Upvote Upvoted 0 Downvote Downvoted
1 2 3
I'm writing a proper guide to performance.
posted in Q/A Help
61
#61
-4 Frags +
mastercomsWhen you're driving a car, the speedometer does not show an average of your speed over the last hour, despite reading "miles per hour".

Good point…

mastercomsAnd you can easily calculate hours per mile by taking 1 over your speed. This is just how units and math work.

I have to be missing something at this point. Frametimes can vary in each frame (e.g one frame could take 100ms to render while another could take 400ms, but 1⁄2 = 500 ms), so why bother deriving the frametime from frame rate when the frame time will vary anyways? I mean frame rate is counted over a second right?

[quote=mastercoms]When you're driving a car, the speedometer does not show an average of your speed over the last hour, despite reading "miles per hour".[/quote]

Good point…

[quote=mastercoms]And you can easily calculate hours per mile by taking 1 over your speed. This is just how units and math work.[/quote]

I have to be missing something at this point. Frametimes can vary in each frame (e.g one frame could take 100ms to render while another could take 400ms, but 1⁄2 = 500 ms), so why bother deriving the frametime from frame rate when the frame time will vary anyways? I mean frame rate is counted over a second right?
62
#62
17 Frags +

You made a whole thread and all this shit just to recommend mastercomms?

https://i.gyazo.com/2cb50a054dce06c7cd88e237ad6cb961.png

You made a whole thread and all this shit just to recommend mastercomms?

[img]https://i.gyazo.com/2cb50a054dce06c7cd88e237ad6cb961.png[/img]
63
#63
-12 Frags +
alfaYou made a whole thread and all this shit just to recommend mastercomms?

https://i.gyazo.com/2cb50a054dce06c7cd88e237ad6cb961.png

That is a huge oversimplification; there is more to this guides than just mastercomfig. I also recommend frametimes over frame rate and tell people that bad servers will cost your server.

[quote=alfa]You made a whole thread and all this shit just to recommend mastercomms?

[img]https://i.gyazo.com/2cb50a054dce06c7cd88e237ad6cb961.png[/img][/quote]

That is a huge oversimplification; there is more to this guides than just mastercomfig. I also recommend frametimes over frame rate and tell people that bad servers will cost your server.
64
#64
20 Frags +

i find it hard to believe this is not all an elaborate troll

i find it hard to believe this is not all an elaborate troll
65
#65
11 Frags +
AimIsADickFrametimes can vary in each frame (e.g one frame could take 100ms to render while another could take 400ms, but 1⁄2 = 500 ms)

That first frame you mentioned is 1 frame / .1 seconds = 10FPS, and the second one you mentioned is 1 / .4 = 2.5FPS. See I made the FPS vary each frame too.

AimIsADickwhy bother deriving the frametime from frame rate when the frame time will vary anyways? I mean frame rate is counted over a second right?

This is like asking why the SI unit hertz exists when we have seconds. They're used for different calculations and comparisons.

[quote=AimIsADick]Frametimes can vary in each frame (e.g one frame could take 100ms to render while another could take 400ms, but 1⁄2 = 500 ms)[/quote]

That first frame you mentioned is 1 frame / .1 seconds = 10FPS, and the second one you mentioned is 1 / .4 = 2.5FPS. See I made the FPS vary each frame too.

[quote=AimIsADick]why bother deriving the frametime from frame rate when the frame time will vary anyways? I mean frame rate is counted over a second right?[/quote]

This is like asking why the SI unit hertz exists when we have seconds. They're used for different calculations and comparisons.
66
#66
17 Frags +
AimIsADickbad servers will cost your server.

https://i.imgur.com/fmk1KMf.png

[quote=AimIsADick]bad servers will cost your server.[/quote]
[img]https://i.imgur.com/fmk1KMf.png[/img]
67
#67
-1 Frags +

posting in legendary thread

posting in legendary thread
68
#68
-9 Frags +
mastercomsAimIsADickFrametimes can vary in each frame (e.g one frame could take 100ms to render while another could take 400ms, but 1⁄2 = 500 ms)
That first frame you mentioned is 1 frame / .1 seconds = 10FPS, and the second one you mentioned is 1 / .4 = 2.5FPS. See I made the FPS vary each frame too.

Thing is that frame rate would be extremely misleading here, since the variance is practically invisible with just frame rate. The frame rate would only match up the total amounts if all of the frames rendered in the same amount of time, which isn't always the case. I'd feels like I'm just repeating my self here tho…

mastercomsAimIsADickwhy bother deriving the frametime from frame rate when the frame time will vary anyways? I mean frame rate is counted over a second right?
This is like asking why the SI unit hertz exists when we have seconds. They're used for different calculations and comparisons.

Can you elaborate on this point actually? I don't understand what you mean.

Hang on a second. I thought we were arguing on the reliability of frame rate, but we're really arguing on whether frame time is the same metric as frame rate. I think I get your logic here (that since frame rate is a frequency, frametime can technically be derived from frame rate and thus be used to get the frame rate), but I just don't agree that derivative metric are the same as their parent metric.

If this turns out to be basic stuff, then sorry about that. I just didn't learn about this beforehand.

[quote=mastercoms][quote=AimIsADick]Frametimes can vary in each frame (e.g one frame could take 100ms to render while another could take 400ms, but 1⁄2 = 500 ms)[/quote]

That first frame you mentioned is 1 frame / .1 seconds = 10FPS, and the second one you mentioned is 1 / .4 = 2.5FPS. See I made the FPS vary each frame too.[/quote]

Thing is that frame rate would be extremely misleading here, since the variance is practically invisible with [i]just[/i] frame rate. The frame rate would only match up the total amounts if all of the frames rendered in the same amount of time, which isn't always the case. I'd feels like I'm just repeating my self here tho…

[quote=mastercoms][quote=AimIsADick]why bother deriving the frametime from frame rate when the frame time will vary anyways? I mean frame rate is counted over a second right?[/quote]

This is like asking why the SI unit hertz exists when we have seconds. They're used for different calculations and comparisons.[/quote]

[s]Can you elaborate on this point actually? I don't understand what you mean.[/s]

Hang on a second. I thought we were arguing on the reliability of frame rate, but we're really arguing on whether frame time is the same metric as frame rate. I think I get your logic here (that since frame rate is a frequency, frametime can technically be derived from frame rate and thus be used to get the frame rate), but I just don't agree that derivative metric are the same as their parent metric.

If this turns out to be basic stuff, then sorry about that. I just didn't learn about this beforehand.
69
#69
newbie.tf
5 Frags +
AimIsADickThing is that frame rate would be extremely misleading here, since the variance is practically invisible in frame rate. The frame rate would only match up the total amounts if all of the frames rendered in the same amount of time, which isn't always the case. I'd feels like I'm just repeating my self here tho…

fps changes every frame. it's not actually measuring the past second.

it is much easier for us to think in terms of fps because it operates on a much easier scale for folks to understand (generally 30-300 except in extremes) as opposed to spf which is weird (.3-.03 i think lol?)

it's like how we measure car speed in miles/hour, but it's not actually the amount of distance we've traveled in the past hour. just the amount we would travel if we were to continue at that exact speed the next hour

AimIsADick(It feels like we're getting nowhere with this conversation…)

I promise you that isn't mastercomm's fault.

[quote=AimIsADick]Thing is that frame rate would be extremely misleading here, since the variance is practically invisible in frame rate. The frame rate would only match up the total amounts if all of the frames rendered in the same amount of time, which isn't always the case. I'd feels like I'm just repeating my self here tho…[/quote]

fps changes every frame. it's not actually measuring the past second.

it is much easier for us to think in terms of fps because it operates on a much easier scale for folks to understand (generally 30-300 except in extremes) as opposed to spf which is weird (.3-.03 i think lol?)

it's like how we measure car speed in miles/hour, but it's not actually the amount of distance we've traveled in the past hour. just the amount we would travel if we were to continue at that exact speed the next hour

[quote=AimIsADick](It feels like we're getting nowhere with this conversation…)[/quote]
I promise you that isn't mastercomm's fault.
70
#70
-9 Frags +
KevinIsPwnAimIsADickThing is that frame rate would be extremely misleading here, since the variance is practically invisible in frame rate. The frame rate would only match up the total amounts if all of the frames rendered in the same amount of time, which isn't always the case. I'd feels like I'm just repeating my self here tho…
fps changes every frame. it's not actually measuring the past second.

it is much easier for us to think in terms of fps because it operates on a much easier scale for folks to understand (generally 30-300 except in extremes) as opposed to spf which is weird (.3-.03 i think lol?)

it's like how we measure car speed in miles/hour, but it's not actually the amount of distance we've traveled in the past hour. just the amount we would travel if we were to continue at that exact speed the next hour

Yeah I noticed that with cl_showfps. It was changing constantly and I don't believed it properly counted the amount of frames rendered in a second. So if FPS changes every frame, then what would be the proper metric? I really think it's just better to display the amount of frames as an amount rather than a frequency.

KevinIsPwnAimIsADick(It feels like we're getting nowhere with this conversation…)I promise you that isn't mastercomm's fault.

It's likely my fault at this point then. I just don't understand this logic that derivative metrics = their parent metrics. Don't derivative metrics measure different stuff from their parents?

[quote=KevinIsPwn][quote=AimIsADick]Thing is that frame rate would be extremely misleading here, since the variance is practically invisible in frame rate. The frame rate would only match up the total amounts if all of the frames rendered in the same amount of time, which isn't always the case. I'd feels like I'm just repeating my self here tho…[/quote]

fps changes every frame. it's not actually measuring the past second.

it is much easier for us to think in terms of fps because it operates on a much easier scale for folks to understand (generally 30-300 except in extremes) as opposed to spf which is weird (.3-.03 i think lol?)

it's like how we measure car speed in miles/hour, but it's not actually the amount of distance we've traveled in the past hour. just the amount we would travel if we were to continue at that exact speed the next hour[/quote]

Yeah I noticed that with cl_showfps. It was changing constantly and I don't believed it properly counted the amount of frames rendered in a second. So if FPS changes every frame, then what would be the proper metric? I really think it's just better to display the amount of frames as an amount rather than a frequency.

[quote=KevinIsPwn][quote=AimIsADick](It feels like we're getting nowhere with this conversation…)[/quote]
I promise you that isn't mastercomm's fault.[/quote]

It's likely my fault at this point then. I just don't understand this logic that derivative metrics = their parent metrics. Don't derivative metrics measure different stuff from their parents?
71
#71
0 Frags +

.

.
72
#72
newbie.tf
1 Frags +
AimIsADickDon't derivative metrics measure different stuff?

no.

miles/hour communicates the same thing as hours/mile. there is no functional or mathematic difference. even if hours/mile were better at communicating the measurement linguistically (there's a decent argument there), there's no functional benefit warranting the switch to an entirely new system.

i have no idea if a frametime of __ ms is good or bad, but i immediately understand what __ fps visually is. that is the crux of what everyone in this thread is trying to tell you

[quote=AimIsADick]Don't derivative metrics measure different stuff?[/quote]
no.

miles/hour communicates the same thing as hours/mile. there is no functional or mathematic difference. even if hours/mile were better at communicating the measurement linguistically (there's a decent argument there), there's no functional benefit warranting the switch to an entirely new system.

i have no idea if a frametime of __ ms is good or bad, but i immediately understand what __ fps visually is. that is the crux of what everyone in this thread is trying to tell you
73
#73
-5 Frags +
KevinIsPwnAimIsADickDon't derivative metrics measure different stuff?no.

miles/hour communicates the same thing as hours/mile. there is no functional difference. even if hours/mile were better at communicating the measurement linguistically (there's a decent argument there), there's no benefit worth switching to an entirely new system.

i have no idea if a frametime of __ ms is good or bad, but i immediately know what __ fps means.

Ohhh. This whole time I conflated representation with function. Yeah nevermind I agree with you and mastercoms now.

[quote=KevinIsPwn][quote=AimIsADick]Don't derivative metrics measure different stuff?[/quote]
no.

miles/hour communicates the same thing as hours/mile. there is no functional difference. even if hours/mile were better at communicating the measurement linguistically (there's a decent argument there), there's no benefit worth switching to an entirely new system.

i have no idea if a frametime of __ ms is good or bad, but i immediately know what __ fps means.[/quote]

Ohhh. This whole time I conflated representation with function. Yeah nevermind I agree with you and mastercoms now.
74
#74
7 Frags +
AimIsADick

How are you still posting on this website

[quote=AimIsADick][/quote]
How are you still posting on this website
75
#75
0 Frags +

who would win? mastercomms who has opened the video game up and optimised it or aimisadick who has contributed nothing to note

who would win? mastercomms who has opened the video game up and optimised it or aimisadick who has contributed nothing to note
76
#76
7 Frags +

74 posts later we arrive at the astonishing conclusion that "which is better x or 1/x" is a meaningless question

74 posts later we arrive at the astonishing conclusion that "which is better x or 1/x" is a meaningless question
77
#77
newbie.tf
9 Frags +
plunkwho would win? mastercomms who has opened the video game up and optimised it or aimisadick who has contributed nothing to note

friendly reminder to anyone in this thread that mastercomms has spend a shitload of time pushing fps configs pretty much as far as we can reasonably expect and that there is a place to donate for that work here

[quote=plunk]who would win? mastercomms who has opened the video game up and optimised it or aimisadick who has contributed nothing to note[/quote]
friendly reminder to anyone in this thread that mastercomms has spend a shitload of time pushing fps configs pretty much as far as we can reasonably expect and that there is a place to donate for that work [url=https://docs.mastercomfig.com/en/latest/support_me/]here[/url]
78
#78
-4 Frags +
plunkwho would win? mastercomms who has opened the video game up and optimised it or aimisadick who has contributed nothing to note

I mean I made a generator to simplify the process of creating command-based voicemenus. I'm also thinking about a TF2 chat translator, and I have a (somewhat abandoned but cant have enough time to work on) project to document launch options from the 2018 TF2 leak.

That's about it tho… mastercoms definitely would win rn.

[quote=plunk]who would win? mastercomms who has opened the video game up and optimised it or aimisadick who has contributed nothing to note[/quote]

I mean [url=https://github.com/WhyIsEvery4thYearAlwaysBad/CmdMenuGenerator]I made a generator[/url] to simplify the process of creating command-based voicemenus. I'm also thinking about a TF2 chat translator, and I have a [url=https://github.com/WhyIsEvery4thYearAlwaysBad/TF2LeakInfo] (somewhat abandoned but cant have enough time to work on) project to document launch options from the 2018 TF2 leak[/url].

That's about it tho… mastercoms definitely would win rn.
79
#79
4 Frags +
plunkaimisadick who has contributed nothing to note

watch how you talk about the man who solved the ubersaw conundrum

[quote=plunk]aimisadick who has contributed nothing to note[/quote]
watch how you talk about the man who solved the ubersaw conundrum
80
#80
-5 Frags +
Adnurakplunkaimisadick who has contributed nothing to notewatch how you talk about the man who solved the ubersaw conundrum

It feels like you're being serious and sarcastic at the same time.

[quote=Adnurak][quote=plunk]aimisadick who has contributed nothing to note[/quote]
watch how you talk about the man who solved the ubersaw conundrum[/quote]

It feels like you're being serious and sarcastic at the same time.
81
#81
13 Frags +

https://i.imgur.com/0SdD5yT.mp4

https://i.imgur.com/0SdD5yT.mp4
82
#82
6 Frags +
AimIsADickIt feels like you're being serious and sarcastic at the same time.

One of those is definitely true

[quote=AimIsADick]It feels like you're being serious and sarcastic at the same time.[/quote] One of those is definitely true
83
#83
3 Frags +
Adnurakplunkaimisadick who has contributed nothing to notewatch how you talk about the man who solved the ubersaw conundrum

youre right i should respect the man himself

[quote=Adnurak][quote=plunk]aimisadick who has contributed nothing to note[/quote]
watch how you talk about the man who solved the ubersaw conundrum[/quote]
youre right i should respect the man himself
84
#84
-3 Frags +
turbochad69https://i.imgur.com/0SdD5yT.mp4

Oh.

det-AimIsADickIt feels like you're being serious and sarcastic at the same time. One of those is definitely true

Sarcastic?

[quote=turbochad69]https://i.imgur.com/0SdD5yT.mp4[/quote]

Oh.

[quote=det-][quote=AimIsADick]It feels like you're being serious and sarcastic at the same time.[/quote] One of those is definitely true[/quote]

Sarcastic?
85
#85
15 Frags +

Can we debate if 1+1 is actually equal to 2 next?

Can we debate if 1+1 is actually equal to 2 next?
86
#86
8 Frags +
Tino_Can we debate if 1+1 is actually equal to 2 next?

Yeah but you can represent that far more effectively as 1(d/dx(x)+ (14^0)) because the inclusion of derivative functions is imperative to identify the mathematical boundaries stated by the expression. 1+1 is just an over simplification.

[quote=Tino_]Can we debate if 1+1 is actually equal to 2 next?[/quote]

Yeah but you can represent that far more effectively as 1(d/dx(x)+ (14^0)) because the inclusion of derivative functions is imperative to identify the mathematical boundaries stated by the expression. 1+1 is just an over simplification.
87
#87
3 Frags +
Tino_Can we debate if 1+1 is actually equal to 2 next?

http://tachyos.org/godel/1+1=2.html

[quote=Tino_]Can we debate if 1+1 is actually equal to 2 next?[/quote]
http://tachyos.org/godel/1+1=2.html
88
#88
5 Frags +

hooking up my fps counter to datadoghq dot com to get Real Observability and be able to make Educated Decisions based on Real World Data

hooking up my fps counter to datadoghq dot com to get Real Observability and be able to make Educated Decisions based on Real World Data
89
#89
1 Frags +

https://www.youtube.com/watch?v=DRoSzCS3fO8

[youtube]https://www.youtube.com/watch?v=DRoSzCS3fO8[/youtube]
1 2 3
Please sign in through STEAM to post a comment.