| Re: IDL --> C++ ? [message #32625] |
Sat, 26 October 2002 21:04 |
muswick
Messages: 10
Registered: April 1998
|
Junior Member |
|
|
Your question has been asked several times over the years, (maybe not
to c++, but a translator to some other language.)
I think you need to ask yourself why you want the translator. I can
think of several, but the big one is:
Once C/C++ is code is made, the results could be compiled and you
could either sell or give your application away, without others
needing to buy runtime licenses or you having to pay royalty fees to
RSI. Given the approximate $500 cost of runtime licenses, this would
be desirable for simple applications. But if you have something
complex that is worth selling, than you basically have two choices,
rewrite the application, or pass the cost of the license onto your
customers. My company, chose the latter, since staying with IDL has
advantages for us.
For simple apps, or research groups, etc. you're using IDL because of
it's power, capabilities and ease of prototyping code, interactive
testing to application. That power and flexibility is what RSI sells,
and we pay for. A lot of work has and continues to go into IDL, and
RSI, needs to pay it's staff and make a profit. For groups, or
company internal use, there are a variety of license methods available
to fit needs and budgets.
If you ever could write such a translator, it would be worth selling,
or charge people for the service of doing the translation, maybe 10
cents per line of IDL code. But keep in mind, that IDL has long
history and is pretty much backward compatible. I am working on a
large application written some time ago, where the original programmer
implemented objects using HANDLES. The code still runs fine under IDL
5.5, but I had to dust cobwebs from my brain cells to remember how to
use them. Those are just some of the obsolete stuff you'd have
include.
Maybe posting why you want the translator, might help someone in the
newsgroup give you some alternatives.
Good luck
Gary
Robert <Rober-Mueller@My-Mail.ch> wrote in message news:<ap9kh3$7cr$1@news.online.de>...
> Hi folks,
>
> is there any application out there to translate IDL source code to C++
> source code ? Even buggy solutions are welcome ,) ...
> If there is no possibility right now, i'm going to programm eigher a full
> freatured ( free ! ) translator or a set of classes/ librarys that'll allow
> easy rewriteing any IDL application within C++ . ( helping hands and minds
> are more then welcome... )
>
> thank you for reading and for any response,
>
> Stefan
|
|
|
|
| Re: IDL --> C++ ? [message #32626 is a reply to message #32625] |
Fri, 25 October 2002 12:26  |
Justin[2]
Messages: 16
Registered: October 2002
|
Junior Member |
|
|
Oops. Looks like I posted this message a few more times than I had meant
to. Sorry about that. Still it was a very good post, so it probably
deserved it...
|
|
|
|
| Re: IDL --> C++ ? [message #32633 is a reply to message #32626] |
Thu, 24 October 2002 16:55 |
Kenneth P. Bowman
Messages: 585
Registered: May 2000
|
Senior Member |
|
|
In article <ap9kjj$7cr$3@news.online.de>,
Robert <Rober-Mueller@My-Mail.ch> wrote:
> Hi folks,
>
> is there any application out there to translate IDL source code to C++
> source code ? Even buggy solutions are welcome ,) ...
> If there is no possibility right now, i'm going to programm eigher a full
> freatured ( free ! ) translator or a set of classes/ librarys that'll allow
> easy rewriteing any IDL application within C++ . ( helping hands and minds
> are more then welcome... )
Are you by any chance referring to Interface Definition Language (IDL),
rather than Interactive Data Language (er, IDL)? This newsgroup is the
latter. The former has its own newsgroup: comp.lang.idl.
Cheers, Ken Bowman
|
|
|
|
| Re: IDL --> C++ ? [message #32634 is a reply to message #32633] |
Thu, 24 October 2002 15:45 |
Justin[2]
Messages: 16
Registered: October 2002
|
Junior Member |
|
|
Robert,
I think this might be a near impossible task for you unless you amass an
army to help, or unless you plan to implement only the most basic
implimentation.
Start with plotting. How do you plan to produce plots in C++ that will
work across different platforms? Even if you just implemented the screen
devices ('Win', 'X', 'MAC') and the 'PS' device, there would be hell of
a lot of coding needed to create a plot that looks the same as an IDL
plot. If you're considering things like widgets and pointers and object
graphics too... well I think it would be tough.
Also if you plan to do a 'simple' translation (i.e. IDL:"a=1" to
C++:"a=1;") without complicated run-time handling of variables, how will
you handle variable type (C++ is strongly typed, IDL weakly typed)? For
instance, in IDL you could happily write:
a = '123'
a = FLOAT(a)
and 'a' will change from string to float as needed. A simple translation
of this would fail in C++ since a variable cannot change type at
run-time. Similarly what about run-time creation of structures
(CREATE_STRUCT), for instance, or calling of statements created at
run-time (EXECUTE)? I think this list is endless...
Now this wouldn't be impossible (I guess IDL is written in C) but I
think it would be mighty hard. I don't think a translator is viable, and
to write an interpreter would be duplicating RSI's effort!
Regards,
Justin
|
|
|
|
| Re: IDL --> C++ ? [message #32635 is a reply to message #32634] |
Thu, 24 October 2002 15:45 |
Justin[2]
Messages: 16
Registered: October 2002
|
Junior Member |
|
|
Robert,
I think this might be a near impossible task for you unless you amass an
army to help, or unless you plan to implement only the most basic
implimentation.
Start with plotting. How do you plan to produce plots in C++ that will
work across different platforms? Even if you just implemented the screen
devices ('Win', 'X', 'MAC') and the 'PS' device, there would be hell of
a lot of coding needed to create a plot that looks the same as an IDL
plot. If you're considering things like widgets and pointers and object
graphics too... well I think it would be tough.
Also if you plan to do a 'simple' translation (i.e. IDL:"a=1" to
C++:"a=1;") without complicated run-time handling of variables, how will
you handle variable type (C++ is strongly typed, IDL weakly typed)? For
instance, in IDL you could happily write:
a = '123'
a = FLOAT(a)
and 'a' will change from string to float as needed. A simple translation
of this would fail in C++ since a variable cannot change type at
run-time. Similarly what about run-time creation of structures
(CREATE_STRUCT), for instance, or calling of statements created at
run-time (EXECUTE)? I think this list is endless...
Now this wouldn't be impossible (I guess IDL is written in C) but I
think it would be mighty hard. I don't think a translator is viable, and
to write an interpreter would be duplicating RSI's effort!
Regards,
Justin
|
|
|
|
| Re: IDL --> C++ ? [message #32636 is a reply to message #32635] |
Thu, 24 October 2002 15:42 |
Justin[2]
Messages: 16
Registered: October 2002
|
Junior Member |
|
|
Robert,
I think this might be a near impossible task for you unless you amass an
army to help, or unless you plan to implement only the most basic
implimentation.
Start with plotting. How do you plan to produce plots in C++ that will
work across different platforms? Even if you just implemented the screen
devices ('Win', 'X', 'MAC') and the 'PS' device, there would be hell of
a lot of coding needed to create a plot that looks the same as an IDL
plot. If you're considering things like widgets and pointers and object
graphics too... well I think it would be tough.
Also if you plan to do a 'simple' translation (i.e. IDL:"a=1" to
C++:"a=1;") without complicated run-time handling of variables, how will
you handle variable type (C++ is strongly typed, IDL weakly typed)? For
instance, in IDL you could happily write:
a = '123'
a = FLOAT(a)
and 'a' will change from string to float as needed. A simple translation
of this would fail in C++ since a variable cannot change type at
run-time. Similarly what about run-time creation of structures
(CREATE_STRUCT), for instance, or calling of statements created at
run-time (EXECUTE)? I think this list is endless...
Now this wouldn't be impossible (I guess IDL is written in C) but I
think it would be mighty hard. I don't think a translator is viable, and
to write an interpreter would be duplicating RSI's effort!
Regards,
Justin
|
|
|
|
| Re: IDL --> C++ ? [message #32637 is a reply to message #32636] |
Thu, 24 October 2002 15:42 |
Justin[2]
Messages: 16
Registered: October 2002
|
Junior Member |
|
|
Robert,
I think this might be a near impossible task for you unless you amass an
army to help, or unless you plan to implement only the most basic
implimentation.
Start with plotting. How do you plan to produce plots in C++ that will
work across different platforms? Even if you just implemented the screen
devices ('Win', 'X', 'MAC') and the 'PS' device, there would be hell of
a lot of coding needed to create a plot that looks the same as an IDL
plot. If you're considering things like widgets and pointers and object
graphics too... well I think it would be tough.
Also if you plan to do a 'simple' translation (i.e. IDL:"a=1" to
C++:"a=1;") without complicated run-time handling of variables, how will
you handle variable type (C++ is strongly typed, IDL weakly typed)? For
instance, in IDL you could happily write:
a = '123'
a = FLOAT(a)
and 'a' will change from string to float as needed. A simple translation
of this would fail in C++ since a variable cannot change type at
run-time. Similarly what about run-time creation of structures
(CREATE_STRUCT), for instance, or calling of statements created at
run-time (EXECUTE)? I think this list is endless...
Now this wouldn't be impossible (I guess IDL is written in C) but I
think it would be mighty hard. I don't think a translator is viable, and
to write an interpreter would be duplicating RSI's effort!
Regards,
Justin
|
|
|
|
| Re: IDL --> C++ ? [message #32638 is a reply to message #32637] |
Thu, 24 October 2002 15:40 |
Justin[2]
Messages: 16
Registered: October 2002
|
Junior Member |
|
|
Robert,
I think this might be a near impossible task for you unless you amass an
army to help, or unless you plan to implement only the most basic
implimentation.
Start with plotting. How do you plan to produce plots in C++ that will
work across different platforms? Even if you just implemented the screen
devices ('Win', 'X', 'MAC') and the 'PS' device, there would be hell of
a lot of coding needed to create a plot that looks the same as an IDL
plot. If you're considering things like widgets and pointers and object
graphics too... well I think it would be tough.
Also if you plan to do a 'simple' translation (i.e. IDL:"a=1" to
C++:"a=1;") without complicated run-time handling of variables, how will
you handle variable type (C++ is strongly typed, IDL weakly typed)? For
instance, in IDL you could happily write:
a = '123'
a = FLOAT(a)
and 'a' will change from string to float as needed. A simple translation
of this would fail in C++ since a variable cannot change type at
run-time. Similarly what about run-time creation of structures
(CREATE_STRUCT), for instance, or calling of statements created at
run-time (EXECUTE)? I think this list is endless...
Now this wouldn't be impossible (I guess IDL is written in C) but I
think it would be mighty hard. I don't think a translator is viable, and
to write an interpreter would be duplicating RSI's effort!
Regards,
Justin
|
|
|
|
| Re: IDL --> C++ ? [message #32639 is a reply to message #32638] |
Thu, 24 October 2002 15:40 |
Justin[2]
Messages: 16
Registered: October 2002
|
Junior Member |
|
|
Robert,
I think this might be a near impossible task for you unless you amass an
army to help, or unless you plan to implement only the most basic
implimentation.
Start with plotting. How do you plan to produce plots in C++ that will
work across different platforms? Even if you just implemented the screen
devices ('Win', 'X', 'MAC') and the 'PS' device, there would be hell of
a lot of coding needed to create a plot that looks the same as an IDL
plot. If you're considering things like widgets and pointers and object
graphics too... well I think it would be tough.
Also if you plan to do a 'simple' translation (i.e. IDL:"a=1" to
C++:"a=1;") without complicated run-time handling of variables, how will
you handle variable type (C++ is strongly typed, IDL weakly typed)? For
instance, in IDL you could happily write:
a = '123'
a = FLOAT(a)
and 'a' will change from string to float as needed. A simple translation
of this would fail in C++ since a variable cannot change type at
run-time. Similarly what about run-time creation of structures
(CREATE_STRUCT), for instance, or calling of statements created at
run-time (EXECUTE)? I think this list is endless...
Now this wouldn't be impossible (I guess IDL is written in C) but I
think it would be mighty hard. I don't think a translator is viable, and
to write an interpreter would be duplicating RSI's effort!
Regards,
Justin
|
|
|
|
| Re: IDL --> C++ ? [message #32640 is a reply to message #32639] |
Thu, 24 October 2002 15:33 |
Justin[2]
Messages: 16
Registered: October 2002
|
Junior Member |
|
|
Robert,
I think this might be a near impossible task for you unless you amass an
army to help, or unless you plan to implement only the most basic
implimentation.
Start with plotting. How do you plan to produce plots in C++ that will
work across different platforms? Even if you just implemented the screen
devices ('Win', 'X', 'MAC') and the 'PS' device, there would be hell of
a lot of coding needed to create a plot that looks the same as an IDL
plot. If you're considering things like widgets and pointers and object
graphics too... well I think it would be tough.
Also if you plan to do a 'simple' translation (i.e. IDL:"a=1" to
C++:"a=1;") without complicated run-time handling of variables, how will
you handle variable type (C++ is strongly typed, IDL weakly typed)? For
instance, in IDL you could happily write:
a = '123'
a = FLOAT(a)
and 'a' will change from string to float as needed. A simple translation
of this would fail in C++ since a variable cannot change type at
run-time. Similarly what about run-time creation of structures
(CREATE_STRUCT), for instance, or calling of statements created at
run-time (EXECUTE)? I think this list is endless...
Now this wouldn't be impossible (I guess IDL is written in C) but I
think it would be mighty hard. I don't think a translator is viable, and
to write an interpreter would be duplicating RSI's effort!
Regards,
Justin
|
|
|
|
| Re: IDL --> C++ ? [message #32641 is a reply to message #32640] |
Thu, 24 October 2002 15:32 |
Justin[2]
Messages: 16
Registered: October 2002
|
Junior Member |
|
|
Robert,
I think this might be a near impossible task for you unless you amass an
army to help, or unless you plan to implement only the most basic
implimentation.
Start with plotting. How do you plan to produce plots in C++ that will
work across different platforms? Even if you just implemented the screen
devices ('Win', 'X', 'MAC') and the 'PS' device, there would be hell of
a lot of coding needed to create a plot that looks the same as an IDL
plot. If you're considering things like widgets and pointers and object
graphics too... well I think it would be tough.
Also if you plan to do a 'simple' translation (i.e. IDL:"a=1" to
C++:"a=1;") without complicated run-time handling of variables, how will
you handle variable type (C++ is strongly typed, IDL weakly typed)? For
instance, in IDL you could happily write:
a = '123'
a = FLOAT(a)
and 'a' will change from string to float as needed. A simple translation
of this would fail in C++ since a variable cannot change type at
run-time. Similarly what about run-time creation of structures
(CREATE_STRUCT), for instance, or calling of statements created at
run-time (EXECUTE)? I think this list is endless...
Now this wouldn't be impossible (I guess IDL is written in C) but I
think it would be mighty hard. I don't think a translator is viable, and
to write an interpreter would be duplicating RSI's effort!
Regards,
Justin
|
|
|
|
| Re: IDL --> C++ ? [message #32642 is a reply to message #32641] |
Thu, 24 October 2002 15:24 |
Mark Hadfield
Messages: 783
Registered: May 1995
|
Senior Member |
|
|
"Robert" <Rober-Mueller@My-Mail.ch> wrote in message
news:ap9qfr$act$1@news.online.de...
> Pavel A. Romashkin wrote:
>
>> I envy your ambitious mind and ample time at hand. I am sure the
>> NG will be anxiously waiting for your program. I am afraid I will
>> not see it as I plan to retire in some 30 years.
>
> By the way , you're answer has not been verry helpful. Instead of
> trying to mess around with me you shall better type something
> useful.
I think you would do well to get used to the style of this newsgroup,
Robert. Flaming is definitely not appreciated but gently barbed humour
(as in Pavel's message) is, and is considered to be something of an
art form. Please do not take it personally.
I think Pavel was being very helpful. He was implying that writing a
useful IDL -> C++ syntax translator would be a huge job. Don't go
there!
One the other hand, pulling off this feat should earn you a permanent
membership on the IDL Expert Programmers Association.
--
Mark Hadfield "Ka puwaha te tai nei, Hoea tatou"
m.hadfield@niwa.co.nz
National Institute for Water and Atmospheric Research (NIWA)
|
|
|
|
| Re: IDL --> C++ ? [message #32644 is a reply to message #32642] |
Thu, 24 October 2002 14:58 |
Robert[1]
Messages: 3
Registered: October 2002
|
Junior Member |
|
|
Pavel A. Romashkin wrote:
> Robert,
> I envy your ambitious mind and ample time at hand.
> I am sure the NG will be anxiously waiting for your program. I am afraid
> I will not see it as I plan to retire in some 30 years.
> Good luck,
> Pavel
>
Hey hi Pavel,
how long do you need to write a "Hello World" ? more or less then a year ?
*g , but no kidding, some work is allready done and I'm sure to write the
core funktions woun't took 30 years or more ,)
By the way , you're answer has not been verry helpful. Instead of trying to
mess around with me you shall better type something useful.
cheers,
Robert
|
|
|
|
| Re: IDL --> C++ ? [message #32645 is a reply to message #32644] |
Thu, 24 October 2002 14:14 |
Pavel A. Romashkin
Messages: 531
Registered: November 2000
|
Senior Member |
|
|
Robert,
I envy your ambitious mind and ample time at hand.
I am sure the NG will be anxiously waiting for your program. I am afraid
I will not see it as I plan to retire in some 30 years.
Good luck,
Pavel
Robert wrote:
>
> Hi folks,
>
> is there any application out there to translate IDL source code to C++
> source code ? Even buggy solutions are welcome ,) ...
> If there is no possibility right now, i'm going to programm eigher a full
> freatured ( free ! ) translator or a set of classes/ librarys that'll allow
> easy rewriteing any IDL application within C++ . ( helping hands and minds
> are more then welcome... )
>
> thank you for reading and for any response,
>
> Robert
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
