KZ-Timer vs REAL-Timer?
23/03/2009 20:29Originally posted by omiloI would like to see maps being beatable by tenth of seconds, for example: WR being 25.2 and in order to beat it you need at least 25.1. And the timer should be amxx or fps death, so that the times are real. That's the fairest way imo.
I agree with this, but sooner or later we are probably gonna hear complaints about this as well, cause oh noes, now clintmo_longerjumper[x] is unbeatable again..........
If such things as timerbugs did not exist, I would absolutely have voted for keeping the Kreedz counter, but as things looks now, I guess it's better to use Fps-Death or an AMXX plugin to check demos.
I got one question:
Will both the Kreedz counter time (may be timerbugged) and Fps-Death/plugin time be listed in the record section?
totally agree with thopa.
if demo will be beaten for 0.1 seconds it will be faster no matter what think about this previous holder of this demo. he ofc will be sad and willl not be agree but it's a fact. i think optimal will be 0.01 sec.
sry for bad eng)
miliseconds. there is a lot of people that have beaten wrs with the same second but faster.
but on the website have to put miniseconds :)
i think some people forget it's still a WR, thus hard to beat.
Originally posted by goodhoopoetotally agree with thopa.
if demo will be beaten for 0.1 seconds it will be faster no matter what think about this previous holder of this demo. he ofc will be sad and willl not be agree but it's a fact. i think optimal will be 0.01 sec.
sry for bad eng)
What about the person who actually beat the time with 0.01 sec. Won't he be sad if his demo isn't added as WR? One person will always be sad, no matter what choice you make. That's why it's more correct to replace the slower demo with the quicker demo. :)
23/03/2009 21:40I've posted some arguments earlier, but I'll just squeeze it in a nutshell now: There's no point of letting the slower run be the WR, and WR should be beatable even with the slightest amount of time (0,01 second I believe, since 100fps is max and you need 1 frame for use).
thopa goes phsycologist :P
Originally posted by m0verFirst one. Reason is that we are in 21st century and if technology allows for perfection, we should use it. KZ timer is obsolete, what I would like is to replace all timers in all maps with perfectly accurate one, even though it is infeasible.
There is no real argument for old KZ timer except nostalgia and fanboys. And I don't care for neither. If someone is faster because he geeked the map, then he is the fastest.
Should we record Usain bolt with old mechanical stopwatches because it is the real spirit of running? Times change, and people should cope with it. So, down with the inaccurate KZ timer, it isn't even used in public servers so why should it be used anyway.
nuff said
miliseconds for sure.
lol, demos are beated by 1 or more sec and you think that demos can be beated by 0.01 seC? lol
I don't even feel like posting the same arguments over and over again.. But thanks faker, for not making this in to a poll.
This will only increase the competition - and an increased competition will only give us even faster demos.
Now I can't think any reasonable arguments for not having the fastests demo as WR. Period!
Now let's get this done.
arguments for a float-timer (mm:ss:xxxxxx):
- every kind of sports or competition in which the records are displayed as time use decimal places to show an exact time to compare the participants
- using decimal places would result in a lot of new records which would make the world-/country-/communityrecords even better
- time is time, saying a 75.6 seconds record is the wr although demos with a time of 75.5 or even less exist is simply wrong
- this could help xj to look much healthier (activity) than it looks now
the time should be as accurate as possible but i also agree with 2 decimal places. but, i would like to remember that a lot of decimal places would be interesting for short maps such as clintmo_longjumpers.
numb talked to me telling me that 2 places are the accuratest way to use a float-timer due to the 100fps which are allowed but i think this is not (fully) correct. cs calculates as many frames as possible per second, for example it calculates in the first second 100 frames, in the second 99. this means it calculated 99.5 frames within 2 seconds. since cs is a frame based game it means that the player has x frames per second in which he can press the buttons, result: the time is not ticking like 0.000001, 0.000002... it is ticking like 0.0000010050... 0.000002010... (each frame 1/average_fps is added). knowing this means that fps will be even more important than they are now.
arguments for amxx:
- the climbtime can be shown on pressing the stopbutton, you do not need to check every demo in fps death before sending it in
(basically an accurate timer system should be used but the argument above is in my eyes a feature which fps death does not provide although it should if it will be used to comapre demos)
i would prefer an amxx-timer since im sure that the timer-system kzm used or me uses in the xj-plugin are kind of accurate (99.9999% due to rounding issues). old timer systems used a system to detect player's pressing the button which's accuracy was based on the server's fps (which is ~100 in case of record-demos). nowadays a new way to detect certain things is available in amxx with which i can trigger players pressing any button as soon as the engine detects this.
i just tried to compare the amxx-time with fps death but unfortunately i could not because fps death does not detect the climbtime in demos which use the xj plugin. i am going to ask edark a few things as soon as i am able to catch him in skype, irc or steam so i can tell you whether amxx is accurate enough for this.
Originally posted by omiloI would like to see maps being beatable by tenth of seconds, for example: WR being 25.2 and in order to beat it you need at least 25.1. And the timer should be amxx or fps death, so that the times are real. That's the fairest way imo.
That's the truth
Beating it by 0.01 wont make XJ active in any way just because if you see your 30 WRs beated by a 0.01 or 0.02 you'll be so pissed off and you'll quit
Kreedz-Timer based on seconds!!!!!!!!!!
OLD FTW
forget about my frame-based thing idea... recording with fps_max 90 would help a lot on long maps... ^^ edited my old post.
Beating by 1/10 second is the best choice. For two reason:
1. This is the minimum time unit that human can clearly sense. Second is too long, and 0.01s is too short.
2. Old timer are out-dated and should be thrown away. Better things will finally be used regardless of how much people love and miss the old, it's exactly the same as why this site is updating new WRs.
Originally posted by fantasistBeating by 1/10 second is the best choice. For two reason:
1. This is the minimum time unit that human can clearly sense. Second is too long, and 0.01s is too short.
2. Old timer are out-dated and should be thrown away. Better things will finally be used regardless of how much people love and miss the old, it's exactly the same as why this site is updating new WRs.
That would just bring up a new discussion in a year, cus someone has improved a record by 0.09 secs and is not rewarded. If you beat the World Record - you got it.
Originally posted by eusunt #cCBeating it by 0.01 wont make XJ active in any way just because if you see your 30 WRs beated by a 0.01 or 0.02 you'll be so pissed off and you'll quit
Of course we wouldn´t see a whole bunch of demos getting beaten by 0.01, what on earth makes you think that?
And even if it did happen, which is highly highly unlikely, who cares? Nobody is forcing you to watch them.
I'm a very strong believer in the faster is better and accuracy.
Though I should not have very much say in this because I do not participate at the pro level, nor do I watch the demos.
But honestly I do not see any reason as to why its a bad thing. I've seen the complaint about its not fun to watch a demo thats identical, and to that I quote cheche "who cares? Nobody is forcing you to watch them."
At first I wrote quite a long piece about my opinions on this, but I deleted it and I've decided that it's very clear that a world record is the fastest time round with the highest accuracy of measuring the time that's practically possible.
As for measuring this, there are a lot of problems when you consider records are set on lan servers for various reasons and the internet isn't at a stage where everybody is running on a connection fast enough or stable enough to create a perfectly fair global online system.
I'm going to go out on a limb and say that most people share the opinion that if you can measure a run to be faster that's a world record, but I think that this discussion brings together 2 discussions at once. The first discussion I feel is the simple one, should we strive for as accurate timing as possible. The second discussion is how we can go about doing this.
I'm no expert when it comes to the coding of CS or the nature of measuring time and relating that to things such as FPS, but I know that there are people out there who are. Using an AMX plugin raises some issues such as it having to be a flawless plugin that couldn't be exploited and could always be proven to be legit in a demo, also with it being on a LAN would an amx plugin be effected by the users fps or the system it was running on? Also, what if it were updated? Everybody has to download the new version in order to register a record? What about the records set with old versions?
I'm going to assume that FPS-death could display complete time more accurately than it does, if so, that is a perfectly reasonable and practical solution. No matter what the ingame timer says and no matter what the run looks like, if, assuming FPS-death can be trusted, a run has a shorter complete time after being analyzed then that is the world record. Beat it by 0.00001 and you hold the world record.
Originally posted by ChaosADOriginally posted by kayneOriginally posted by KunquiI voted for miniseconds :)
Please log in to post replies.