Mopolauta

Elasto Mania board
It is currently 22 Sep 2017, 19:45

All times are UTC [ DST ]




Post new topic Reply to topic  [ 125 posts ]  Go to page 1, 2, 3  Next
Author Message
PostPosted: 28 Oct 2014, 17:25 
Offline
Cheatless
User avatar

Joined: 19 May 2002, 17:05
Posts: 4246
As anyone else, I always wondered what kind of times are theoretically possible in Elma. So some years ago I decided to find out.
The plan was as follows :
a) first I code standard cheating tool with kinds of saveload-like tools, as user friendly as possible
b) I drive WRs with these tools
c) I code an optimization tool which would take b) as initial solution and try various methods

After a few days of trying it became quite clear, that it is pretty hard to beat most WRs even with cheat, if you don't have skills or touch for a particular level. So I decided to included some pros in the project. Original plan was to include Zweq and John, but John gave up after a few days and hasn't driven any decent times.

So me and Zweq started to beat WRs one-by-one, meanwhile I improved the cheating tool bit-by-bit, quickly reaching some 34low totaltime. The first problems appeared when I realized how the bounces work. Using simple tools we "optimize" the keypresses to get any strong bounce any direction. At that point we had to define, what's acceptable and what's not. Over time you somehow get blind and tend to accept stronger and stronger bounces. Often we jsut argued with Zweq if something is fine if it's 5% stronger or now. Eventually we reached that 32.3x total time with and released it on 1st of April:). We played a bit more and improved some of the times, but after the release we kinda lost motivation. Also that bug-bouncing contributed significantly to the end of the 32tt project. At some point we also started to drive bug WRs, where any bug bounce was acceptable and got pretty amazing times (6.xx in 44, 7.xx in 47, 8.xx in 01, 9.xx in 02). But we didn't drive such times for every level. The bug-bounce problem made optimization part c) impossible. Any decent method would end up optimizing the strength of bounces. Too sad:/

Eventually, maybe 70% of the recs were driven by Zweq, and 30% by me. I did for example - 02, 06, 21, 38, 44, 53 and several parts in all other levs (2nd half of 07, 2nd half of 01, last zigzag part and so on), often I just did some 1sec parts in the middle of some long levs. Zweq drove the rest:). He is pro:) I was just hoyling. Often at work:). Large majority of styles were found by Zweq (with some exceptions, such as bowling). At some point we thought, that romy4 might have hacked our site, because most of his styles we had some recs for:).

Originally, I wanted this project to be forgotten. But Zweq convinced me that maybe it is a good idea to release the cheating tool, so that all the top players could try to get together the best times possible. Discuss.

_________________
[carebox]


Top
 Profile  
 
PostPosted: 28 Oct 2014, 17:37 
Online
39mins club

Joined: 29 Dec 2002, 19:17
Posts: 1517
Location: Bratislava, Slovakia
I was always curious about the best possible WRs. Afaik after Smibu released his s/l, the internal activity rose and people tried impsy styles with it. I think tools like this (also hourglass) gave us pretty impressive recs by many pros. I would also like to see your 32tt recs:) I'm all for releasing the tool, because I think it would help the internal scene (at least for some time).

_________________
Image


Top
 Profile  
 
PostPosted: 28 Oct 2014, 17:41 
Offline
Kuski
User avatar

Joined: 24 May 2004, 16:22
Posts: 451
Location: #ranks & #rscoring
I don't see any huge harm in releasing the cheat tools, but many people would be against it. So I'm somewhere in the middle I think. Whether you release cheat tools or not, I think you should at least release recs.

_________________
( igge) i will not quit elma before i finish this level
My YouTube channel


Top
 Profile  
 
PostPosted: 28 Oct 2014, 18:06 
Offline
38mins club
User avatar

Joined: 2 May 2005, 14:20
Posts: 907
Location: Hungary
Ye recs would be better than progs. If cheat tool released then it must be detectable at least, even if some pro-coder tries to change code.

_________________
Team SPEED

Image

Q: Which number Bjenn plays on roulette?
A: 29.


Top
 Profile  
 
PostPosted: 28 Oct 2014, 18:22 
Offline
35mins club
User avatar

Joined: 28 Nov 2002, 15:54
Posts: 3701
Location: suo mesta
I will obviously have ultra much to write about this and I will keep posting here as I keep remembering all the million things to say. First and foremost I'm happy I can finally 100% openly talk about stuff, for example it was rather tragicomical to answer questions like "have you seen all 54 mila recs?", yeah maby I had seen them. Also I think it went somehow wrong with uploading recs to youtube xd, although I'm pretty sure I said I'm ok with it.

Labs wrote:
Ye recs would be better than progs. If cheat tool released then it must be detectable at least, even if some pro-coder tries to change code.


Dunno what you're smoking, maybe you haven't noticed there are 2 undetectable cheat tools already xd. I think it would be very exciting to start getting the theoretical total time down in a co-effort (you can share the rides as datafiles and other guy can continue from any part)

_________________
Image


Top
 Profile  
 
PostPosted: 28 Oct 2014, 18:40 
Offline
Hot kuski
User avatar

Joined: 18 Aug 2003, 23:33
Posts: 484
Location: Sweden
Well, I would say releasing the tools could be a good idea, watching some tas video of elma would be amazing, even if any bugs accepted video, even if no bugs video osv.
I don't think releasing the replays is a good idea, maybe as a video, but there are still some styles that are hidden and interesting to some mans. Like int02 osv.

_________________
Image
Image
Image
Signatür ruined by SveinR - smaller plz :*


Top
 Profile  
 
PostPosted: 28 Oct 2014, 19:31 
Offline
38mins club
User avatar

Joined: 2 May 2005, 14:20
Posts: 907
Location: Hungary
I just not wanne to see cheated wr's on the table. Smibu's tool is oke since you can't use it online, dunno about hourglass, is weird prog. Maybe you can modify smibu's tool too and change it to approve online play tho, i'm not a coder. At least if new tools released there should be some protection of going online with them.

_________________
Team SPEED

Image

Q: Which number Bjenn plays on roulette?
A: 29.


Top
 Profile  
 
PostPosted: 28 Oct 2014, 20:27 
Offline
36mins club
User avatar

Joined: 4 Jul 2004, 12:10
Posts: 5167
thanks for sharing the story, would've been quite sad if we just forgot about it.

Agreeing with most things that have been said, especially
Labs wrote:
At least if new tools released there should be some protection of going online with them.

This is quite obvious. But if this isn't an issue, then I'm all for releasing the tools and plsplspls make an TAS Elma Done Quick video.

Also if cheat tools would be released it would be quite nice to have a cheat-WR-table somewhere (something a little more "official" then some thread on lauta would be nice) since I also agree that Smibu releasing S/L was a very important step to keep the internal scene alive. This could be quite a motivation-booster for the entire internal-community, especially when it's possible that several kuski drive one rec. would be mega awesome to see a int35 cheat-WR shared by zweq, anp, kazan and spef or something like that.

Of course at the moment the WR-table not being updated might be the bigger issue here at the moment, but I'd love to see new styles and new awesome times on the normal WR-table too. and maybe we'll get 33 non-cheat-TT someday.

_________________
TT: 36:59:53 || Avg TT: 38:09:65


Top
 Profile  
 
PostPosted: 28 Oct 2014, 20:44 
Offline
38mins club
User avatar

Joined: 6 Nov 2003, 06:04
Posts: 918
Location: Russia
Knew it >:D

Imo release tools, would be a nice motivation boost for some mans, and possibly good brainstorming coming out of it. And not release recs, at least in the next couple of years :p

There're some weird consequences coming out of this though. So, zweq had had access to fps limiter before anyone else and before it was released, and now it also appears that he'd also had access to sl (and even more tools?) before others as well? Being teh best player and all, it doesn't really matter, but still, doesn't it make it a little unfair to have such secret tools at your disposal when you're an active wr-making player?


Top
 Profile  
 
PostPosted: 28 Oct 2014, 21:21 
Offline
35mins club
User avatar

Joined: 28 Nov 2002, 15:54
Posts: 3701
Location: suo mesta
ye life is unfair :( i had to fight against darmoed for 2 years and rusmans had SL for 20 years anyway. But FPS limiter incident was pretty bad, I didn't fully understand how much better 30 fps is compared to 60 fps in flat track. Go 6AH zweq.

_________________
Image


Top
 Profile  
 
PostPosted: 28 Oct 2014, 21:45 
Offline
Kuski
User avatar

Joined: 27 Jul 2003, 06:44
Posts: 611
Location: Ikrény, Hungary
release recz or youtube video, whatever... but we must see those great times!!

also need to replace px... we need events like these to keep the scene more and more alive!!

_________________
Elasto Mania - 34:37.34


Top
 Profile  
 
PostPosted: 28 Oct 2014, 22:25 
Offline
Kuski
User avatar

Joined: 15 Apr 2007, 01:05
Posts: 3299
Zweq wrote:
Dunno what you're smoking, maybe you haven't noticed there are 2 undetectable cheat tools already xd.
ye, berh, death and some other(s) that can but don't. they're not totally ruining everything even if they can, but if they do, they're pretty much noticeable and fixable. I think community can use it in a good way, rare are those who can actually hack it anyway.

I'm up to release it, would be more interesting to me to use on externals, will be fun to try out :)

_________________
Website || TT:41:45:64 || Team Image
ImageImage


Top
 Profile  
 
PostPosted: 29 Oct 2014, 01:44 
Offline
Kuski
User avatar

Joined: 28 May 2007, 08:47
Posts: 915
People have been begging milagros for years to release the replays, it's starting to sound like a broken record. Not sure if it matters anymore to keep asking him or not, I'd assume he gets the point, seems like a smart enough guy. I will ask him directly.

Milagros are you considering releasing these replays or should we stop dicking around asking you for them?

_________________
God Bless America


Top
 Profile  
 
PostPosted: 29 Oct 2014, 07:58 
Offline
Hot kuski
User avatar

Joined: 18 Aug 2003, 23:33
Posts: 484
Location: Sweden
Why should the replays be released:
* Because cool good times höhö osv I am non pr0.

Why shouln't the replays be released:
* Many of the times are shit and completely uninteresting.
* The few interesting ones are: Hidden style that still confuse some people, Near wr times where wr rec is unpublic, osv. Because these are likely to lead to wr being beat.
* Sharing now would be saying fuck you to many wrs and the internal scene today.
* osv.

For example:
Jarkko has bene keeping int05 replay secret for years, lets decide as a community to release a 0.02 better rec with xiit so people can use in game merging and stuff and suddenly its easy to find out cool stuff and wring is easier because fuck you jarkko.
int13,14,18,44,45,53 osv are similar near wr, maybe more orka kolla. int10 i heard some 10.4x time from mila earlier also maybe zweq drove with new style who knows.

A better idea is to release the tools and let people drive 0.62 better int05 or let someone else drive 0.02 better and share up to them.
The point of this whole thing is however to get pro mans to höyl some 1:07.56 time in int40 not to share some bad 1:17.90 replay. osv. In the long run we will get much better replays to watch.
Like a year or two from share of tools we will watch TEDQ because tas tt is under 28 osv. It also includes TEDQ (any bug) with tt under 13.

It would however be fun to watch some silly fuck bounce in int28 or lal straight to roof brutal in int24 replays now maybe, but for nothing more than mindless entertainment of my feeble brain.

_________________
Image
Image
Image
Signatür ruined by SveinR - smaller plz :*


Top
 Profile  
 
PostPosted: 29 Oct 2014, 08:41 
Offline
Kuski
User avatar

Joined: 27 Jul 2003, 06:44
Posts: 611
Location: Ikrény, Hungary
many tools are already released... i don't think that significant styles and improvements will be done if only tools get published..

seeing OMFGWTF IMPOSSIBLE HOW??? recs would motivate mans much more i believe

_________________
Elasto Mania - 34:37.34


Top
 Profile  
 
PostPosted: 29 Oct 2014, 11:10 
Offline
Kuski
User avatar

Joined: 10 May 2014, 19:23
Posts: 171
bene wrote:
Why shouln't the replays be released:
* Many of the times are shit and completely uninteresting.
* The few interesting ones are: Hidden style that still confuse some people, Near wr times where wr rec is unpublic, osv. Because these are likely to lead to wr being beat.
* Sharing now would be saying fuck you to many wrs and the internal scene today.
* osv.
Maybe mila should release only 1 or 2 recz just for demo

_________________
The best animal hardcore porn ever!


Top
 Profile  
 
PostPosted: 29 Oct 2014, 12:20 
Offline
Cheatless
User avatar

Joined: 19 May 2002, 17:05
Posts: 4246
demo rec


Attachments:
28pro.rec [13.37 KiB]
Downloaded 239 times

_________________
[carebox]
Top
 Profile  
 
PostPosted: 29 Oct 2014, 12:28 
Offline
Hot kuski
User avatar

Joined: 18 Aug 2003, 23:33
Posts: 484
Location: Sweden
13.37 KiB!

There are many recs already released.
int01,int07,int47,int48,int51 are all on youtube.
28pro,30pro,31pro recs are somewhere and should be easy to find.

Am I missing some?

_________________
Image
Image
Image
Signatür ruined by SveinR - smaller plz :*


Top
 Profile  
 
PostPosted: 29 Oct 2014, 13:58 
Offline
38mins club
User avatar

Joined: 6 Nov 2003, 06:04
Posts: 918
Location: Russia
imo not missing. That seriously looks fun to play around with, both for normal stylefinding/detailnitpicking, and for exploring the insane TAS-only routes. Inb4 mila decides to release tools and 1 year later everyone will become bugbounceblind, and we'll get legit tt under 30min.


Top
 Profile  
 
PostPosted: 29 Oct 2014, 19:13 
Offline
Kuski
User avatar

Joined: 28 May 2007, 08:47
Posts: 915
If the tool allowed Zweq to have an ability to find styles that he wouldn't have otherwise realized existed then the tool and all the recs should be released so nobody has any unfair advantage.

Lengthy sentence but that's the truth. Its the same shit that happened with saveload, some had it and some didn't and then it got released to everyone because of ethics.

P.S. not hating on you zweq just being real xoxo

_________________
God Bless America


Top
 Profile  
 
PostPosted: 29 Oct 2014, 19:28 
Offline
Cheatless
User avatar

Joined: 19 May 2002, 17:05
Posts: 4246
styles were found because zamppe is pro, not because he had some tool
edit: btw does teh 28pro look acceptable?:)

_________________
[carebox]


Top
 Profile  
 
PostPosted: 29 Oct 2014, 20:23 
Offline
38mins club
User avatar

Joined: 23 Sep 2009, 10:56
Posts: 1420
Location: Some pub in Prague
Ofc not looking acceptable. I've never ever in my years of playing seen such bounce with this speed from this angle.

Imo that's the only thing that rly sucks about it - that there's no real line between buggish and "normal" bounce. Which is funny, coz "normal" bounce is bug as well.

I've always thought, that max speed you can get after bnc cannot be greater than initial speed. In other words - that bnc just changes direction of riding. But apparently that is not true, so dunno

_________________
http://elmaonline.net/players/Bludek
Image


Top
 Profile  
 
PostPosted: 29 Oct 2014, 21:36 
Offline
Cheatless
User avatar

Joined: 19 May 2002, 17:05
Posts: 4246
here you have maybe 70% of the speed you had before the bounce

_________________
[carebox]


Top
 Profile  
 
PostPosted: 29 Oct 2014, 21:59 
Offline
36mins club
User avatar

Joined: 4 Jul 2004, 12:10
Posts: 5167
i still don't understand the whole bug-bounce thing. are bounces like this reproducable, in the sense that you get the excact same bounce when you land on the excact same spot, in the excact same angle with the excact same keypresses? i just don't understand how a bug bounce is produced, what happens in the code. and why the hell does this bounce look so smooth and unbuggy although it's obviously unnormal?

_________________
TT: 36:59:53 || Avg TT: 38:09:65


Top
 Profile  
 
PostPosted: 29 Oct 2014, 22:18 
Offline
Kuski
User avatar

Joined: 14 Apr 2008, 17:58
Posts: 1458
Location: jkl
Lukazz wrote:
i still don't understand the whole bug-bounce thing. are bounces like this reproducable, in the sense that you get the excact same bounce when you land on the excact same spot, in the excact same angle with the excact same keypresses? i just don't understand how a bug bounce is produced, what happens in the code. and why the hell does this bounce look so smooth and unbuggy although it's obviously unnormal?

Yeh these are the important questions. Is that int 28 bounce like int 30 bounce; reproducable but just really fucking hard to get exactly right?

_________________
Team MiE - MiE Cup 1
Prestigious member of 14.6x Tutor14 club


Top
 Profile  
 
PostPosted: 29 Oct 2014, 22:32 
Offline
35mins club
User avatar

Joined: 28 Nov 2002, 15:54
Posts: 3701
Location: suo mesta
gimp wrote:
If the tool allowed Zweq to have an ability to find styles that he wouldn't have otherwise realized existed then the tool and all the recs should be released so nobody has any unfair advantage.

Lengthy sentence but that's the truth. Its the same shit that happened with saveload, some had it and some didn't and then it got released to everyone because of ethics.

P.S. not hating on you zweq just being real xoxo


Yes of course it should be public, I've been suggesting it to mila for a long time. And yes there is a morale issue - would I find some styles if I wasn't doing recs for mila, I don't know. When you spend time playing, you find styles. I don't know how much it matters where you spend that time, in train lev or saveload, but it matters of course. Often new style comes from an initial idea. Initial idea can be many things, from entirely new route to brake tapping in start. And then the idea just has to be tested. Most of the time the testing part is very easy even in train lev, but not always. Tunnel terror end is a good example of difficult testing. Of course I found my tunnel terror end style while doing rec(50.74) using the tool, anyone with a brain can probably guess that. It would have been harder to find in a train lev, because building a good train lev for that spot is tricky. After finding style I was STOKED of course and went to höyl nonstop in eol until WR. Then I felt bad because of morale issues and asked john if I should send the WR in and he had no idea what I was talking about. Duno anything.


Bludek wrote:
I've always thought, that max speed you can get after bnc cannot be greater than initial speed. In other words - that bnc just changes direction of riding. But apparently that is not true, so dunno


Haven't you seen darms pipe bug bounce? Or maybe I misunderstood.

Lukazz wrote:
i still don't understand the whole bug-bounce thing. are bounces like this reproducable, in the sense that you get the excact same bounce when you land on the excact same spot, in the excact same angle with the excact same keypresses? i just don't understand how a bug bounce is produced, what happens in the code. and why the hell does this bounce look so smooth and unbuggy although it's obviously unnormal?


28pro bounce should be pretty easy to reproduce. Well, when you see your wheel popping in random (not random) direction, you have got a bugbounce, gz. When the (apparently random) wheel pop is orthogonal to the surface of the polygon you get a smooth looking bounce like 28pro, and to do it orthogonal just takes some trial & error. Sometimes it takes 1hour, sometimes 2mins, depending on the case (here it took 2 mins). For some reason in some cases bug bounces tend to be likely to be orthogonal to the surface, for example new pipe style. I have no idea why that is, someone smarter will have to find it out.

And how to get a bug bounce is not a secret, for example ropelli found it out and revealed the mechanic in eol. Mila had told me about it, but first time I saw it in public was by ropelli, and it has also been explained in other threads.

_________________
Image


Top
 Profile  
 
PostPosted: 29 Oct 2014, 22:58 
Offline
37mins club
User avatar

Joined: 28 Feb 2003, 13:35
Posts: 1906
Location: B.C. Canada
I am just shocked that Mila picked zweq over myself as his pro player. Probably be sub 30tt by now.

I agree with the majority here though, replays and tool should be released. Obviously only if the tool is detectable. I haven't been playing in a while but this is something that would peak my interest in the scene again.

Side note, Mila, you sound like a master coder. Make me a bot for Clash of Clans. Easy Canadian $ coming your way.

_________________
Religious Man
|| TT: 37:47:70 || EX New Wave WR + Animal Farm WR || 24 Canadian Records || TEC Bronze Medal || HHIT in the 37min club || http://elastomaster.tripod.com/
||
Image


Top
 Profile  
 
PostPosted: 29 Oct 2014, 23:13 
Offline
Kuski
User avatar

Joined: 23 May 2002, 13:59
Posts: 6296
Location: In a northern danish city beating YOUR record.
bene said some wise things. I don't find s/l recs interesting for viewing pleasure personally, only for style if I'm playing the level. Reason to not release tool could be that people will just find and drive some sick time in the tool and not bother to drive it for real afterwards. Seem to remember some pro internalist saying he felt like that sometimes. Of course cat is already out of the bag with hourglass and smibu's s/l.

Lukazz wrote:
i still don't understand the whole bug-bounce thing. are bounces like this reproducable


If there is no RNG in the code, which to my knowledge there is not, then everything is reproducable, in theory. Then it depends how does it listen to keypresses. Is there a certain interval between each listen and action so that a human might be able to reproduce a run, if there is a saved run might be the same when loaded. If there is not a saved run might be different when loaded due to where the computer currently are in things like cpu cycles which could cause small random delays. Of course this is just assumptions, I don't know anything.

Bludek wrote:
I've always thought, that max speed you can get after bnc cannot be greater than initial speed. In other words - that bnc just changes direction of riding. But apparently that is not true, so dunno


Has been a common way to define it yeah, but not really official in any way I guess. And impossible to read the speed from replay properly as far as I know.

_________________
Kopasite ^ 39 level packs and 1193 levels for your höyling pleasure
LaMe world of Kopa ^ Team LaMe
Elma Online ^ Moposite Records ^ Upload


Top
 Profile  
 
PostPosted: 29 Oct 2014, 23:39 
Offline
36mins club
User avatar

Joined: 4 Jul 2004, 12:10
Posts: 5167
Zweq wrote:
And how to get a bug bounce is not a secret, for example ropelli found it out and revealed the mechanic in eol. Mila had told me about it, but first time I saw it in public was by ropelli, and it has also been explained in other threads.

could you link to any of the threads or explain it?

_________________
TT: 36:59:53 || Avg TT: 38:09:65


Top
 Profile  
 
PostPosted: 29 Oct 2014, 23:55 
Offline
Kuski
User avatar

Joined: 15 Apr 2007, 01:05
Posts: 3299
Kopaka wrote:
And impossible to read the speed from replay properly as far as I know.
why not or what do you mean with properly? simple using RM (replay manager by smibu). bike has coordinates every frame

_________________
Website || TT:41:45:64 || Team Image
ImageImage


Top
 Profile  
 
PostPosted: 30 Oct 2014, 00:32 
Offline
Cheatless
User avatar

Joined: 19 May 2002, 17:05
Posts: 4246
yeah, bounces are reproducible
this 28 style can be done even more often then pipe brutal i guess
it is the same bug
originally zweq did 28 in 2 mins, i did this one in 5:)
without cheat it would take maybe an hour or two

_________________
[carebox]


Top
 Profile  
 
PostPosted: 30 Oct 2014, 01:45 
Offline
Kuski
User avatar

Joined: 24 May 2004, 16:22
Posts: 451
Location: #ranks & #rscoring
I have to say bene's comment has swayed me, releasing recs can be very unfair to some WR holders. Many/most of the recs are probably impossible styles anyway? IMO leave it up to mila/zweq judgement which recs should be released.

As for the morality issue, I don't see anything wrong with zweq using tool to find styles. Everyone is free to make their own programs that can help find styles etc. Just because only a few are able to doesn't mean it is immoral for them to do so. I'm sure many chess players for example have hired programmers to make software for them to help analyze things, and that does not give them an unfair advantage. An advantage, sure, but for me not unfair.

_________________
( igge) i will not quit elma before i finish this level
My YouTube channel


Last edited by danielj on 30 Oct 2014, 02:51, edited 1 time in total.

Top
 Profile  
 
PostPosted: 30 Oct 2014, 01:52 
Offline
Kuski
User avatar

Joined: 14 Apr 2008, 17:58
Posts: 1458
Location: jkl
IMO can release the buggiest most insane recs that could never be replicated in a WR-legit way, just for fun. If there are styles that might be useful in making actual WRs, don't include them.

_________________
Team MiE - MiE Cup 1
Prestigious member of 14.6x Tutor14 club


Top
 Profile  
 
PostPosted: 30 Oct 2014, 05:08 
Online
39mins club

Joined: 29 Dec 2002, 19:17
Posts: 1517
Location: Bratislava, Slovakia
Mila told me once that there are like 5 recs that could lead to possible new WRs. Most are just insanely good optimized times or fak like 51.

_________________
Image


Top
 Profile  
 
PostPosted: 30 Oct 2014, 08:40 
Offline
Kuski
User avatar

Joined: 5 Dec 2008, 16:19
Posts: 1013
Location: a position or point in physical space
Release whatever you want. Just make sure the tools doesn't work online.

_________________
Lousku wrote:
could you mayke shorter sig please :( mega annoying and also against rules :()


Top
 Profile  
 
PostPosted: 30 Oct 2014, 08:52 
Offline
Hot kuski
User avatar

Joined: 18 Aug 2003, 23:33
Posts: 484
Location: Sweden
culinko wrote:
Mila told me once that there are like 5 recs that could lead to possible new WRs. Most are just insanely good optimized times or fak like 51.


I said already, there are many that are bad and using old moves and with no effort behind them.

For example look at int40:
Kazan did 1:17.90, he did not use mila sl strength bounces at apples, do this and suddenly faster time then add some more stuff and some other crazy unpublic styles and you can improve several seconds.
Factor in that mila sl is better in general compared to whoregals with whatever crazy tools we can find or create and kazan can improve some more.
Then add 3 more mans (Madness,Zweq,Anpdad) in a co op effort and suddenly some more seconds improved or whatever.
Even fakin bene knows about 2 faster styles that kazan didn't use. I know there are faster styles that 32tt project didn't use without even seeing that rec. Because newsflash, styles are invented by mans and not by tools and kazan is not a complete idiot.
Suddenly 1:07.56 replay. osv.

The project does not contain insanely good optimized times, it's normal runs with good tools and old moves, at least in the original list of times that mila posted.
Sure there are some more höyled times like 1,2,44 but for exampel int40 is something zweq did in 5 minutes while taking a shit at 4am in the morning.
We will get better times and see better replays if mans are interested in tasing with proper tools this is a fact.

After looking at the list I would only be interested in seeing the ones that I can use for making wrs 1,2,3,13,14,18,34,44,45,53 to name a few from memory osv.
Having for example a 36.19 rec in int34 when wr was 36.37 is a huge advantage. I think you misunderstood mila when he told you there were only 5 interesting recs.

Many of the hidden styles are quite obvious when you see the time and something that even anpdad can find and test. Some styles might be harder to find like int51 dunno orka brain everything now.
Finding the styles is also part of the fun, maybe someone will dislike being handed a solution.

I don't think zweq wants to share a replay that is as bad as 32tt project int40 even and now you know why.
About the whole subject about zweq having sl before others I can say I don't care at all. Probably cause I am mostly raised in the era of widespread saveload dunno.

Is it possible to share some anybag replay like int02 9.xx?

_________________
Image
Image
Image
Signatür ruined by SveinR - smaller plz :*


Top
 Profile  
 
PostPosted: 30 Oct 2014, 09:41 
Offline
39mins club
User avatar

Joined: 7 Apr 2007, 12:15
Posts: 6291
Location: Stockholm, Sweden
I kinda see Mila and Zweq teaming up to create sick styles just like any other team in elma, only just so happens they are a really good and expansive team. No one has been complaining about anpdad and kazan and whoever else is in this super-style finding, russian wr-making team when they have been SLing styles and making wrs. If I'm not mistaken they've done it since before Smibu's easily accessible SL too, and instead done it through hourglass which I myself have no idea how to get a hold of and get to work properly etc...

Sure, maybe the cheats used should be released, but I don't think it will change anything drastically. Tools similar to Mila's (I assume) have always been available; and what's really needed is a brain to use them. In the case of the 32ttproject, the tools required for making these sick styles and times were not-so-much the xiits created by mila, but rather the skills and creativity provided by Zweq. Just look at shit like table 265 and all the fucking amazing styles this guy has come up with. It's not because of the tools that he's able to think of them; tools just help him express them.

You're all saying elma is some sort of material-sport. It's like saying O'sullivan is better cause he has a better cue or analysis team, or that Lance Armstrong has better bike or steroids (point here being SL might be considered a xiit but everyone can use it so who gives a damn), or Picasso has a better paint brush and canvas... In the end tools are just a way for the person to show their true selves. Everyone can use them, some better than others. It all just boils down to raw ability.

tl;dr: zweq is a tool

_________________
John: lol hittade ett popcorn i naveln
(19:52:06) (@Madnezz) The Golden Apple Award goes to.....
(19:52:36) (@Madnezz) ib9814.lev by igge!!!
Zweq wrote:
99.9999% of nabs haven't even opened the book yet and most of those that have are still on the first pages


Top
 Profile  
 
PostPosted: 30 Oct 2014, 10:10 
Offline
Hot kuski
User avatar

Joined: 18 Aug 2003, 23:33
Posts: 484
Location: Sweden
Igge wrote:
Tools similar to Mila's (I assume) have always been available; and what's really needed is a brain to use them.

I don't know exactly how mila sl works but if you read his first post he is talking about simple tools for "optimizing" bounces for any direction and speed they want. This sounds like a bounce helper and does not exist in hourglass.

maligaros wrote:
The first problems appeared when I realized how the bounces work. Using simple tools we "optimize" the keypresses to get any strong bounce any direction.

Maybe the tools are just frame advancing and sick understanding of how bounces work which the community doesn't know 100%. There is a few random comments about it here and there but not enough for replicating them in hourglass in a few minutes of playing I think.
Sharing the tools would mean sharing the knowledge and suddenly more people can reasonably make these recs. Because no sane person is going to blindly ram into a wall for 3 hours in hourglass and hope they get a bounce.

There might be more tools specialized for elma that we don't know of yet that wouldn't be present in hourglass.

_________________
Image
Image
Image
Signatür ruined by SveinR - smaller plz :*


Top
 Profile  
 
PostPosted: 30 Oct 2014, 10:39 
Offline
35mins club
User avatar

Joined: 28 Nov 2002, 15:54
Posts: 3701
Location: suo mesta
rough cunt is 25-30 recs are worth watching for anyone, even for bene, none of those seem to be in mila's list. Mila's list is mostly controversial rides with small bugs here and there, sometimes just one bug (new wave). But bene is partially right, many levs haven't been touched barely at all.

bene wrote:
Igge wrote:
Tools similar to Mila's (I assume) have always been available; and what's really needed is a brain to use them.

I don't know exactly how mila sl works but if you read his first post he is talking about simple tools for "optimizing" bounces for any direction and speed they want. This sounds like a bounce helper and does not exist in hourglass.



Just to clear further misunderstandings. All times are 100% self driven, and they are all theoretically possible within the simulation. (for example there is no editing the state of bike). Input is given to the game, and the game simulates an outcome, it's just we could change fps where-ever-the-fuck we wanted and we could proceed one frame at a time, or go back a little if fuked something up.

An ultra bug time would often take 10hours and 200 internal errors, even if it's simple short lev like over and under. But it takes less time and less errors the more you start to understand about bugs and internal errors and stüf

And there never was any computer optimization done on anything, at least not in this project. At least I never heard of it.

_________________
Image


Top
 Profile  
 
PostPosted: 30 Oct 2014, 11:32 
Offline
Kuski
User avatar

Joined: 5 Dec 2008, 16:19
Posts: 1013
Location: a position or point in physical space
So will you release anything or it's just one big attention whore project where you just show couple of things every year or so, to make sure people will still talk about it?

_________________
Lousku wrote:
could you mayke shorter sig please :( mega annoying and also against rules :()


Top
 Profile  
 
PostPosted: 30 Oct 2014, 11:39 
Offline
36mins club
User avatar

Joined: 4 Jul 2004, 12:10
Posts: 5167
Chris wrote:
So will you release anything or it's just one big attention whore project where you just show couple of things every year or so, to make sure people will still talk about it?

pls stfu

_________________
TT: 36:59:53 || Avg TT: 38:09:65


Top
 Profile  
 
PostPosted: 30 Oct 2014, 12:00 
Offline
Kuski
User avatar

Joined: 5 Dec 2008, 16:19
Posts: 1013
Location: a position or point in physical space
Lukazz wrote:
Chris wrote:
So will you release anything or it's just one big attention whore project where you just show couple of things every year or so, to make sure people will still talk about it?

pls stfu

No. Had they released the whole thing years ago, everybody would have moved on years ago. The only point of this topic is to talk about that 32tt project. Given that this would be not first publicly available cheat tool, I don't see what further damage can it do if at all. All wrs and any other times have to be driven in EOL anyway, which I guess can protect against such a s/l tools. It's not that people would suddenly use only these tools and abandon legitimate WRing, because it didn't happen even after Smibu released his tool. It's just mila and zweq that can't make their minds or they don't want to do this.

Image

_________________
Lousku wrote:
could you mayke shorter sig please :( mega annoying and also against rules :()


Top
 Profile  
 
PostPosted: 30 Oct 2014, 12:12 
Offline
35mins club
User avatar

Joined: 28 Nov 2002, 15:54
Posts: 3701
Location: suo mesta
During short discussion IRC with mila few days ago, my condition for this thread to exist was that the tool will be published, because otherwise this thread would look exactly how chris described. I'm not sure what mila wants to change in the tool, but at least there is a rare bug with 5+ min rides that needs ficksing.

[17:28] <zamppe> yeah go for it if you will publlish the tool also later
[17:42] <zamppe> or just do the post when you can also post the exe
[17:48] <zamppe> then ppl would speculate less about the point of the post and instead spend time with the tool and stuff

_________________
Image


Top
 Profile  
 
PostPosted: 30 Oct 2014, 12:33 
Offline
Kuski
User avatar

Joined: 27 Jul 2003, 06:44
Posts: 611
Location: Ikrény, Hungary
just link the recz worth watching and done... :)

_________________
Elasto Mania - 34:37.34


Top
 Profile  
 
PostPosted: 30 Oct 2014, 13:10 
Offline
38mins club
User avatar

Joined: 2 May 2005, 14:20
Posts: 907
Location: Hungary
Recs and tool actually could give some health to the internal scene which is dying day from day. Why so big problem for bene the new styles or some new moves? Then you should not watch any edq or other players new styles either.

_________________
Team SPEED

Image

Q: Which number Bjenn plays on roulette?
A: 29.


Top
 Profile  
 
PostPosted: 30 Oct 2014, 13:30 
Offline
Kuski
User avatar

Joined: 27 Jul 2003, 06:44
Posts: 611
Location: Ikrény, Hungary
smarter ppl already realized that there is no point in knowledge if you don’t share it with others... this is true for this situation as well. the whole community will benefit from it

_________________
Elasto Mania - 34:37.34


Top
 Profile  
 
PostPosted: 30 Oct 2014, 14:04 
Offline
Hot kuski
User avatar

Joined: 18 Aug 2003, 23:33
Posts: 484
Location: Sweden
oke zweq interesting that I am even partially right since I am jsut guessing based on a dated list.

I would lose all interest in the non tas internal scene after a share if there are 25-30 recs/styles/whatever that impact too much, as the game would turn into a mindless boring höyl for a few years that is better suited for tas.
But fak it, let's be smarter people and share everything to kill this game finally so I can move on with my life and start playing some other fun game with anpdad or someone else. Is it still poe that is hip to paly???

The only way I would be motivated to play after a share like this is if the share is in private to me, because then I can beat all the wrs and confuse igge))))

_________________
Image
Image
Image
Signatür ruined by SveinR - smaller plz :*


Top
 Profile  
 
PostPosted: 30 Oct 2014, 14:29 
Offline
Cheatless
User avatar

Joined: 19 May 2002, 17:05
Posts: 4246
i never got that rare bug zweq mentioned, so not sure how to fix and what causes it
maybe because i haven't played 10mins externals, might be some overflow somewhere
i don't really plan to do any update, i also think zweq has the latest working version (mine might be a bit fucked atm and the compiler does not work in win8)
so i guess he can release it any time he likes:)

_________________
[carebox]


Top
 Profile  
 
PostPosted: 30 Oct 2014, 14:33 
Offline
36mins club
User avatar

Joined: 4 Jul 2004, 12:10
Posts: 5167
bene wrote:
But fak it, let's be smarter people and share everything to kill this game finally

smibu's S/L didn't kill the game, i don't see why this tool should. i love watching TAS-recs for inspiration and gerenal awesomness, but playing without tools is still the only real deal. i don't think this will change with a few other tools available.

GOGO ZWEQ!

_________________
TT: 36:59:53 || Avg TT: 38:09:65


Top
 Profile  
 
PostPosted: 30 Oct 2014, 14:51 
Offline
35mins club
User avatar

Joined: 28 Nov 2002, 15:54
Posts: 3701
Location: suo mesta
bene wrote:
oke zweq interesting that I am even partially right since I am jsut guessing based on a dated list.

I would lose all interest in the non tas internal scene after a share if there are 25-30 recs/styles/whatever that impact too much, as the game would turn into a mindless boring höyl for a few years that is better suited for tas.
But fak it, let's be smarter people and share everything to kill this game finally so I can move on with my life and start playing some other fun game with anpdad or someone else. Is it still poe that is hip to paly???

The only way I would be motivated to play after a share like this is if the share is in private to me, because then I can beat all the wrs and confuse igge))))


What are you smoking again( it's not like kazan 49 jaws is useful to normal playing in any way. Neither will any of these recs be. Imo it has been apparent for a long time already that TAS and nonTAS are quite far from each other and this tool will only make the gap larger. Or at least I think it's more flexible than hourglass because it has been tailored specifically for elma

_________________
Image


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 125 posts ]  Go to page 1, 2, 3  Next

All times are UTC [ DST ]


Who is online

Users browsing this forum: culinko, Haruhi and 2 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB® Forum Software © phpBB Group