comp.lang.idl-pvwave archive
Messages from Usenet group comp.lang.idl-pvwave, compiled by Paulo Penteado

Home » Public Forums » archive » IDL and OPENGL
Show: Today's Messages :: Show Polls :: Message Navigator
E-mail to friend 
Return to the default flat view Create a new topic Submit Reply
Re: IDL and OpenGL [message #17299 is a reply to message #17065] Wed, 29 September 1999 00:00 Go to previous messageGo to previous message
ushomirs is currently offline  ushomirs
Messages: 14
Registered: May 1993
Junior Member
I think what SGI folks meant is that IDL routines are **not** compiled
into **machine** code, the way C programs are. So even an IDL routine
is "compiled", it is transformed into a bytecode for it's own virtual
machine. So unlike compiled C code, where the processor executes your
program directly, the processor executes the code for the "IDL virtual
machine", which in turn reads your program (or the "compiled" bytecode)
and interprets it step by step. So there is an extra level of
indirection involved in running IDL code. that's what the SGI people
were refering to.

If IDL code were truly compiled in the same sense as C code, then
this loop
FOR i=0,9 do foo[i]=2.*foo[i]

would take exactly as long as

foo=2.*foo

It doesn't, because the IDL "virtual machine" has to interpret every
iteration of the FOR loop, but it can just map the vector operation onto
a machine code for loop.

Now, in case of OpenGL, this distinction is a wash. So long as most of
the time is spent inside OpenGL, some slowness associated with
interpreting rather than executing machine code directly doesn't matter.

greg




THere's some discussion of this in IDL's external development guide.


In this respect IDL is much more like Java, where th


In article <37F22743.52AC5811@ssec.wisc.edu>,
Liam Gumley <Liam.Gumley@ssec.wisc.edu> wrote:
> Richard Tyc wrote:
>> Our group was having a meeting with some SGI folks yesterday and
some
>> interesting points were brought up which I hope some IDL experts
could shed
>> some light on. We were discussing the performance of IDL vs. OpenGL
source
>> based app.. Their point was that IDL is an interpreted language and
as such
>> is not optimized (or tuned as they put it) for OpenGL and thus runs
>> significantly slower than a custom C app using OpenGL.
>
> IDL does indeed run in an interpreting mode (at the command line),
where
> each statement is interpreted and executed separately. However once
you
> 'compile' an IDL procedure, it is no longer 'interpreted' each time it
> is called.
>
> Perhaps someone in the know could give us a brief description of how
IDL
> procedures are transformed from source code to CPU instructions.
>
> Cheers,
> Liam.
>
> --
> Liam E. Gumley
> Space Science and Engineering Center, UW-Madison
> http://cimss.ssec.wisc.edu/~gumley
>


Sent via Deja.com http://www.deja.com/
Before you buy.
[Message index]
 
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Previous Topic: !path problem
Next Topic: howto make a fits cube in IDL?

-=] Back to Top [=-
[ Syndicate this forum (XML) ] [ RSS ] [ PDF ]

Current Time: Wed Oct 08 19:59:02 PDT 2025

Total time taken to generate the page: 0.00417 seconds