That's a very good point. I plan on doing some bench work with the Propeller Activity Board and become more familiar with propeller operations.
I learned to program C while developing the CNC Plasma Cutting Machine and found C to be a very nasty language to learn. I suspect that doing the timing you suggest will require learning SPIN...which doesn't look fun at all.
I could be wrong though.
Discovery
Spin is fun I don't like C too; I can use it when I have to, else I prefer something other. Pascal, Spin, asm...
pik33, I'm curious why the downer of C and a preference for Pascal? Conceptually they are both the same as far as I can tell. Pascal has a more verbose syntax and is perhaps a bit more fussy about type checking. Swings and roundabouts.
Personally I can't get along with any language whose name starts with "P": Perl, Python, Pascal, Prolog...:)
Let's see, languages that I know...Borland Basic, Parallax Basic, FORTRAN, Pascal, and now C. By far the easiest language to use that gets acceptable results very quickly is PBasic. I program a lot of BS2s for a wide variety of applications. Whoever designed the BS2 architeture and developed the PBasic for it really deserves a pat on the back.
Next is FORTRAN, The shear power, speed, and ease of use makes it tops for big engineering design and control applications.
Language comparisons are relative. It is easy to go to the PBasic library of functions, read the explanations with examples, and start coding programs right away.
With C...such was not the case. reading explanations of the commands was confusing, no examples to emulate, and the code (90% of the time) would not run and generated error codes that were nowhere close to the actual error. The real saving grace for C was members of the forum who supplied working programs that included an instruction that I wanted to use...so I was able to copy it. Trying to figure out how to use the instruction from the instruction list was in short futile.
C is not pedigocically nice. (=, ==, &&, !=, etc.) whereas equal, not equal, greater than, +, -, /, etc. are nice to work with. It appears to me that C was designed for a special group of coders who would rather keep the language to themselves much like prescriptions written by doctors to keep instructions contained within their group.
I plan to write a few programs in SPIN and PASM to get a feel for them.
I plan to write a few programs in SPIN and PASM to get a feel for them.
Discovery
Good plan. I think you'll find Spin pretty easy to learn although it has even more obscure operators than C. PASM is quite easy as well. There are a few quirks like needing to use self-modifying code for things like function calls and COG memory indexing but there are lots of people here who can help you with those.
Discovery,
With your background, you will find spin easy. Just makes sure you get indentation correct (turn on the indentation marks if they are off).
But spin will not be as fast as C. It has some of the peculiarities of C like ==, !=, etc. Just print the couple of pages reference and you will mostly be fine.
PASM is not difficult either, as long as you ease your way into it.
If you are happy with C then stick to it. Just separate out some of the functions into other cogs and then they could be replaced with PASM if you wanted to speed it up, or need deterministic code (able to guarantee number of clocks in code).
The prop is a great chip enabling the time dependant parts of the code to be run in separate cogs (cores), and then a main program to bring those cores together in an unconstrained main program. It ultimately makes debugging easier because you can prove each cores code almost independently.
Personally I can't get along with any language whose name starts with "P": Perl, Python, Pascal, Prolog...:)
Now Heater mind your Ps and Qs. At least the first P. Don't go dissin' Perl. Powerful magic in that even on the windows versions. And while some may find C to be user hostile, nothing compares for obfuscation as Perl can. Even had an annual contest running for the most obfuscated Perl possible. I'm kinda language agnostic. whatever lets me get the job done. But Perl was my tool of choice for automating various processes.
But I really like the prop and PASM. looking at all the embedded options available, the prop is the only chip that will let me continue down the Analog path doing capture on multiple parallel channels with no skew in sample time between them. I don't think there is a PIC that can do this, though by now there may be a multicore ARM that may be able to do this. Now if the prop had a higher clock freq, I could get above the 100ksps rate............
pik33, I'm curious why the downer of C and a preference for Pascal? Conceptually they are both the same as far as I can tell. Pascal has a more verbose syntax and is perhaps a bit more fussy about type checking. Swings and roundabouts.
Personally I can't get along with any language whose name starts with "P": Perl, Python, Pascal, Prolog...:)
I don't hate C. I can use it when needed.
I simply learned Pascal first, C later.
The first programming language I learned was Fortran; then Pascal - both of them used on ancient Polish mainframe computer called Odra, using perforated cards The computer worked under the George 3 operating system.
Then I got 8-bit Atari and learned Basic and then 6502 asm... then there was other machines and Z80 asm.. 68000 asm...
Then the university I studied in bought some IBM PC and we had Turbo Pascal on it... and x86 asm.
I learned C and C++several years later and this is main reason why I don't like programming in it. It is like speaking a foreign language.
Then I don't like the philosophy of making the program with several hundreds small files where I cannot find anything.
I have no intention to start a language war here. As far as I'm concerned languages like FORTRAN, Pascal, C, Spin are all the same thing. They are all conceptually 1970's vintage block structured languages with constructs for sequence, selection and iteration and a stack based approach to function calls and variable scope. They are all ALGOL with different bells and whistles added, and different syntax.
To that group we can also throw in many other languages like Ada, PL/M, Coral and so on.
No, I was more interested in that pedagogical thing. Why would C, or Spin be perceived as an up hill struggle compared to other very similar languages?
Discovery, has mentioned some issues he has had:
...reading explanations of the commands was confusing... no examples to emulate...the code would not run...
These are of course not issues with the language itself. Over in the parallel universe of the Arduino people are introduced to programming with C++. An even more complex and twisted language. They seem to get along just fine. How so? A scan of the Arduino site turns up a lot of very nice and simple documentation. Including a very simple intro to programming with C++. Presumably all their examples do "just work". Then there are tons of working libraries for the Arduino. Then the IDE is extremely simple to use.
It would seem C for the Propeller has not reached that level of support yet.
...generated error codes that were nowhere close to the actual error...
That can be a problem. I guess there is not much we can do about the error messages from GCC. But then again, those Aduino guys have the same compiler and the same error messages. The issue of errors being reported in places you don't expect is common to all compiled high level languages. The compiler cannot know your intention it can only know what you wrote. If a type error occurs in an expression at the bottom of your file is the error there? Or is the error that you specified the wrong type for a variable at the top of the file? Or in a header file? How could the compiler know?
Discovery, I'd like to see an example of such an error message that was troublesome.
C is not pedigocically nice. (=, ==, &&, !=, etc.) ....
That's debatable, Pascal has similar but different operators. Spin has similar but different operators. As do the other languages. I guess it depends what you are used to. Pedagogically, for a beginner who has never programmed before, does it make any difference which symbols are used for what operators? (Assuming some connection with previous experience like "+", "-", "/". Even using "*" for multiply is a bit suspect).
Then I don't like the philosophy of making the program with several hundreds small files where I cannot find anything.
You know, that sometimes drives me nuts too.
I guess that's why people love IDE's like VisualStudio. Click on a symbol and boom you can be taken to it's definition or see where else it is used. It gives the illusion that there are no files, just a huge mass of code you can jump around in.
I'm not sure how that relates to C though. I have worked on large projects in Pascal, PL/M, Coral, Ada and others. They all had masses of files. Conversely C does not stop you thowing your entire project into a single file if you want.
Files are of course good for organization and reuse of code. Carried to extreme it's a pain though.
...mind your Ps and Qs. At least the first P. Don't go dissin' Perl
Sadly I can't do any dissin' of Perl. I have never used it. Never even really looked at any.
Of course my point about not wanting to use languages whoes names start with a "P" is that most of the language war arguments one sees are kind of emotional and based on a lot of trivial fluff that ultimately does not make a lot of difference. Well, appart from the fact that I noticed that all the worlds crappy languages have names starting with a "P"
...nothing compares for obfuscation as Perl can...
However I can challeng that statement
The following is a hello world program in JavaScript, you can run it in your browser or under node.js to get the message.
After spending some time writing and running C programs on the Propeller Activity Board, I came to the conclusion that I can use this program. It does require keeping track of the braces, brackets, commas, semicolons, etc. I was reminded recently by my daughter who was talking with a friend on the phone and switched languages several times between English, Russian, and Spanish. She mentioned that some things don't translate in one language but do in another. That is prettly much the same on computer programming. One language is definitely better that another depending on the use. I use InterLisp for expert system development but it could be accomplished in Basic; but, if given the choice...I would choose InterLisp.
The major stumbing block I encounter using BS2s and Propellers is the very strong need for a library of small chunks of code that show a user, such as myself, how to code perriferial devices to the microcontrollers such as DS1302 clock chips. It took a while asking members of the forums for representative code. But after a while, the BS2 microcontroller was able to communicate with devices on the one wire and I2C. Four channel ADCs, DIs/DOs, temperature, humidity, etc.
We are way OT. However, since the OP (original poster) has joined the fray I guess this is OK.
I have found SPIN to be
1. Quite simple to learn and use, but there are a few gotcha's.
2. Very small code footprint due to interpreter
3. Slow, unfortunately due to interpreter
It's not a full-blown language, hence its ease in learning.
I have found PASM to be
1. Quite simple to learn. This is due to regular instruction set.
2. Fast, but limited in size. Mostly not a problem for small drivers and the like. Not useful for the "main" program.
SPIN & PASM go together naturally. There are some old, but good, tutorials for programmers in the PropTool library.
As for the P1/prop/propeller/P8X32A, there are only a few registers that matter when one starts on the prop.
There are 16 of which 3 are not used because of no PORTB. But to start with PAR, CNT, DIRA, INA, OUTA are the only ones to use first up.
In fact, to flash a LED simply, you only require DIRA and OUTA.
This is what makes the prop so simple - no complex registers to set up, and no interrupts to worry about.
I had to check. Never heard of most of them. B was OK, being the for runner of C, but now obsolete. BCPL, BASH and BASIC are to be avoided. So yes you have a point
Now you remind me that I forgot PHP, Postscript, Processing, PL/M from my list of P* languages. Three of which I have had to deal with and found to be horrible, one of which I have only briefly seen to be horrible.
So there we have it, PHP, Postscript, Processing, PL/M, Perl, Python, Pascal, Prolog.
The evidence is overwhelming, there is a black hole in language design around the letter P.
Comments
Spin is fun
pik33, I'm curious why the downer of C and a preference for Pascal? Conceptually they are both the same as far as I can tell. Pascal has a more verbose syntax and is perhaps a bit more fussy about type checking. Swings and roundabouts.
Personally I can't get along with any language whose name starts with "P": Perl, Python, Pascal, Prolog...:)
Next is FORTRAN, The shear power, speed, and ease of use makes it tops for big engineering design and control applications.
Language comparisons are relative. It is easy to go to the PBasic library of functions, read the explanations with examples, and start coding programs right away.
With C...such was not the case. reading explanations of the commands was confusing, no examples to emulate, and the code (90% of the time) would not run and generated error codes that were nowhere close to the actual error. The real saving grace for C was members of the forum who supplied working programs that included an instruction that I wanted to use...so I was able to copy it. Trying to figure out how to use the instruction from the instruction list was in short futile.
C is not pedigocically nice. (=, ==, &&, !=, etc.) whereas equal, not equal, greater than, +, -, /, etc. are nice to work with. It appears to me that C was designed for a special group of coders who would rather keep the language to themselves much like prescriptions written by doctors to keep instructions contained within their group.
I plan to write a few programs in SPIN and PASM to get a feel for them.
Discovery
With your background, you will find spin easy. Just makes sure you get indentation correct (turn on the indentation marks if they are off).
But spin will not be as fast as C. It has some of the peculiarities of C like ==, !=, etc. Just print the couple of pages reference and you will mostly be fine.
PASM is not difficult either, as long as you ease your way into it.
If you are happy with C then stick to it. Just separate out some of the functions into other cogs and then they could be replaced with PASM if you wanted to speed it up, or need deterministic code (able to guarantee number of clocks in code).
The prop is a great chip enabling the time dependant parts of the code to be run in separate cogs (cores), and then a main program to bring those cores together in an unconstrained main program. It ultimately makes debugging easier because you can prove each cores code almost independently.
Now Heater mind your Ps and Qs. At least the first P. Don't go dissin' Perl. Powerful magic in that even on the windows versions. And while some may find C to be user hostile, nothing compares for obfuscation as Perl can. Even had an annual contest running for the most obfuscated Perl possible. I'm kinda language agnostic. whatever lets me get the job done. But Perl was my tool of choice for automating various processes.
But I really like the prop and PASM. looking at all the embedded options available, the prop is the only chip that will let me continue down the Analog path doing capture on multiple parallel channels with no skew in sample time between them. I don't think there is a PIC that can do this, though by now there may be a multicore ARM that may be able to do this. Now if the prop had a higher clock freq, I could get above the 100ksps rate............
I don't hate C. I can use it when needed.
I simply learned Pascal first, C later.
The first programming language I learned was Fortran; then Pascal - both of them used on ancient Polish mainframe computer called Odra, using perforated cards
Then I got 8-bit Atari and learned Basic and then 6502 asm... then there was other machines and Z80 asm.. 68000 asm...
Then the university I studied in bought some IBM PC and we had Turbo Pascal on it... and x86 asm.
I learned C and C++several years later and this is main reason why I don't like programming in it. It is like speaking a foreign language.
Then I don't like the philosophy of making the program with several hundreds small files where I cannot find anything.
To that group we can also throw in many other languages like Ada, PL/M, Coral and so on.
No, I was more interested in that pedagogical thing. Why would C, or Spin be perceived as an up hill struggle compared to other very similar languages?
Discovery, has mentioned some issues he has had: These are of course not issues with the language itself. Over in the parallel universe of the Arduino people are introduced to programming with C++. An even more complex and twisted language. They seem to get along just fine. How so? A scan of the Arduino site turns up a lot of very nice and simple documentation. Including a very simple intro to programming with C++. Presumably all their examples do "just work". Then there are tons of working libraries for the Arduino. Then the IDE is extremely simple to use.
It would seem C for the Propeller has not reached that level of support yet. That can be a problem. I guess there is not much we can do about the error messages from GCC. But then again, those Aduino guys have the same compiler and the same error messages. The issue of errors being reported in places you don't expect is common to all compiled high level languages. The compiler cannot know your intention it can only know what you wrote. If a type error occurs in an expression at the bottom of your file is the error there? Or is the error that you specified the wrong type for a variable at the top of the file? Or in a header file? How could the compiler know?
Discovery, I'd like to see an example of such an error message that was troublesome. That's debatable, Pascal has similar but different operators. Spin has similar but different operators. As do the other languages. I guess it depends what you are used to. Pedagogically, for a beginner who has never programmed before, does it make any difference which symbols are used for what operators? (Assuming some connection with previous experience like "+", "-", "/". Even using "*" for multiply is a bit suspect).
I guess that's why people love IDE's like VisualStudio. Click on a symbol and boom you can be taken to it's definition or see where else it is used. It gives the illusion that there are no files, just a huge mass of code you can jump around in.
I'm not sure how that relates to C though. I have worked on large projects in Pascal, PL/M, Coral, Ada and others. They all had masses of files. Conversely C does not stop you thowing your entire project into a single file if you want.
Files are of course good for organization and reuse of code. Carried to extreme it's a pain though.
Of course my point about not wanting to use languages whoes names start with a "P" is that most of the language war arguments one sees are kind of emotional and based on a lot of trivial fluff that ultimately does not make a lot of difference. Well, appart from the fact that I noticed that all the worlds crappy languages have names starting with a "P"
However I can challeng that statement
The following is a hello world program in JavaScript, you can run it in your browser or under node.js to get the message.
The major stumbing block I encounter using BS2s and Propellers is the very strong need for a library of small chunks of code that show a user, such as myself, how to code perriferial devices to the microcontrollers such as DS1302 clock chips. It took a while asking members of the forums for representative code. But after a while, the BS2 microcontroller was able to communicate with devices on the one wire and I2C. Four channel ADCs, DIs/DOs, temperature, humidity, etc.
Discovery
I have found SPIN to be
1. Quite simple to learn and use, but there are a few gotcha's.
2. Very small code footprint due to interpreter
3. Slow, unfortunately due to interpreter
It's not a full-blown language, hence its ease in learning.
I have found PASM to be
1. Quite simple to learn. This is due to regular instruction set.
2. Fast, but limited in size. Mostly not a problem for small drivers and the like. Not useful for the "main" program.
SPIN & PASM go together naturally. There are some old, but good, tutorials for programmers in the PropTool library.
As for the P1/prop/propeller/P8X32A, there are only a few registers that matter when one starts on the prop.
There are 16 of which 3 are not used because of no PORTB. But to start with PAR, CNT, DIRA, INA, OUTA are the only ones to use first up.
In fact, to flash a LED simply, you only require DIRA and OUTA.
This is what makes the prop so simple - no complex registers to set up, and no interrupts to worry about.
Discovery
I had to check. Never heard of most of them. B was OK, being the for runner of C, but now obsolete. BCPL, BASH and BASIC are to be avoided. So yes you have a point
Actually, I'm not sure what you are getting at.
So there we have it, PHP, Postscript, Processing, PL/M, Perl, Python, Pascal, Prolog.
The evidence is overwhelming, there is a black hole in language design around the letter P.