Anyone's afraid of BlitzMax speed?
-
- Enthusiast
- Posts: 423
- Joined: Fri Apr 25, 2003 5:22 pm
- Contact:
I'm still wondering what you guys do quantify there:
e.g. 2069 m/secs
Who or what ist running at that speed ?
It must be an atomic-driven-car or something like that.
Or perhaps David Hasselhoff with his new red-rescue-rocket ?
e.g. 2069 m/secs
Who or what ist running at that speed ?
It must be an atomic-driven-car or something like that.
Or perhaps David Hasselhoff with his new red-rescue-rocket ?
Last edited by Froggerprogger on Tue Mar 21, 2006 10:11 pm, edited 1 time in total.
%1>>1+1*1/1-1!1|1&1<<$1=1
-
- Addict
- Posts: 1126
- Joined: Wed Oct 15, 2003 12:40 am
- Location: Sweden
- Contact:
Froggerprogger wrote:I'm still wondering what you guys do quantify there:
e.g. 2069 m/secs
Who or what ist running at that speed ?
It must be an atomic-driven-car or something like that.
Or perhaps David Hasselhoff with his new red-rescue-rocket ?


(\__/)
(='.'=) This is Bunny. Copy and paste Bunny into your
(")_(") signature to help him gain world domination.
he meant that m/secs would be what we call - eine Maßeienheit für Geschwindigkeiten- and it would have to be named ms for milliseconds
Visit www.sceneproject.org
dagcrack wrote:Of course mines bigger.

needed that!
( PB6.00 LTS Win11 x64 Asrock AB350 Pro4 Ryzen 5 3600 32GB GTX1060 6GB)
( The path to enlightenment and the PureBasic Survival Guide right here... )
( The path to enlightenment and the PureBasic Survival Guide right here... )
I added a new benchmark to the testfile. It's the Dhrystone benchmark and it's ported from Pascal to BMX and PB by Agent Smith from the codersworkshop forum. This time, PB loses BIG time:
PureDhry - V4.00beta7 - Size: 13KB
1000000000 runs
Microseconds for one run through Dhrystone: 1.154687
Dgrystones per Second: 866035.562500
DhryMax - V1.12demo - Size: 81.5KB
1000000000 runs
Microseconds for one run through Dhrystone: 0.498836994
Dgrystones per Second: 2004662.88
Dhrystone relies heavy on Int and String operations. And here, BMX seems to be more than 2 times faster than PB O.o Anyone can explain what's going on???
PureDhry - V4.00beta7 - Size: 13KB
1000000000 runs
Microseconds for one run through Dhrystone: 1.154687
Dgrystones per Second: 866035.562500
DhryMax - V1.12demo - Size: 81.5KB
1000000000 runs
Microseconds for one run through Dhrystone: 0.498836994
Dgrystones per Second: 2004662.88
Dhrystone relies heavy on Int and String operations. And here, BMX seems to be more than 2 times faster than PB O.o Anyone can explain what's going on???
1) I wouldnt move to one of his products just because of a bunch of ms in difference between a few tests (which are no real life tests, as someone suggested). Plus those benchmarks are all wrong... But even if BMX can work twice as fast doing a simple task such as converting an integer into a string or something like that. I wonder how fast it would work in a real-life environment...
2) Fred & co will beat bmx up whenever they feel like. That's having the power! - They can have Mark as their pet whenever they want, I bet you
.
Even though I hate Fred a little (wait, I think I hate everyone in this world a little) for not taking into account math routines in PB - (even today they are lacking, and yes, they are a must if you ask me).. Mark got nothing that would beat PB's stability, well-thought design and most of all, the support and community (although you guys should quit that thing with David already, let go!).
Heck, that guy should think about selling oranges or something already...
Come on Mark, I would buy you a kilogram or two, if they taste good. However I hope you don't bloat the oranges or try to sell me painted rocks.
I feel that Mark has always been living it the easy way, taking shortcuts for everything and making it as simple as he could, still, of course.. getting all the credit for it. That's something I cant remotely stand.
Indeed I have something Against Mark "SIBLINGS".... And I can't stand him taking the credits for something he didnt really do.. I mean theres more work in PB than in BMX in my opinion... Yes I've tryed BMX demo and I think the open source libraries idea is good, just like if you were working with your own custom libs in Cpp. But in no way it feels as good as PB does.
Plus, if you've been working with PB for a long time, I bet you a 6pack of cold beers that you already have your own Libraries that you include in your projects
!
About OGL support... At least on this PC I see PB rendering 100.000s of textured triangles at really nice speeds
- Forgive me, but I will have to hurry up now, so I can prove to the BMX kiddies how powerful PB is then, and, of course.. sell them a library or two.

Mark is really lame, He took weeks / months to fix a few minor bugs on B3D.. And He couldnt port it to DX8 / DX9 because he had virtually hard-coded DX7 routines inside B3D... GO LAUGH!!!.
2) Fred & co will beat bmx up whenever they feel like. That's having the power! - They can have Mark as their pet whenever they want, I bet you

Even though I hate Fred a little (wait, I think I hate everyone in this world a little) for not taking into account math routines in PB - (even today they are lacking, and yes, they are a must if you ask me).. Mark got nothing that would beat PB's stability, well-thought design and most of all, the support and community (although you guys should quit that thing with David already, let go!).
Heck, that guy should think about selling oranges or something already...
Come on Mark, I would buy you a kilogram or two, if they taste good. However I hope you don't bloat the oranges or try to sell me painted rocks.
I feel that Mark has always been living it the easy way, taking shortcuts for everything and making it as simple as he could, still, of course.. getting all the credit for it. That's something I cant remotely stand.
Indeed I have something Against Mark "SIBLINGS".... And I can't stand him taking the credits for something he didnt really do.. I mean theres more work in PB than in BMX in my opinion... Yes I've tryed BMX demo and I think the open source libraries idea is good, just like if you were working with your own custom libs in Cpp. But in no way it feels as good as PB does.
Plus, if you've been working with PB for a long time, I bet you a 6pack of cold beers that you already have your own Libraries that you include in your projects

About OGL support... At least on this PC I see PB rendering 100.000s of textured triangles at really nice speeds


Mark is really lame, He took weeks / months to fix a few minor bugs on B3D.. And He couldnt port it to DX8 / DX9 because he had virtually hard-coded DX7 routines inside B3D... GO LAUGH!!!.
Seems like BMX kiddies has waaaay too much time on their hands, lets give them something to do, here, read: (if you can read, of course.. Else, ask your parent).
1) In order to correctly compare 2 languages, you must run a benchmark per routine this languages share.
2) There would be 10 tests per routine comparison on each language
3) You will then take the average of the results on each language.
4) Finally, after you've drunk 9 cokes (like if it had alcohol..., jeez.. kids!) And spent several days on it, Make a sum of the results on each language.
Its obvious, the smaller ending number would be the winner.
Now, you must make it all in code. So each benchmark would have 10 iterations, plus, the fixed constant iterations of the benchmark itself.
Of course that running a benchmark like this would take virtually a day to accomplish if all functions are benchmarked with lots of iterations.
But then, dont you dare to compare results with intel against amd results. We dont care how fast your processor is (or how small your thingie is, in that case) Because what matters here, are the raw numbers. But, Intel and AMD (which are, the most common processors you'd be using) Are very, very different. So many operations are slower on Intel, and some others might be faster on Intel. Just take that into account. (it also doesnt matter much, since you want to know which language is faster on your system...) But then again, it must run fast on any system. Jerky!
1) In order to correctly compare 2 languages, you must run a benchmark per routine this languages share.
2) There would be 10 tests per routine comparison on each language
3) You will then take the average of the results on each language.
4) Finally, after you've drunk 9 cokes (like if it had alcohol..., jeez.. kids!) And spent several days on it, Make a sum of the results on each language.
Its obvious, the smaller ending number would be the winner.
Now, you must make it all in code. So each benchmark would have 10 iterations, plus, the fixed constant iterations of the benchmark itself.
Of course that running a benchmark like this would take virtually a day to accomplish if all functions are benchmarked with lots of iterations.
But then, dont you dare to compare results with intel against amd results. We dont care how fast your processor is (or how small your thingie is, in that case) Because what matters here, are the raw numbers. But, Intel and AMD (which are, the most common processors you'd be using) Are very, very different. So many operations are slower on Intel, and some others might be faster on Intel. Just take that into account. (it also doesnt matter much, since you want to know which language is faster on your system...) But then again, it must run fast on any system. Jerky!
Not one second have I thought about move to BMX! All I want to do is to prove that PB beats the hell out of BMX.1) I wouldnt move to one of his products just because of a bunch of ms in difference between a few tests (which are no real life tests, as someone suggested). Plus those benchmarks are all wrong... But even if BMX can work twice as fast doing a simple task such as converting an integer into a string or something like that. I wonder how fast it would work in a real-life environment...
I want them to do (just a little bit so all users of PB can have a lough whenever they hear "Blitz....something")!2) Fred & co will beat bmx up whenever they feel like. That's having the power! - They can have Mark as their pet whenever they want, I bet you
DON'T CALL ME A BMX KIDDIE! I sold my BMX license long time ago because I have full trust in Fred and the team! For me, PB is the best language around. But what is false if I want to compare the speed of BMX and PB? I just want to know _why_ BMX is faster and it would be nice to know if Fred is still working on increasing the speed of PB.Seems like BMX kiddies has waaaay too much time on their hands, lets give them something to do, here, read: (if you can read, of course.. Else, ask your parent).
Not you dude, people on coderscockshop ..
I'm duneinc at there in case you wanted to know! but I dont browse the forums at all. I often go help some people with b3d questions (or common game design).
Sure blitz-something will always be less than PB!. And for games as you know (or may not know) I've spent an year developing a real killer... For several reasons it was on a "halt" (Number one reason: investment, Number two: personal life).
I now say that in about 6 months or less I will be posting compatibility tests and a feature show off.
I'm saying it now because I think I'm doing a good job with this. I'm not porting anyones libraries nor stealing code from others like people has been doing for the eternity of time
(And even worst, charging for their "WRAPPERS"). I've developed everything from the ground, optimized as much as time would let me optimize, and designed every system and library to be developer-friendly and as I said, very well thought and fast.
This are no promises though
I'm no politician
Time will tell. So far every single function beats B3D's, that was the main goal. Giving developers the ease of use that b3d could give them, but, with the power that only the most expensive engines are able to provide them. This, at a fraction of cost.
No one can beat this
- Hence I'm saying it now.
This is no cold steel, It's hot, very hot. So dont touch it , or it will burn your hand.
Theres nothing worst than coldsteal!, ehm, sorry, steel. Or would you like to sleep around cold steel?.
I'm duneinc at there in case you wanted to know! but I dont browse the forums at all. I often go help some people with b3d questions (or common game design).
Sure blitz-something will always be less than PB!. And for games as you know (or may not know) I've spent an year developing a real killer... For several reasons it was on a "halt" (Number one reason: investment, Number two: personal life).
I now say that in about 6 months or less I will be posting compatibility tests and a feature show off.
I'm saying it now because I think I'm doing a good job with this. I'm not porting anyones libraries nor stealing code from others like people has been doing for the eternity of time

This are no promises though

I'm no politician

Time will tell. So far every single function beats B3D's, that was the main goal. Giving developers the ease of use that b3d could give them, but, with the power that only the most expensive engines are able to provide them. This, at a fraction of cost.
No one can beat this

This is no cold steel, It's hot, very hot. So dont touch it , or it will burn your hand.
Theres nothing worst than coldsteal!, ehm, sorry, steel. Or would you like to sleep around cold steel?.
I tried to leave that one uncommented because it's really OT but let me just say this: ColdSteel (when it will be 1.0) will be more than a wrapper for Irrlicht. It is already more than a wrapper because it's even easier to use than Irrlicht. I understand that you have to goof on ColdSteel because you did your own engine. But you should keep things professional. I never saw Fred goofing on Blitz products (that's our jobTheres nothing worst than coldsteal!, ehm, sorry, steel.

Well, back2topic shall we

P.S.:
Good to know. CW is so full of Cobra and Blitz fanboys - I'm fighting a war over thereNot you dude, people on coderscockshop ..

I don't need to keep things "professional" with this people.
Got my own market, my own customers and many years on this, which, this people clearly doesnt have. Jedive, the guy from spain.. I know him, hes just a wrap-a-holic. Not a real developer.
Since 6 months this engine is usable. But why would I come tell you that?, theres no need.
If I was going to wrap someones work, I would already have it done.
Got my own market, my own customers and many years on this, which, this people clearly doesnt have. Jedive, the guy from spain.. I know him, hes just a wrap-a-holic. Not a real developer.
Since 6 months this engine is usable. But why would I come tell you that?, theres no need.
If I was going to wrap someones work, I would already have it done.