| Re: Anyone used FastDL? [message #51952] |
Tue, 19 December 2006 15:31 |
ronn
Messages: 123
Registered: April 1999
|
Senior Member |
|
|
Mike,
There is another alternative to FastDL and that is my interface to PVM
from Oak Ridge. The link is here:
http://www.kilvarock.com/software/idltopvm.htm
PVM itself is free and runs on windows and *nix, as does my interface.
Plus the cost of my code is an order of magnitude less :)
-Ronn
Ronn Kling
KRS, inc.
email: r...@rlkling.com
Use these books to upgrade your IDL skills.
"Application Development with IDL"
"Calling C from IDL, Using DLM's to extend your IDL code". SECOND
EDITION includes C++ and debugging!
"Power Graphics with IDL, A Beginner's Guide to Object Graphics",
http://www.kilvarock.com
|
|
|
|
| Re: Anyone used FastDL? [message #51962 is a reply to message #51952] |
Tue, 19 December 2006 02:35  |
Robbie
Messages: 165
Registered: February 2006
|
Senior Member |
|
|
A few points I'd like to clarify:
I don't think that ITTVIS is setting the benchmark for these expensive
prices. The price of FastDL and IDL Analyst are more likely based on
the original prices that TechX and Visual Numerics charged. ITTVIS is
really just passing on the cost.
There are many ways to achieve clustered computing and none of them are
cheap so far as time or money is concerned. Most solutions involve
copious amounts of recoding. Even a simple batch processing script
(like you suggest) over several computers can end quickly in a train
wreck. I've done some preliminary work on network semaphores and I'm
hoping to find a solution to avoid colliding batch processes.
A better example of this kind of behavior is IDLffDICOMex,
IDLExportBridge and IDLDataMiner. My understanding is that these
products are wholly owned by ITTVIS. I think that releasing them as
pricey add ons stagnates the language. Customers are expected to recode
their DICOM applications when they upgrade from IDLffDICOM to
IDLffDICOMex. Customers are expected to fiddle with musical sav files
until IDLDataMiner becomes absolutely necessary. It's bad for the
customer and it's bad for the language.
Robbie
|
|
|
|
| Re: Anyone used FastDL? [message #51964 is a reply to message #51962] |
Mon, 18 December 2006 17:50 |
Richard French
Messages: 173
Registered: December 2000
|
Senior Member |
|
|
I'm still a big fan of IDL, Rick, but $3000 for an educational license for
FastDL for an 8 CPU cluster + $450/yr maintenance is not small change. I'll
follow your suggestion and find out about a trial FastDL license. I don't
have any experience in parallel processing programming, and I have no doubt
that the devil is in the details in how much of an improvement one can
really get. What's unclear to me at this point is how much harder it would
be for me to develop my own communication links between separate IDL
processes on different CPUS than to rewrite the code using FastDL to do more
or less the same thing. Here is where having some substantial examples coded
in different ways would really help. For example, if I have 100 images to
process using identical code, it would not be that hard to start 8 IDL
processes on different CPUS and have each one grab the next available image.
I'm not sure how much more than this TaskDL does for this kind of operation.
Of course, doing parallel processing a finer level of things (miIDL) is a
different story, and here I know that I have a lot to learn.
Dick
On 12/18/06 8:06 PM, in article em7dqu$bm6$1@news.nems.noaa.gov, "Rick
Towler" <rick.towler@nomail.noaa.gov> wrote:
> That's an unfair comparison. Have you done an IDL vs. MATLAB pricing
> comparison lately? Compared to MATLAB, IDL delivers some serious value.
> And while I haven't asked in a while, RS-err ITTVIS has always been
> very generous in providing trial licenses to their products. So I'm sure
> if you wanted to try IDL Analyst you could, free of charge.
>
> -Rick
>
|
|
|
|
| Re: Anyone used FastDL? [message #51965 is a reply to message #51964] |
Mon, 18 December 2006 17:06 |
Rick Towler
Messages: 821
Registered: August 1998
|
Senior Member |
|
|
Richard G. French wrote:
> It seems to me that IDL is going the way of MATLAB - you start paying (quite
> a bit) for extra features. Do you want reliable numerical routines? Pay for
> IDL Analyst. Do you want to do parallel processing? Pay (quite a bit) for
> FastDL. These extras are actually quite expensive, and in some cases it is
> hard to know what is really going to be useful until you fork over the dough
> and try it out.
That's an unfair comparison. Have you done an IDL vs. MATLAB pricing
comparison lately? Compared to MATLAB, IDL delivers some serious value.
And while I haven't asked in a while, RS-err ITTVIS has always been
very generous in providing trial licenses to their products. So I'm sure
if you wanted to try IDL Analyst you could, free of charge.
-Rick
|
|
|
|
| Re: Anyone used FastDL? [message #51966 is a reply to message #51965] |
Mon, 18 December 2006 16:12 |
David Fanning
Messages: 11724
Registered: August 2001
|
Senior Member |
|
|
Richard G. French writes:
> It seems to me that IDL is going the way of MATLAB - you start paying (quite
> a bit) for extra features. Do you want reliable numerical routines? Pay for
> IDL Analyst. Do you want to do parallel processing? Pay (quite a bit) for
> FastDL. These extras are actually quite expensive, and in some cases it is
> hard to know what is really going to be useful until you fork over the dough
> and try it out.
Well, it's not a mom and pop shop anymore, that's for
sure. But the FastDL guys are part of the high-end
physics community, where you don't just start your
IDL program and go get a cup of coffee. You start your
program and go teach a class, get a haircut, stop at
Starbucks, etc. Those guys are pretty clear they could
use something like this. :-)
Cheers,
David
--
David Fanning, Ph.D.
Fanning Software Consulting, Inc.
Coyote's Guide to IDL Programming: http://www.dfanning.com/
Sepore ma de ni thui. ("Perhaps thou speakest truth.")
|
|
|
|
| Re: Anyone used FastDL? [message #51967 is a reply to message #51966] |
Mon, 18 December 2006 16:00 |
Richard French
Messages: 173
Registered: December 2000
|
Senior Member |
|
|
It seems to me that IDL is going the way of MATLAB - you start paying (quite
a bit) for extra features. Do you want reliable numerical routines? Pay for
IDL Analyst. Do you want to do parallel processing? Pay (quite a bit) for
FastDL. These extras are actually quite expensive, and in some cases it is
hard to know what is really going to be useful until you fork over the dough
and try it out.
Dick French
On 12/18/06 2:10 PM, in article MPG.1ff0957c2553eafd989e56@news.frii.com,
"David Fanning" <news@dfanning.com> wrote:
> Mike writes:
>
>> There have been occasional marketing messages regarding a product
>> called FastDL in this news group, but I have not seen any posts from
>> people who've used it. Has anyone out there used this product? Does
>> anyone have an opinion strong enought to recommend it or not?
>
> I haven't seen this final product, so I can't say much about
> it, but I have worked with these folks in the past. I know
> they are technically competent and very bright. I expect
> this product does exactly what they say it does. I wouldn't
> hesitate to check it out.
>
> Cheers,
>
> David
|
|
|
|
| Re: Anyone used FastDL? [message #51969 is a reply to message #51967] |
Mon, 18 December 2006 11:10 |
David Fanning
Messages: 11724
Registered: August 2001
|
Senior Member |
|
|
Mike writes:
> There have been occasional marketing messages regarding a product
> called FastDL in this news group, but I have not seen any posts from
> people who've used it. Has anyone out there used this product? Does
> anyone have an opinion strong enought to recommend it or not?
I haven't seen this final product, so I can't say much about
it, but I have worked with these folks in the past. I know
they are technically competent and very bright. I expect
this product does exactly what they say it does. I wouldn't
hesitate to check it out.
Cheers,
David
--
David Fanning, Ph.D.
Fanning Software Consulting, Inc.
Coyote's Guide to IDL Programming: http://www.dfanning.com/
Sepore ma de ni thui. ("Perhaps thou speakest truth.")
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
