Discussion on:
View:
Show:
Its true compiler languages have come a long way since the days of GeeWizz Basic but PowerBasic is just one of a few note worthy compilers. My personal preference is HotBasic which in my opinion is leagues ahead of most competitors. Why? because It takes Basic style syntax and converts it to ASM Mnemonics then links it. The result is like having written your application in assembler. IMHO, nothing results in smaller or faster executable code.
I use Powerbasic for business software.
It's perfect under every point of view (easy, fast, reliable), but the more remarkable feature for my purposes is that there is no need for installation; no runtime, no dependencies, no registry entry ... only stand-alone executables.
One of my customers add a client to its intranet? All what's he need is to create a desktop icon pointing to my main application on the server and my software is running ... a lot of time saved!
It's perfect under every point of view (easy, fast, reliable), but the more remarkable feature for my purposes is that there is no need for installation; no runtime, no dependencies, no registry entry ... only stand-alone executables.
One of my customers add a client to its intranet? All what's he need is to create a desktop icon pointing to my main application on the server and my software is running ... a lot of time saved!
Hi Mike--
It would be good to note that the general algorithm you describe applies to all native code compilers. They convert source code to machine code which can then be executed. Some do it as a multi-step process... Compile Basic source to ASM source, then compile the assembler source to object code, then finally link the object code to the final .EXE file. Often, this type of process involves the use of additional external programs like an assembler and a linker, so the Basic compiler has much less to do.
Other native code compilers, like PowerBASIC, perform the entire process in a single step. Basic source is translated to machine code directly, at up to 20,000,000 lines of source per minute. That means a large application will usually compile in just a fraction of a second. That can make a big difference during debug sessions that we all have to endure. {smile}
One other quick point... I think that direct comparison of the speed of executable code generated by PowerBASIC to any other Basic will prove to be a real eye-opener for you. Check it out!
Best regards,
Bob Zale
PowerBASIC Inc.
It would be good to note that the general algorithm you describe applies to all native code compilers. They convert source code to machine code which can then be executed. Some do it as a multi-step process... Compile Basic source to ASM source, then compile the assembler source to object code, then finally link the object code to the final .EXE file. Often, this type of process involves the use of additional external programs like an assembler and a linker, so the Basic compiler has much less to do.
Other native code compilers, like PowerBASIC, perform the entire process in a single step. Basic source is translated to machine code directly, at up to 20,000,000 lines of source per minute. That means a large application will usually compile in just a fraction of a second. That can make a big difference during debug sessions that we all have to endure. {smile}
One other quick point... I think that direct comparison of the speed of executable code generated by PowerBASIC to any other Basic will prove to be a real eye-opener for you. Check it out!
Best regards,
Bob Zale
PowerBASIC Inc.
Everyone is entitled to their opinions which carry more weight when justified by real world facts. So here is my challenge to anyone with an open mind. Visit the hotbasic.org web site. Try the free CMD version of HotBasic and the free online compile of the Windows version. If you don't find it easier to use, requiring less declarations, faster, producing smaller executables, stick with what you know.
For the record I have no vested or economic interest in HotBasic other than using it to create my IM client/server, PopNote.
TallyHo,
Mike
For the record I have no vested or economic interest in HotBasic other than using it to create my IM client/server, PopNote.
TallyHo,
Mike
Hi Mike--
I find your exuberance kind of exciting... so I had to give it a try! You said nothing is faster than HotBasic? What a challenge! Here's a simple program in HotBasic:
$APPTYPE CONSOLE
DIM u AS DOUBLE
DIM i AS DOUBLE
DIM a AS DOUBLE
DIM b AS DOUBLE
DIM c AS DOUBLE
u=TIMER
FOR i=1 TO 1000000000
b=i/2
a=i+1
c=a*(i MOD b)
NEXT i
PRINT STR$(TIMER-u)
pause
END
On a slow machine, it took 84.3 seconds. Not bad, I guess. So, how about the equivalent PowerBASIC program?
FUNCTION PBMAIN ()
DIM u AS DOUBLE
DIM i AS DOUBLE
DIM a AS DOUBLE
DIM b AS DOUBLE
DIM c AS DOUBLE
u=TIMER
FOR i=1 TO 1000000000
b=i/2
a=i+1
c=a*(i MOD b)
NEXT i
PRINT STR$(TIMER-u)
END FUNCTION
Same computer-- it took just 15.25 seconds with PowerBASIC. I guess that means PowerBASIC was 553% faster. What do you think?
How about this program?
$APPTYPE CONSOLE
defdbl x, y, i
defsng t
x = 1
y = 1.000001
t = TIMER
FOR i = 1 TO 100000000
x = x * y
NEXT
t = TIMER - t
print STR$(t)
PAUSE
END
Same slow machine -- 8.72 seconds. Not bad.
The equivalent code in PowerBASIC?
FUNCTION PBMAIN () AS LONG
x## = 1
y## = 1.000001
t! = TIMER
FOR i& = 1 TO 100000000
x## = x## * y##
NEXT
t! = TIMER - t!
PRINT STR$(t!)
END FUNCTION
Note the PowerBASIC code was calculated using extended precision variables, considerably more complex than just doubles. What was the elapsed time? 0.17 seconds. That's 5129% faster than HotBasic. It might be a good idea to consider alternative compilers.
Thanks for contributing here!
Bob Zale
PowerBASIC Inc.
I find your exuberance kind of exciting... so I had to give it a try! You said nothing is faster than HotBasic? What a challenge! Here's a simple program in HotBasic:
$APPTYPE CONSOLE
DIM u AS DOUBLE
DIM i AS DOUBLE
DIM a AS DOUBLE
DIM b AS DOUBLE
DIM c AS DOUBLE
u=TIMER
FOR i=1 TO 1000000000
b=i/2
a=i+1
c=a*(i MOD b)
NEXT i
PRINT STR$(TIMER-u)
pause
END
On a slow machine, it took 84.3 seconds. Not bad, I guess. So, how about the equivalent PowerBASIC program?
FUNCTION PBMAIN ()
DIM u AS DOUBLE
DIM i AS DOUBLE
DIM a AS DOUBLE
DIM b AS DOUBLE
DIM c AS DOUBLE
u=TIMER
FOR i=1 TO 1000000000
b=i/2
a=i+1
c=a*(i MOD b)
NEXT i
PRINT STR$(TIMER-u)
END FUNCTION
Same computer-- it took just 15.25 seconds with PowerBASIC. I guess that means PowerBASIC was 553% faster. What do you think?
How about this program?
$APPTYPE CONSOLE
defdbl x, y, i
defsng t
x = 1
y = 1.000001
t = TIMER
FOR i = 1 TO 100000000
x = x * y
NEXT
t = TIMER - t
print STR$(t)
PAUSE
END
Same slow machine -- 8.72 seconds. Not bad.
The equivalent code in PowerBASIC?
FUNCTION PBMAIN () AS LONG
x## = 1
y## = 1.000001
t! = TIMER
FOR i& = 1 TO 100000000
x## = x## * y##
NEXT
t! = TIMER - t!
PRINT STR$(t!)
END FUNCTION
Note the PowerBASIC code was calculated using extended precision variables, considerably more complex than just doubles. What was the elapsed time? 0.17 seconds. That's 5129% faster than HotBasic. It might be a good idea to consider alternative compilers.
Thanks for contributing here!
Bob Zale
PowerBASIC Inc.
Per notification from mike.k4hum, I might add several comments:
1. We find Mr. Zale saying "5129% faster than HotBasic", which, of course, is physically impossible.
Other relevant points may be mentioned. In his PB version, i& is INTEGER, but he uses DOUBLE in the HB version, which is known to be slower. In general, he can cherry-pick whatever examples he wants, but the public record shows PB and HB are close in speed, as we posted for years (see www.hotbasic.org/hot/hotspeed.html). Also notice in the first example, Mr. Zale did not use HB's integer math which would have been faster. My impression is that Mr. Zale is not credible in these posts and PB users are too smart to be fooled by bogus speed data.
2. Everybody knows that migration between PB and HB has been 100% unidirectional -- namely from PB to HB, so PB users do know the truth and choose HB.
3. What everybody may not know is that Mr. Zale has repeatedly visited the HB site over the years, which suggests that he knows who the leader is.
4. I know of nobody who thinks that PB will ever catch up to HB.
Cordially, Maj. Hog, Spokespig, www.hotbasic.org
1. We find Mr. Zale saying "5129% faster than HotBasic", which, of course, is physically impossible.
Other relevant points may be mentioned. In his PB version, i& is INTEGER, but he uses DOUBLE in the HB version, which is known to be slower. In general, he can cherry-pick whatever examples he wants, but the public record shows PB and HB are close in speed, as we posted for years (see www.hotbasic.org/hot/hotspeed.html). Also notice in the first example, Mr. Zale did not use HB's integer math which would have been faster. My impression is that Mr. Zale is not credible in these posts and PB users are too smart to be fooled by bogus speed data.
2. Everybody knows that migration between PB and HB has been 100% unidirectional -- namely from PB to HB, so PB users do know the truth and choose HB.
3. What everybody may not know is that Mr. Zale has repeatedly visited the HB site over the years, which suggests that he knows who the leader is.
4. I know of nobody who thinks that PB will ever catch up to HB.
Cordially, Maj. Hog, Spokespig, www.hotbasic.org
I've a backround of more than 25 years of software development using Mainframe and Windows OS . In my opinion the fast, lean but mighty PowerBasic environment in combination with the (free) support of all the experts, who share their knowledge in the unique PowerBasic Forums is hard to beat in the industry concerning software development productivity. Try it!
100% true. I used a lot of compilers and a lot of languages during the last decades. When it comes to Windows programming, PowerBasic is always my first choice. For me 2 things are crucial: understanding my own code a year or two later and absolutely stable executables. In C I have to write tons of comments to remember what I have been doing. In PowerBasic I am happy with a line of comment here and there. And I have never been able to crash an EXE compiled with Powerbasic with perhaps only one exception: accessing arrays beyond its boundaries will result in a well managed shutdown rather than a crash. I you avoid this, you get unparalleld stable code. I really hope PowerBasic will be around for the rest of my lifetime.
I'm a freelance developer and software author. I use PowerBASIC, which I consider a real craftsman's tool for the current Windows platforms. Excellent support, both from PowerBASIC Inc and the user forums which it hosts. Knowing that the support is there lends itself to a fearless approach to development! The native IDE is very simple and bling-free - Eclipse it ain't. The compiler is very fast and has very few bugs. If you are evaluating it, spend an hour or two looking at the thousands of source code listings published in the forums, join the forums, ask questions...
PB gives me the best balance of productivity and executable speed of any enviromment I've used. Re deployment it runs and deploys on everything from win95 to win7(64) out of the box. It has built-in high level features that don't compromise on performance but also lets you get down into asm and win32API.
I rely on PowerBasic to handle complex system and platform integration problems. I develop web applications for operational use within my company. But these applications cover such a wide range of technologies that often I need to go to PowerBasic to tie everything together. PB and its user community offers great power that is immediately accessible.
I've been using Powerbasic for many years. With the help of of the Powerbasic forums, I taught myself how to create in-house utilities for my company and applications that are used by federal, state, and local agencies. These applications help people understand the impacts that fresh water needs have on aquatic habitat. I don't profess to know everything about the Windows operating system, but then I don't need to, because with Powerbasic and other third party tools like EZGUI, the learning curve wasn't so steep. This is a serious platform made for a wide variety of skill levels.
I have been using PowerBASIC since its very early days, back when it was called TurboBASIC. I started with their DOS version. Loved it. Some people are still using apps I wrote several years ago with PB-DOS.
The Windows version of PB is a force! I will not use much of anything else. PB is my first choice for everything I do now.
Don't take my view. Give it a try. You WILL be amazed!
The Windows version of PB is a force! I will not use much of anything else. PB is my first choice for everything I do now.
Don't take my view. Give it a try. You WILL be amazed!
I have programmed in Fortran, Pascal, MFC, Java, C, C++, Masm32... etc..
none of them touch PowerBASIC in terms of ease of use, speed & power....
A few years back, I got a call from someone asking me to help them build some high speed graphic routines... after hearing what they wanted... I replied that is easy and I started to explain what to do... they asked if I would send them a demo... So I asked who they were... and they said they were a satellite facility of Virginia Tech... So I asked them why don't they have the Computer Sciences Dept help them ?... they replied that they were the CS dept... and actually the CS Phd dept... and that they only know C++ .. and they can't do what they want easily with C++... hehe... I have never looked back....
further, PowerBASIC, along with Masm32 are the most knowledgeable and helpful programming forums on the internet...
none of them touch PowerBASIC in terms of ease of use, speed & power....
A few years back, I got a call from someone asking me to help them build some high speed graphic routines... after hearing what they wanted... I replied that is easy and I started to explain what to do... they asked if I would send them a demo... So I asked who they were... and they said they were a satellite facility of Virginia Tech... So I asked them why don't they have the Computer Sciences Dept help them ?... they replied that they were the CS dept... and actually the CS Phd dept... and that they only know C++ .. and they can't do what they want easily with C++... hehe... I have never looked back....
further, PowerBASIC, along with Masm32 are the most knowledgeable and helpful programming forums on the internet...
I use PowerBasic in science and engineering work. Unlike with C you needn???t be a code-cracker to read Basic ??? and PowerBasic is just as fast as C. PB compiles your source code to an executable that runs by itself, no extra libraries are required. The Console version of PowerBasic makes it fairly easy to convert old DOS programs so they run directly under Windows.
Nothing against PB. I actually own the latest versions of both PBWin and PBCC and they're great. I just found the interview a little on the light side, coming off as more of an advert than anything else. When will PB produce 64-bit executables and how about Linux and/or OSX support?
BSD Unix support.
Actually, if they just took a POSIX-compliant approach to porting it to FreeBSD, they could probably get NetBSD, OpenBSD, PC-BSD, Dragonfly BSD, and MacOS X (almost) for free. Linux support would just be a tweak or two away.
Actually, if they just took a POSIX-compliant approach to porting it to FreeBSD, they could probably get NetBSD, OpenBSD, PC-BSD, Dragonfly BSD, and MacOS X (almost) for free. Linux support would just be a tweak or two away.
While the language has some interesting and impressive features, the company's poor business practice and ignoring all emails about those practices leave me with a very negative attitude and bank account.

































