• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Developing a benchmark program

Originally posted by :LJ:
Depends what you're trying to test. If you're testing the hardware performance, then you write the most efficient code possible which hammers the subsystem in question. If you're testing a database system, then you'll probably concentrate on hammering it with the most load in particular areas. If you're testing an enterprise software platform, you'll most likely test memory handling performance (not just bandwidth) and network performance.

Unlike STL, you have to realise that the end product should result from a fit-for-purpose design, rather than any personal prejudice.
Click to expand...

It depends why you're benchmarking aswell. If you want to see how fast something is (to see a if a specific product is suitable for a specific project), rather than how fast it is relative to other systems (PC upgrades), then efficiency matters alot more.
 
Code efficiency doesn't really matter here to me, now it would be very important if I was writing software where speed mattered. I don't think speed matters here, what matters is relative speed. I'm not writing a compression algorithm so that it can be the fastest, not that it would be a bad thing, but so that it can be tested on multiple computers.

Actually I feel bad about saying code efficiency doesn't matter because it actually does. The benchmark interface itself has to be efficient. The algorithms don't have to be efficient in this case, but the interface does. I'm not writing winzip, I'm just writing a benchmark. This is not meant to be encryption or compression software where speed matters.

I have another question. As far as a score for the benchmark, my thought right now is to have it just be a time. You'll have an overall time and a time for each test, probably ten. A 3d test would be in fps, which actually screws things up a little. Should the test output a score when you run all the tests or should it output time, and not count the 3d test, if there is one. It could just output the fps as a sidebar. The users would also be allowed to run as many tests as they want.
 
[Whatsisname]
> Yes because a buggy and slow, inefficiant process is the right way to create and
> effective benchmark. Way to go.

No. Benchmarks should be written naturally.

> Also, you suck at the quotes.

They're called USENET quotes.

[FuzzyDonkey]
> Efficient code isn't what matters. Calibration is what matters. The code only needs to
> be efficient enough so that you spend the majority of the time doing what you think
> the benchmark is supposed to measure.

Right.

[:LJ:]
> And STL : Have you ever considered that there may be a *reason* you're getting
> rejected from these places?

People are idiots.

> If you opened your mind

That doesn't even mean anything. All languages and all approaches are not created equal.

> If your project manager says "Do this in Java/VB/Python/whatever"

I refuse to work for any company that wants me to write Java or Visual Basic. But that's okay. I also refuse to work for any company that wants me to lift heavy boxes over my head in hot trailers.
 
Originally posted by turbowrx
As far as a score for the benchmark, my thought right now is to have it just be a time. You'll have an overall time and a time for each test, probably ten. A 3d test would be in fps, which actually screws things up a little. Should the test output a score when you run all the tests or should it output time, and not count the 3d test, if there is one. It could just output the fps as a sidebar. The users would also be allowed to run as many tests as they want.
Click to expand...

I can see this in two ways. One, individual times in most cases I can think of would work well. But two, an overall time wouldn't mean muchof anything. Even if you do decide to go this route and base it entirely on time, maybe you should make a way to score it as well, where the user could see both. This could make the benchmark accessible to a wider range of users.

If this benchmark works out well, it may be something that other people may want to add on to later. If you release it, will it be an open source license?

I also had another thought about what you could do interface-wise. You could make something like "profiles," some of which come with the proggie and some of which could be customized by the user. The profiles could be sets of tests to run. For example, there could be a "Compression" profile that runs all of the audio and video tests, or a "3D Graphics" profile that runs those tests. This idea would only truly be useful if more tests are added over time, however, so these ideas kind of go together I guess.

Just a few thoughts to bounce off of you. Keep us updated. I don't know why, but I just love benchmarks.
 
The open source idea is something i've pondered. It would make for a better benchmark in the long run and it could become quite popular I suppose.

Your profile idea is something that could be done at a later date when there are a wider range of tests.

I'll keep everyone updated on what's going on with this, but it is going to move slow if I'm the only one working on it. I have too much to do that is higher priority. I'll do my best. Maybe I should setup a website with a forum to talk about it. Depends on the interest I see in the project. Anyway, keep the ideas flowing as I want to develop what you, the users, want to use.
 
Originally posted by turbowrx
The open source idea is something i've pondered. It would make for a better benchmark in the long run and it could become quite popular I suppose.
Click to expand...

It would, but then the main function of the program would be the interface setup, moreso than the actual benchies. I'm not sure that that's what you wanted to do. Of course, it could work out anyway, making it easier for you to add/change the benchies on your own as well.

Originally posted by turbowrx
Your profile idea is something that could be done at a later date when there are a wider range of tests.
Click to expand...

I know, but depending on how you design this, the "profile" thing could make the initial design different. I'm not trying to be pushy, now. I just like your idea, and am trying to offer my 2 cent suggestions.

Originally posted by turbowrx
I'll keep everyone updated on what's going on with this, but it is going to move slow if I'm the only one working on it. I have too much to do that is higher priority. I'll do my best. Maybe I should setup a website with a forum to talk about it. Depends on the interest I see in the project. Anyway, keep the ideas flowing as I want to develop what you, the users, want to use.
Click to expand...

Yeah, see what I said above regarding my pushiness. :D Don't feel compelled to use my ideas, either. They may be lousy, as I am more academic than "real world" regarding this sort of application. I can see it filling a pretty large niche, though.
 
Originally posted by STL
[:LJ:]
> And STL : Have you ever considered that there may be a *reason* you're getting
> rejected from these places?

People are idiots.

> If you opened your mind

That doesn't even mean anything. All languages and all approaches are not created equal.

> If your project manager says "Do this in Java/VB/Python/whatever"

I refuse to work for any company that wants me to write Java or Visual Basic. But that's okay. I also refuse to work for any company that wants me to lift heavy boxes over my head in hot trailers.
Click to expand...

That's fair enough - I just hope that you're prepared to spend resonably large chunks of your life unemployed for the sake of your principles (that's not meant as a dig - just a reality that several of my friends who are equally talented and equally strong of opinion have had to face). For my part, I'm willing to prostitute my convictions for the sake of supporting my family and my real hobby, which isn't anything to do with computers or programming.
 
[:LJ:]
> That's fair enough - I just hope that you're prepared to spend
> resonably large chunks of your life unemployed for the sake of your
> principles

Are you saying that if you aren't willing to program in Java and Visual Basic, you can't get a job? Ha, ha, ha.
 
Originally posted by STL
[:LJ:]
> That's fair enough - I just hope that you're prepared to spend
> resonably large chunks of your life unemployed for the sake of your
> principles

Are you saying that if you aren't willing to program in Java and Visual Basic, you can't get a job? Ha, ha, ha.
Click to expand...

Why must you always take things literally in order to get a laugh? (it's only you that's laughing, by the way)

I'm saying that if you're so inflexible or limited that you won't/can't use anything other than C++, you're going to struggle to get consistent employment, since you won't be able to decide the development platform when you get out of college and start life as a low-level grunt, as we all have. You have to earn the right to do that.
 
[:LJ:]
> I'm saying that if you're so inflexible or limited that you
> won't/can't use anything other than C++

I'm open to learning other tools. For example, if I need to learn Python in order to do something, or Perl, I can pick up as much of that as I need.

I already know more Scheme than probably anyone else on the forums.

But I refuse to write Java or Visual Basic (or COBOL, for that matter).

> you're going to struggle to get consistent employment

Don't you worry your pretty little head about my future.
 
You suck at life.

Originally posted by STL
People are idiots.
...
That doesn't even mean anything. All languages and all approaches are not created equal.
....
I refuse to work for any company that wants me to write Java or Visual Basic. But that's okay. I also refuse to work for any company that wants me to lift heavy boxes over my head in hot trailers.
Click to expand...
 
Guys, I realize that I was the first one dumb enough to try, but didn't I already prove that its pointless to talk sense to a kid who thinks he knows everything?

This started off as a really good post.

i would like to know what you, the user, wants to see in a benchmark? What tests you would like to see and what kind of user interface?
Click to expand...

It would be nice for someone to post a few ideas besides me. The poor guy will stuck in the position of fielding the validity of my views instead of taking the best ideas or combinations.

How about some good ideas for turbowrx?

*Edited for grammar boo-boo*
 
[Time2mod]

> It would be nice for someone to post a few ideas besides me.

I did, and considering the activity of the thread - other people read it and had the oppoturnity to. If turbowrx isn't content with the responses he got he can try again and be more explicit about what he wants.
 
why do you too bother with the retarded quote thats more work and harder to read?

Originally posted by jblair
[Time2mod]

> It would be nice for someone to post a few ideas besides me.

I did, and considering the activity of the thread - other people read it and had the oppoturnity to. If turbowrx isn't content with the responses he got he can try again and be more explicit about what he wants.
Click to expand...
 
Originally posted by Whatsisname
why do you too bother with the retarded quote thats more work and harder to read?
Click to expand...

Probably because it's less work in alot of cases, and it's not harder to read.
 
Originally posted by jblair
I did, and considering the activity of the thread - other people read it and had the oppoturnity to. If turbowrx isn't content with the responses he got he can try again and be more explicit about what he wants.
Click to expand...

I know you did, and a few others too. I was exaggerating for emphasis. I'm afraid the thread got off kilter, and I feel partly responsible. That's all. No offense meant.
 
[Whatsisname]

> why do you too [sic] bother with the ... quote thats [sic] more work

There is a very good reason for putting the text you quote above your response and for quoting only what is necessary for your response.

A: Top posters.
Q: What is the most annoying thing on Usenet?

The reason why putting responses under quotations is better is due to the discussion process. You read something and then you respond to it. Especially when responses are long and often filled with quotes from other sources outside the discussion, it's important to know up front what is being responded to.

Quoting only what is necessary also follows the discussion process. One person brings up a point, and the other person confirms whether the point is accepted or needs to be futher discussed. This allows you to keep posts short, easy to read and people know exactly what's confusing you.

As for the USENET styles with the > symbol (and the name of who you are responding to up top.) It's simpler to simply click quote and hammer out a response, but the [QUOTE] tags are a huge clutter on the textarea where you type the post (along with the [B] tags aswell)

Through quoting only what is necessary and responding to one point at a time you often have to quote multiple times within the same post. Using UBB's quoting functionality means that you have horrendously big horizontal rules and redundancy (as you have mention who said what for every quote in order to get it to look consistent.)

Finally, quoting quotes is often needed and to do that with UBB tags I would have to write:
[QUOTE]Person A came out of the closet to say:
[QUOTE]Person B came out of the closet to say:
One of these days I'll throw you in the Total Perspective Vortex
[/QUOTE]
What the heck?
[/QUOTE] You must run out and purchase 42 copies of the Hitchhiker's Trilogy right now!

And then there's what it'll look like when posted:
Person A came out of the closet to say:
Person B came out of the closet to say:
One of these days I'll throw you in the Total Perspective Vortex
Click to expand...
What the heck?
Click to expand...
You must run out and purchase 42 copies of the Hitchhiker's Trilogy right now!

But with USENET styles, all I have to do is this:
[Person A]
> [Person B]
> > One of these days I'll throw you in the Total Perspective Vortex
> What the heck?

You must run out and purchase 42 copies of the Hitchhiker's Trilogy right now!

By using > symbols, you don't waste screen space with unecessary horizontal rules and by putting the name of the person you are responding to once, you don't waste space/time saying "Whatsisname said this" along with every quote.

[FuzzyDonkey]

> Probably because it's less work in alot of cases

When quoting multiple times across many people, it's a real timesaver.

> and it's not harder to read.

It takes the same amount of thinking to recognise what you've said no matter what surrounds it. Unless I shove it in a vault and charge $50 just to look inside. :p
 
My new quoting style goes,

[Mlar Johnson]
> Oink

[Superfly Myers]
> Meow

I don't like nested > > > > >.
 
Only quoting the part you are going to discuss is a valid reason, but you can just chop out what you dont need from the quote vBull provides you with. Worrying about nested quotes and shit is hardly a reason to spend alot of time copying and pasting manually, when most often you merely have to quote their entire post. Two other things though

1) This is not usenet. If we were there, you'd have a point. Not to mention theres no other way on usenet because they only have text in their messages, not pretty html that we can afford here.

2) I top quote intentionally so you read my message, read the quote, then read my message once again. It's a control thing.

Originally posted by jblair
[Whatsisname]

> why do you too [sic] bother with the ... quote thats [sic] more work

There is a very good reason for putting the text you quote above your response and for quoting only what is necessary for your response.

A: Top posters.
Q: What is the most annoying thing on Usenet?

The reason why putting responses under quotations is better is due to the discussion process. You read something and then you respond to it. Especially when responses are long and often filled with quotes from other sources outside the discussion, it's important to know up front what is being responded to.

Quoting only what is necessary also follows the discussion process. One person brings up a point, and the other person confirms whether the point is accepted or needs to be futher discussed. This allows you to keep posts short, easy to read and people know exactly what's confusing you.

As for the USENET styles with the > symbol (and the name of who you are responding to up top.) It's simpler to simply click quote and hammer out a response, but the [QUOTE] tags are a huge clutter on the textarea where you type the post (along with the [B] tags aswell)

Through quoting only what is necessary and responding to one point at a time you often have to quote multiple times within the same post. Using UBB's quoting functionality means that you have horrendously big horizontal rules and redundancy (as you have mention who said what for every quote in order to get it to look consistent.)

Finally, quoting quotes is often needed and to do that with UBB tags I would have to write:
[QUOTE]Person A came out of the closet to say:
[QUOTE]Person B came out of the closet to say:
One of these days I'll throw you in the Total Perspective Vortex
[/QUOTE]
What the heck?
[/QUOTE] You must run out and purchase 42 copies of the Hitchhiker's Trilogy right now!

And then there's what it'll look like when posted:

What the heck?
[/QUOTE] You must run out and purchase 42 copies of the Hitchhiker's Trilogy right now!

But with USENET styles, all I have to do is this:
[Person A]
> [Person B]
> > One of these days I'll throw you in the Total Perspective Vortex
> What the heck?

You must run out and purchase 42 copies of the Hitchhiker's Trilogy right now!

By using > symbols, you don't waste screen space with unecessary horizontal rules and by putting the name of the person you are responding to once, you don't waste space/time saying "Whatsisname said this" along with every quote.

[FuzzyDonkey]

> Probably because it's less work in alot of cases

When quoting multiple times across many people, it's a real timesaver.

> and it's not harder to read.

It takes the same amount of thinking to recognise what you've said no matter what surrounds it. Unless I shove it in a vault and charge $50 just to look inside. :p
Click to expand...
 
[Whatsisname]
> Worrying about nested quotes and shit is hardly a reason to spend
> alot of time copying and pasting manually

My quoting style is for ease of reading, not ease of writing.

You know, the Wrath Of Khan Principle and all that.

It should be noted that I copy and paste exactly once, and then edit the rest by hand.

> This is not usenet.

USENET is likely Older Than You, and certainly Smarter Than You.

> If we were there, you'd have a point. Not to mention theres no
> other way on usenet because they only have text in their messages,
> not pretty html that we can afford here.

HTML markup is lost in flattening; USENET quotes aren't.

Given HTML, I'd prefer colored text or bolded text or something, but those are lost in flattening too.

> I top quote intentionally

Top quoting is moronic.
 
[Whatsisname]

> not pretty html that we can afford here.

Actually, the so-called "pretty HTML" is one of the reasons that I mentioned against UBB quoting. This ties in with quoting multiple times, as the "pretty HTML" is excessively large and has unecessary distinction. (The indentation is sufficent distinction, but we also have to have two horizontal rules, and a line saying "Quote by X".)

> Only quoting the part you are going to discuss is a valid reason ...

If you discuss one point at a time and quote minimally for that purpose, then you are going to have to quote multiple times for one response.

When quoting multiple times across one response you not only have to chop up the quote template UBB provides but also manually copy and paste the quote template for the other quotations. The presence of markup (both [QUOTE] and [B]) means that you have to deal not only with where the beginning is, but also where the end is. This complication is completely unecessary.

> you read my message, read the quote, then read my message once again.

How many times do you have to read my post to get my point? Once. Anything more is an insult to comprehension.
 
You must log in or register to reply here.
Back
Top