@Tom94

Tom Long Dong

Ask @Tom94

Sort by:

LatestTop

Previous

Skystar Toumei HDHR could be another 600pp score, he has 6x100 missed without HD and that would've been 593pp, with HD I'm guessing it's about 650?

I just named some examples. There are plenty of opportunities for him to get >600pp scores. Not like >600pp scores are the only ones which are actually amazing.

This will Probably not get answered but Why do you say nobody is close to C are your eyes closed or something C at the moment cant Even Set a 600 PP Score and there are other Players that Are as good in nomod as him

He played Image Material exactly three times when he got unbanned. The first two runs each had the highest accuracy achieved on the map _ever_. He interrupted the third run, because his acc dropped under 99%. He _never_ missed in any of the runs. Sadly he sliderbroke twice.
Do I need to mention his two FCs on https://osu.ppy.sh/b/724015?m=0 (the one before he got banned was 2x 100, mind you, https://www.youtube.com/watch?v=Yr7PPqaI6MUTom94’s Video 134437556686 Yr7PPqaI6MUTom94’s Video 134437556686 Yr7PPqaI6MU )? Or this SS https://www.youtube.com/watch?v=3rd7xxbfRoETom94’s Video 134437556686 3rd7xxbfRoETom94’s Video 134437556686 3rd7xxbfRoE ? Or time freeze? Or his random play mendes DT FC?
Jesus, he has been unbanned for 3 days and already set score nobody has ever done before on some of the hardest maps. Not having a 600pp score right now doesn't mean anything, given that my system is far from perfect. When he FCs time freeze (which he was more than close to) he'll have one at the very least.

View more

Tom94’s Video 134437556686 Yr7PPqaI6MUTom94’s Video 134437556686 Yr7PPqaI6MU

I heard that Cookiezi's old scores were calculated to be about 9800pp. Assuming that's true, if he had kept those scores, would be be #1 with the scores he's set since being unbanned?

I have no way of knowing whether that's true without checking the calculation myself, which would be very time consuming. Sorry, I got better stuff to do.

Related users

Do you think cookiezi can compete with hvick and raise despite lacking 10.3+ burst speed and reading, as well as lacking top level hd reading?

His HD is pretty solid actually. Also not sure whether his burst speed is really much lower. I've seen him FC 300bpm maps with bursts (albeit at ~95%) back in the day.
Liked by: pyon Kala~

Thoughts on Rafis sniping shigetoras scores?

Isn't that exactly what we've been so hyped for? The current top players battling it out with Cookiezi? I don't get people who are against _competing_ in a _competetive_ game. Tough luck if your #1 scores get taken. Make a better score or move on and practice until you can. Imagine if in football it was considered inappropriate to score goals against weaker enemies. What the fuck?
Sorry for the rant, I know you didn't condemn Rafis in your question, but seeing some people's reactions disgusts me. I want him to continue sniping. Don't make it easy for Cookiezi to climb back to the top!

http://ask.fm/Tom94/answer/134310517198 Excuse my ignorance here, but since I believe this is the first time a high level player officially starts from a new, rather than an account with old scores, will he have to slowly build up scores? or will like going for 300-500pp at the start be a bad idea?

He is free to start with 500pp scores. No reason to artifically limit his pp gain. He'll have to build up the 416 bonus pp from having many scores, though. To be honest, not having these extra pp makes it even more interesting for me, since he'll be at a small disadvantage compared to other high ranking players, pushing him even harder.

So now that cookiezi's unbanned, do you reckon he will take number 1 within a couple months?

I hope he will. Would be fun to watch. :)
If he can consistently make 500+pp scores he might even make it within a couple of days, but he'd have to play a lot. With pp getting 50-100 good scores is enough for a super high rank.
Liked by: DroppedBass

How good is good enough to appreciate cookiezies plays? Top 100, 1000, 5000? Any specific plays you have in mind?

This is not something with a well-defined threshold. The better you are, the better your insight into other people's osu! skills becomes. In fact, it's probably not even connected that much to your actual skill, but rather to your experience with the game, i.e. how much you played, tried, observed, etc.

I don't really like it when people mention players who have stopped playing (rrtyui in your earlier answer), there's nothing really wrong with it, but it somehow ticks me off when people compare active players to inactive players, thoughts on this?

I'm completely fine with discussion players who stopped playing. I don't understand why it'd be wrong, and well, you said it yourself, neither do you.
Liked by: DroppedBass Shira

http://ask.fm/Tom94/answer/134268259790 well I agree with you mostly, but I could confidently say that hvick's and Rafis speed/double time skill is by far superior than Cookiezi's

While he has bandages / whatever around his arm and complains about pain all the time, maybe. I remember the days where he had just the same insane speed as rafis / hvick have now. No clue whether his obvious health issues will persist, though. If there's one lesson I learned from the past few years: Never underestimate Cookiezi.

This question will probably not be answered, but do you have any thoughts on the Shigetora appeal?

I want to see him wreck the leaderboards, GOD DAMNIT. His plays sometimes legitimately make me laugh out loud because of how unimaginably ridiculous they are. You need to be a pretty decent player yourself to even comprehend how crazy some of this stuff is. Yet whenever he disappears for a few months everyone comes and is all "OMG I think WWW finally reached Cookiezi's level". No he didn't, for fuck's sake. And neither did rrtyui. Or hvick. Or *insert any other player here*. [/rant]
No thoughts on the appeal specifically. All I can do is hope he is not a dick, so that he actually gets unbanned. Never actually interacted with him. :p

https://www.youtube.com/watch?v=2M49ylPzHus Rafis uploaded a highlight with you joking about nerfing DT at the end. I just want to inform you that your joke wasn't ruined at all because it also works as a reference to WWW taking #1 and you nerfing DT right after to increase his lead over Hvick

Did I? I don't remember. Don't think it was targeted at specific players, though.
Liked by: DroppedBass

Haskell has unboxed types. https://hackage.haskell.org/package/vector-0.11.0.0/docs/Data-Vector-Unboxed.html https://wiki.haskell.org/Performance/GHC#Unboxed_types https://downloads.haskell.org/~ghc/6.12.3/docs/html/users_guide/primitives.html

Just with builtins, though. You can't come and define your own unboxed types. Unboxed tuples may help a bit, but doing more than just a tiny bit of code with these seems kind of clunky.
Also I already linked your last site in an earlier answer.

http://programmers.stackexchange.com/questions/184586/devising-test-of-haskells-value-reference-semantics

I assume you refer to the accepted answer. Reference vs Value semantics is indeed irrelevant in terms of program behavior, but _not_ with respect to speed. Your programs will output the same, but at different rates. If all your objects are scattered across the heap, then your processor's cache is ineffective. If they are arranged tightly in memory, then your processor's cache will have a very large hit rate.
Liked by: DroppedBass

Nah, Haskell does allow usage of mutable arrays through Monads. https://hackage.haskell.org/package/vector-0.11.0.0/docs/Data-Vector-Mutable.html Haskell is a lot more flexible than most people think. You can get up to 95% the performance of C whilst having a much higher assurance of correctness.

"Mutable boxed vectors". Are you sure this thing has by-value semantics? Boxing sounds like it's doing, you know, boxing, which is defeating most of the point. I wonder if it's smart enough to _not_ box vectors of ints. It'd be hilariously bad if it boxed single ints.
Actually, looks like Haskell doesn't even support value semantics on custom data structures. https://downloads.haskell.org/~ghc/6.12.3/docs/html/users_guide/primitives.html
So... no matter what kind of container you make, in frequent relevant situations you'll be far behind languages which support such things.
Liked by: DroppedBass

I suppose if you're writing extremely performance critical code, C++ does win out over those languages. I've even seen some scientific computing code written in fortran because it's slightly more performant than C++ due to disallowing aliasing of pointers.

Well, it's not hard to label non-aliasing pointers with "__restrict" in your bottlenecking C++ code, so I'm not sure if this is really a plus for fortran. :P
Also I wouldn't say "extremely performance critical". Using linked lists (correct me if I am wrong, but I believe that's the only easily usable container in Haskell) instead of arrays (I'm including dynamic memory-local structures like std::vector here) gives you 100-fold performance loss in many cases. If your language doesn't support by-value semantics for objects you'll have similar issues, even if you can make a memory-local array of references.
Liked by: DroppedBass Benpoi

But there are lots of compiler frontends that output LLVM bytecode, LLVM compiles to really fast native code. Languages like Julia/Haskell/Scala are nearly the speed of C++, but are programmable at a much higher level of abstraction.

Have fun programming in Haskell with cache-local containers. Even if those languages emit LLVM bytecode, the instructions in it are not necessarily optimized or even optimizable to the same degree as what an equally simple C++ program would result in. I'm not saying these languages are bad - functional programming actually is amazing - but in the general case (there are some exceptions) it's a _lot_ easier to write programs optimized for the hardware they are running on (i.e. fast processors with slow memory and a bunch of caches plus some other quirks you can read up on if you'd like) in C++.
If someone has more specific questions about this to ask, feel free to. I've already written some compilers and done a bunch of optimization, so I do know some theory and how it works in practice.
Liked by: DroppedBass

So I heard that C++ is really difficult to learn and some people say it's a pain in the ass because you need much more lines of code for a simple program. Since you are already an advanced programer can you confirm that this is true ? Should I reconsider learning C++ or is it worth it ?

Needing more lines for a simple program is no longer as true with the new stuff C++ has been getting in the recent years. It is quite a lot harder to master than other languages, though, but on the flipside lets you do things you just can't do with other languages. Also if you know how to use it it's blazingly fast.
For me C++ is definitely worth it since I write all kinds of performance-critical programs and C++ has tons of nice features alternatives don't have. Also it has well-optimizing compilers and a crazy amount of third party libraries for almost anything you can imagine. That doesn't mean it's the best choice for you, though. Definitely consider it.

http://ask.fm/Tom94/answer/134068123342 funny, isn't that basically what you did? made TP and let peppy use it as ppv2

Yup. Well, "letting use" is a bit wrong, since I'm the one who programmed it again (for ppv2 we needed an implementation running on linux and support for a lot more scores than what tp had to deal with) and integrated it into osu! myself, including the then new star difficulty.
Anyway, I don't have time to work on osu! in the forseeable future (thesis work + upcoming PhD stuff), let alone ppv2, so someone else will have to take it from here. When osu! goes fully open source maybe some more people will tinker with it.
Liked by: DroppedBass

Next

Language: English