1 or smth, he could still play with his low fps? Bug bounces appear on high fps also, just rarer than in low fps i think. Or i misunderstood and the phisics will not depend at all from fps? I dont like pop thing and animal farm vsync bugs so i would like the "new" game.
Team SPEED

Image

i love everyone :*

User avatar
jonsykkel
Kuski
Posts: 916
Joined: 24 Nov 2009, 20:53
Team: BAP
Contact:

Re: another new eol question mark topic

Post by jonsykkel » 26 Jul 2019, 13:50

Labs wrote:
26 Jul 2019, 13:17
How you make the game doing constant high fps? If one using some pentium 1 or smth, he could still play with his low fps? Bug bounces appear on high fps also, just rarer than in low fps i think. Or i misunderstood and the phisics will not depend at all from fps? I dont like pop thing and animal farm vsync bugs so i would like the "new" game.
yes physics would be decoupled from graphics so even though fraps say 45fps on ur pentium 1 physics would still run at 1000fps (or watever)
bug bounces would be fixed so they cant occur even at high fps
Image
Image
Prestigious member of 16.6x IL101 club

User avatar
pawq
38mins club
Posts: 6196
Joined: 24 Aug 2008, 19:56
Team: TR
Location: Southampton, UK

Re: another new eol question mark topic

Post by pawq » 26 Jul 2019, 14:56

jonsykkel wrote:
26 Jul 2019, 13:50
bug bounces would be fixed so they cant occur even at high fps
How do you plan to do that without altering the physics significantly? dunno how the physics code works exactly, but I understand that bugbounces are the result of the denominator (wheel distance from the centre of the bike) being too close to 0. Can you simply ignore this calculation when the wheel distance crosses a certain threshold, or? If yes, how determine the threshold?

User avatar
jonsykkel
Kuski
Posts: 916
Joined: 24 Nov 2009, 20:53
Team: BAP
Contact:

Re: another new eol question mark topic

Post by jonsykkel » 26 Jul 2019, 18:42

pawq wrote:
26 Jul 2019, 14:56
jonsykkel wrote:
26 Jul 2019, 13:50
bug bounces would be fixed so they cant occur even at high fps
How do you plan to do that without altering the physics significantly? dunno how the physics code works exactly, but I understand that bugbounces are the result of the denominator (wheel distance from the centre of the bike) being too close to 0. Can you simply ignore this calculation when the wheel distance crosses a certain threshold, or? If yes, how determine the threshold?
yes, the reason they happen is this place in the code where you have 1.0 divided by the length of bike->wheel vector, so you can just pretend that length is slightly longer than it is if its too close to 0, that fixes it without making normal bounces any different
the threshold can be determined by experimentation, checking against borderline bug .dats etc. tested it quickly now with threshold = 0.01 and with fix in place it runs all the normal .dats (including normal bounces) i tested perfectly but bug .dats dont work anymore
Image
Image
Prestigious member of 16.6x IL101 club

User avatar
pawq
38mins club
Posts: 6196
Joined: 24 Aug 2008, 19:56
Team: TR
Location: Southampton, UK

Re: another new eol question mark topic

Post by pawq » 26 Jul 2019, 21:04

Nice. Follow-up question: is it possible to "accurately" determine the wheel distance from .recs?

User avatar
jonsykkel
Kuski
Posts: 916
Joined: 24 Nov 2009, 20:53
Team: BAP
Contact:

Re: another new eol question mark topic

Post by jonsykkel » 26 Jul 2019, 21:17

pawq wrote:
26 Jul 2019, 21:04
Nice. Follow-up question: is it possible to "accurately" determine the wheel distance from .recs?
depends how accurate it needs to be, the wheel position is stored with a resolution of 0.001 elmameters which is not too bad but the reks are 30fps so there might be important frames missing if you are looking for minimum distance for example
Image
Image
Prestigious member of 16.6x IL101 club

User avatar
pawq
38mins club
Posts: 6196
Joined: 24 Aug 2008, 19:56
Team: TR
Location: Southampton, UK

Re: another new eol question mark topic

Post by pawq » 27 Jul 2019, 13:50

Accurate enough to be able to determine which current-elma internal (or not) recs are buggy and which ain't. This question has been our nemesis for years, but if it's possible to implement such a filter in a potential fixed-physics elma, then it should also be possible to start using it as a determinant of which recs should be accepted in the current elma (should the new elma not happen, for whatever reason).

Also regarding the missing frames at 30fps: since physics is calculated at the same framerate, then it should only be the wheel distance at those framerates that matters, no? Anything in between being basically lost (not even calculated)? or?

User avatar
jonsykkel
Kuski
Posts: 916
Joined: 24 Nov 2009, 20:53
Team: BAP
Contact:

Re: another new eol question mark topic

Post by jonsykkel » 27 Jul 2019, 14:22

pawq wrote:
27 Jul 2019, 13:50
Accurate enough to be able to determine which current-elma internal (or not) recs are buggy and which ain't. This question has been our nemesis for years, but if it's possible to implement such a filter in a potential fixed-physics elma, then it should also be possible to start using it as a determinant of which recs should be accepted in the current elma (should the new elma not happen, for whatever reason).
recs arent good enoguh for that, mostly because 30fps thing and also its not enoguh to know just the bike->wheel distance, there also needs to be a torque applied to the wheel at the same time (from either brake or throttle) for a bug to occur, and you cant tell when brake is presed in recs
pawq wrote:
27 Jul 2019, 13:50
Also regarding the missing frames at 30fps: since physics is calculated at the same framerate, then it should only be the wheel distance at those framerates that matters, no? Anything in between being basically lost (not even calculated)? or?
the physics are never calculated at 30fps, always higher than ~78 so there is guaranteed to be missing frames in all recs
Image
Image
Prestigious member of 16.6x IL101 club

User avatar
pawq
38mins club
Posts: 6196
Joined: 24 Aug 2008, 19:56
Team: TR
Location: Southampton, UK

Re: another new eol question mark topic

Post by pawq » 27 Jul 2019, 18:06

oke, ty for the explanations :>

User avatar
Lousku
Kuski
Posts: 2894
Joined: 5 Feb 2010, 00:25
Team: BAP
Location: expensive land of dads

Re: another new eol question mark topic

Post by Lousku » 27 Jul 2019, 23:48

- you can make anti xiit that actualy works vhere fysics change a tiny bit (unperceptible but engouh to prevent saveloading or replaying moves) based on seeds received from the server (maxdamantus idea)
If it's really pasibel to make it imperceptible and "irrelevant" to players, then this idea sounds amazing. To a nab it sounds like there might be megaspesifik benetricks that rely on seed luck. If so, it still sounds like the best solution. If not, this is amazing.
then again i don't know anything
maybe easier not to think abouut alöl things thought than not things thought ... or something..=?

User avatar
jonsykkel
Kuski
Posts: 916
Joined: 24 Nov 2009, 20:53
Team: BAP
Contact:

Re: another new eol question mark topic

Post by jonsykkel » 28 Jul 2019, 15:55

need help from my linus & macos boys!
please download this virus code and compile it with clang/gcc, run it and send me the output in any way. just "make" shud gompile and run

http://kopasite.net/up/2s0gc85tqgxld2l/fptest.zip
Image
Image
Prestigious member of 16.6x IL101 club

Spef
34mins club
Posts: 108
Joined: 25 Jun 2011, 23:33

Re: another new eol question mark topic

Post by Spef » 29 Jul 2019, 14:44

Me like - wheelpop, vsync, bounce, brutals, finding right FPS values for levels, randomness except for bounce strength.
I can continue to enjoy or suffer those things in current EOL even if okeol makes changes. Then an issue could be current EOL dying and server not beneing kept up in favor of okeol. How make WR in EOL if can't be online?
[10:47:49] (Hosp) tol thinks that all there is in life is a gf, status, money blabl
[10:48:19] (ANATOLIY) not all hosp =)
[10:48:29] (ANATOLIY) but elma must be at second screen

User avatar
Labs
38mins club
Posts: 1038
Joined: 2 May 2005, 14:20
Team: SPEED
Location: Hungary
Contact:

Re: another new eol question mark topic

Post by Labs » 29 Jul 2019, 16:39

I think 2 thing can run on the same server.
Team SPEED

Image

i love everyone :*

User avatar
pawq
38mins club
Posts: 6196
Joined: 24 Aug 2008, 19:56
Team: TR
Location: Southampton, UK

Re: another new eol question mark topic

Post by pawq » 29 Jul 2019, 17:38

I think it would be ultimately damaging for this community if two versions of the game were "allowed" to run simultaneously. I think that if a decision is made to commit to fixing the physics, the resultant game would essentially have to replace the current EOL. Of course nobody would prevent anyone from still running the old EOL offline for fun (or WRs...), but hosting both at the same time would fragment the community, which would all but ensure it's survival over the long-term.

User avatar
Labs
38mins club
Posts: 1038
Joined: 2 May 2005, 14:20
Team: SPEED
Location: Hungary
Contact:

Re: another new eol question mark topic

Post by Labs » 29 Jul 2019, 18:48

What if "old eol" still could run on server for internal höyl of wrs for "old moposite" tables, with the battle mode (and cups) disabled. That way it wouldnt really separate community. Dunno if its damaging, but maybe worths a try for a test phase.
Team SPEED

Image

i love everyone :*

User avatar
ArZeNiK
39mins club
Posts: 486
Joined: 30 Jul 2016, 09:18
Team: Ferrari

Re: another new eol question mark topic

Post by ArZeNiK » 29 Jul 2019, 20:13

Labs wrote:
29 Jul 2019, 18:48
What if "old eol" still could run on server for internal höyl of wrs for "old moposite" tables, with the battle mode (and cups) disabled. That way it wouldnt really separate community. Dunno if its damaging, but maybe worths a try for a test phase.
sounds good idea for me since vsync and bug things are kinda controversial in balles anyway
I like turtles
Image

User avatar
pawq
38mins club
Posts: 6196
Joined: 24 Aug 2008, 19:56
Team: TR
Location: Southampton, UK

Re: another new eol question mark topic

Post by pawq » 29 Jul 2019, 20:27

Labs wrote:
29 Jul 2019, 18:48
That way it wouldnt really separate community.
It would separate those too attached to start hoyling ints with fixed physics from the others...

User avatar
jonsykkel
Kuski
Posts: 916
Joined: 24 Nov 2009, 20:53
Team: BAP
Contact:

Re: another new eol question mark topic

Post by jonsykkel » 29 Jul 2019, 22:12

i agree with pavq it would have to be one or the other. the whole point of this is to try to revive the game a litle, but if it means that sum players are gona be left behind höyling the old game, that would kinda defeat the purpose and i will just use the original fysics for new eol

about haveing both versions of the physics in the same game: not really an option, it would mean a lot more vork dealing with 2 difrent engines, 2 completely difrent antixiit systems, etc. already big enough projeckt as it is
Image
Image
Prestigious member of 16.6x IL101 club

User avatar
skint0r
Kuski
Posts: 744
Joined: 16 Jun 2002, 07:36
Location: Oslo, Norway

Re: another new eol question mark topic

Post by skint0r » 30 Jul 2019, 02:00

imo insane mind blow that people unwilling to move forwards, no wander community die. also mind blow that considering not make new thing just to appease few nostalgia seeking people who won't let go of old times, and then just make the game slowly die instead of possibly reviving it and fix all the stupid shit that stops game from being even better.

me: ??????? real wategak
Prestigious member of 14.6x Tutor14 club

User avatar
jonsykkel
Kuski
Posts: 916
Joined: 24 Nov 2009, 20:53
Team: BAP
Contact:

Re: another new eol question mark topic

Post by jonsykkel » 30 Jul 2019, 03:24

skint0r wrote:
30 Jul 2019, 02:00
imo insane mind blow that people unwilling to move forwards, no wander community die. also mind blow that considering not make new thing just to appease few nostalgia seeking people who won't let go of old times, and then just make the game slowly die instead of possibly reviving it and fix all the stupid shit that stops game from being even better.

me: ??????? real wategak
good coment i agre with it 500 percent

trust me im not cumsidering going with old physics to appease nostalgia seeking ppl. they can go play across in dosbox

its apsolutely no doubt crystal clear to me that the right choice is fixed physics
my only konsern is - if do that might be sum smal chacne somoene wud stop playing because of it. not talking about nostalgia seeking nabs but the kind of ppl who owns half the vr table. that would be bad adn the opposite of wat shud hapen

im think most likely outcome though would be that ppl sadik at first but then move to the new eol and later they wil see how much beter it is
Image
Image
Prestigious member of 16.6x IL101 club

User avatar
8-ball
39mins club
Posts: 4469
Joined: 9 May 2003, 13:30
Team: BAP
Location: Valmiera, Latvia

Re: another new eol question mark topic

Post by 8-ball » 30 Jul 2019, 11:25

skint0r wrote:
30 Jul 2019, 02:00
imo insane mind blow that people unwilling to move forwards, no wander community die. also mind blow that considering not make new thing just to appease few nostalgia seeking people who won't let go of old times, and then just make the game slowly die instead of possibly reviving it and fix all the stupid shit that stops game from being even better.

me: ??????? real wategak
this

also
jonsykkel wrote:
22 Jul 2019, 08:35
- you can drive a good time and there is no question about whether this bounce or that pop was valid or sach nonsense, its just a good time
- you can make anti xiit that actualy works vhere fysics change a tiny bit (unperceptible but engouh to prevent saveloading or replaying moves) based on seeds received from the server (maxdamantus idea)
- you dont have to wonder if you have been höyling with the wrong fps for years and al that time was wasted
- you dont have to wonder if randomnes in fysics fuked ur ride
- you can focus on playing the game instead of mesing with fps numbers
- you can play without huge input lag and horror framerate al the time
- the fysics would be the same across computers/platforms (they arent now, example: zamppe elma doesnt do same thing as bene elma and my elma at unlimited fps. it also acts difrent at limited fps)
the opposite of all these being true and the resulting fps meta is what made me lose interest in the game and looks like i wasn't alone

i feel for mans who actually enjoyed that (as someone already worded it) esoteric gameplay but it's ultimately killing the game and its possible appeal to new players who could ensure its perpetuity

i think old skoal mans could be given old eol to host themselves to continue competition for legacy physics wrs as long as they maintain interest themselves which honestly i doubt would last long

this has been the most pogchamp topic in years thanks jon
39:37,91

culinko
39mins club
Posts: 1532
Joined: 29 Dec 2002, 19:17
Location: Bratislava, Slovakia
Contact:

Re: another new eol question mark topic

Post by culinko » 30 Jul 2019, 12:39

I feel the same as 8-ball. I lost interest in eol after every new WR came with a wheelpop in it. Fixed physics (no wheelpops, bugbounces, vsync/fps stuff) sounds amazing to me! Another thing that can be super good is automatic WR/record updates every week/biweekly/monthly or even instantly if ppl want it (could be non-WRs instantly and WRs non-instantly, etc.). Currently, the totals on Moposite (also individual times) are not even updated except WRs which are inserted manually into records...

I also feel for people who like Spef who likes all the things that would be eliminated in new eol. Usually I am also the one who dislikes any new changes and likes "things the way they are" (still hate the new lauta layout) but I am 100% convinced that eol needs this change in order to stay relevant.

Then there's also the question: Do we keep all of the people's times/records/totals/WRs from the old eol or do we start with a clean slate? I feel like in order for the best transition we need to start from the scratch, which means it would suck for people who hoyled shit for decades (like me lol) and also everyone would have a chance to get a WR in new eol (this is also pretty funny). Maybe we keep old WRs but only as placeholders? E.g. in order to make WR in new eol you would need to beat the time of the WR in old eol, but the WR in old eol wouldn't necessary be a WR in new eol. However this also means that some of the WRs from the old eol would probably never be beaten (e.g. animal farm).

As for both new and old eol running simultaneously, even for just internals, I feel like that won't really work and as was already said would just fracture the community. We're in an impossible situation here, aren't we? :D
Image

User avatar
Labs
38mins club
Posts: 1038
Joined: 2 May 2005, 14:20
Team: SPEED
Location: Hungary
Contact:

Re: another new eol question mark topic

Post by Labs » 30 Jul 2019, 13:10

About the wr table, it doesnt make sense to wait with updates there, just make it instant, with automated rec upload. Old px system would be awful.
Team SPEED

Image

i love everyone :*

User avatar
nick-o-matic
Kuski
Posts: 829
Joined: 27 Jan 2007, 12:53
Location: Lappeenranta

Re: another new eol question mark topic

Post by nick-o-matic » 30 Jul 2019, 13:59

First of all, huge shoutout for jon for wannaing to make this project happen!! :beer: :beer: :beer: Do we finally have The Chosen One...? =o

Secondly, imo at this point this is by a mile the #1 question:
Kopaka wrote:
24 Jul 2019, 13:23
any thoughts on the legal issues?
Finding an agreement with Balazs and the ability to make OKEOL official would boom the importance of project with 999%. It would also make the project a lot more rewarding to do when it is coming to a lot bigger audience. It could hopefully even become a business and help the developer(s) updating and upgrading it.
jonsykkel wrote:
24 Jul 2019, 14:15
legal issues: buying the rights for the game seems impsy so far, one possibility i have been thinking about is to try to enter some kind of agreement with balazs where the game is sold under the elasto mania name at 5 dolars or something (maybe put it on steam also, cud be an ez way to advertise it) and balazs receives 50% of the revenue or something, rest goes to server costs
alternatively if balazs isnt interested, it would just continue like todays eol i guess
Balazs' iOS elma was apparently quite big fail, but apparently he does not want to entirely sell elma. But he does not have to. Something like this would be good for us (not having to find impsy moneies to pay for him at the day one) and good for him (he still has elma rights and he does not have to do anything to get some money out of it). We should take immediate action to contact Balazs about this, especially since jon himself apparently agrees.

What it comes to the physics, battles are the bread and butter of EOL and they would be that also in the new version. It is an awesome idea if everyone could have the same fun-to-play high-fps physics in battles - some people playing with low fps have significant disadvantage at least in some levels (and often vice versa). But the internals are a more complex question. The history of the internal hoyling is a huge richness of elma and imo noteworthy in the history of videogame speedrunning. It should not be neglected. I have always dreamt that in the new version of elma you could go to an internal and from some menu see the WR history and eye all the shared WR recs and also see the world top 1000 and see all the shared recs osv osv. That should be amazingly fun and interesting also for new players who are learning to play. But that would not exactly fit well with fixed physics for internals... Thus I give as strong vote as ever poss for this:
Labs wrote:
29 Jul 2019, 16:39
I think 2 thing can run on the same server.
So if you enter an internal, you will have the classical physics, and if you enter an external, you would have the fixed high-fps physics. This system has some obvious idiot system flaws (already pointed out), but a flawless solution is simply not poss here and imo this is the most optimal solution to this situation. Two different WR tables would be the most horror ever imo, sorry. Also some mongoness of the classical physics could be avoided by making FPS change poss to be made in-game and to be level-specific. You could set a default FPS for int 31 without affecting other levs. And you could do that from the level itself, without going to some settings menu in the main menu of the game.
jonsykkel wrote:
24 Jul 2019, 11:41
i agre, i also been thinking it might be something that prevents new players from wanting to gggget good and go for wrs etc since its not enoguh to just play the game, also have to have al kinds of knowlege about fps tuning and exploiting glitches in the physics to be "competitive".
Höyling for WRs is a really miniscule part of elmaing already nowadays, there's some 5 players doing that. The importance of internals is imo in:

1) The awe-effect of yeing all the WR recs.
2) Following the progress of WRs (spying WR hoylers, anticipating the new WR table and waiting for the recs to be published in some videos etc osv).
3) Schooling new players in mega basic playing (the original purpose of internals, FPS does not matter at all).
4) Schooling not-so-new-anymore players in getting a little bit good by competing against some target times and strying to enter some top-5000 or top-1000 or top-100 in the all-time lists (FPS osv matter mostly in 27 and 31, but being able to lower FPS in-level would help).
5) Eyeing all the historical WR recs and the WR style development and the top-20 or top-100 of each lev (this also goes hand-in-hand with point #4).
....
29) Actually ggggettin' good at internals and going for WRs yourself. Even with fixed internal physics new players would not start competing with Spef, Zweq and others. There is simply so much other stuff to play and where to achieve something at. Even nowadays most people who have gggggot good do not play ints. We totally should not worry about new pro players never going for int WRs.

Moreover, you could have the classical physics also in some manually chosen old hoyled externals (there needs to be some extensive official in-game levelpack system) and you could also start battles with classical physics. Vsync stuff and pops and other low-fps things are also an integral part of external playing too imo, and this would completely eliminate them. My hope is there will be enough players for having multiple different kind of battles going on at the same time, so this kind of oldfag stuff should not confuse and disturb new players too much. Also in classical low-fps battles you could have the option for the battle starter to decide the fps, which would then be the same for everyone in the battle. This would add some equality even though there's always the fps-built-in randomness. These things are quite small details anyway. In the big picture the int physics should remain classical.

Bug bounce detection could be made automatic by introducing a new .rek format where every frame is saved and thus it would be poss to check if the bounce frame had center of bike and wheel too close. Or the game could be coded to throw internal error when a bugbounce occurs. =D

Antixiit things are super preliminary at this point. Tiny randomness would not really help in fighting against autoplay imo. You can always drive a SL ride to the start of some level and then autoplay perfect starts every try because randomness should not be big enough to affect starts of levels. That is already an enormous advantage especially in some levels. Imo the only solution is to store all the buttonpresses with some gamestates and to spy the player playing the level to analyze by both programs and eye if the hoyling process looks natural (some admins should be able to spy also hidden players).

Sad if Smibu's Elma 2 can't be used as a base. But perhaps some blocks of code, or at least some ideas in Elma 2 code, could be useful...? Perhaps after all his work Smibu could give some comments and advice and ideas for this project...?

User avatar
8-ball
39mins club
Posts: 4469
Joined: 9 May 2003, 13:30
Team: BAP
Location: Valmiera, Latvia

Re: another new eol question mark topic

Post by 8-ball » 30 Jul 2019, 14:11

i'm not convinced, i think there should be no exceptions for legacy physics. turning over the page in history doesn't invalidate the impact of legacy records. everything that was worth seeing has already been immortalized in compilation videos, not to mention that you could still browse history and download recs from moposite and watch them in any legacy elma if you feel like it, and there would still be recsource. i think it's a niche concern that would unnecessarily delay work on more important things in game. fixed fps wr table would be 34 mins in no time anyway and that in and of itself would be exciting to witness.
39:37,91

Post Reply