| Re: Which is better WAVE or IDL? [message #3354] |
Sun, 22 January 1995 11:18  |
rep2857
Messages: 28
Registered: December 1994
|
Junior Member |
|
|
In article <3fm85f$lkr@lyra.csx.cam.ac.uk>,
Iarla Kilbane-Dawe <iarla@atm.ch.cam.ac.uk> wrote:
> I'm curious to know what anyone who may have used both systems
> might have to say about comparing the two. We use PV-WAVE
> but I've been told that IDL is
> a) cheaper, and,
> b) more robust and bug free in usage.
> Would anyone have any comments to make on this?
>
Hi Iarla -
I'm using both. I have heard that IDL was cheaper, but I haven't
checked the pricing for Wave for a couple years. If Wave is more
expensive, it could probably be justified if you use all of the
additional components provided by Wave such as the database access and
symbolic math. Each of these are additional items to the base cost, but
I believe they are unique to Wave. IDL has additional cost items, but
these are better described as applications rather than access functions.
One is a package called ENVI which is especially suited for working
with multi- and hyper-spectral datasets. They also mentioned a medical
application in a recent newsletter which I haven't had time to check
out, yet.
As fas as being more robust or bug-free, I think both are very robust
and bugs are certain to be in both products due to their code/library
sizes, nature of functions (math, statistics, widgets, etc.) provided,
and fairly limited number of users which explore the edges of these
types of functions.
Wave does not support 24-bit color for widget-based drawing
applications on Sun's and HP's. Currently, I am only working with 8-bit
displays on the Sun's I am using for Wave development, so I am hardware
limited. Some of my IDL development is making extensive use of 24-bit
display capabilities on the Sun's used for a different project.
Wave is in a partnership with IMSL which may give them a more extensive
math library, although IDL doesn't seem to be lacking anything that I
normally use.
I feel IDL is easier to program widgets in, but Wave has a larger set
of widgets. Either product allows creation of new widgets. If you are
used to Motif widgets, Wave widgets will seem more familiar. IDL
widgets make better use of keywords for adding additional functionality
to widgets, whereas Wave provides similar abilities by having to
include handler statements. The end user would be hard pressed to
determine whether widget appliations were written in IDL, Wave or
Motif, though.
This is probably a minor point, but for me it is worth mentioning: Wave
is not available for Mac or PowerMac, platforms. The IDL demo I used
several months ago for a PowerMac 7100/66 ran at 3/4 the speed of our
Sun 10/51 which I tested it against. Roughly a 4:1 price ratio (similar
memory and peripherals) and a 4:3 performance ratio between the Sun and
PowerMac (for running IDL only). These aren't latest and greatest
systems by either manufacturer, but they are adequate for simple
price:performance comparisons. Both manufacturers have cheaper and
faster systems at this time.
Mike Schienle Hughes Santa Barbara Research Center
rep2857@sbsun0010.sbrc.hac.com 75 Coromar Drive, M/S B28/87
Voice: (805)562-7466 Fax: (805)562-7881 Goleta, CA 93117
|
|
|
|
|
|
| Re: Which is better WAVE or IDL? [message #3364 is a reply to message #3363] |
Fri, 20 January 1995 05:24  |
pjclinch
Messages: 27
Registered: May 1993
|
Junior Member |
|
|
Iarla Kilbane-Dawe (iarla@atm.ch.cam.ac.uk) wrote:
: I'm curious to know what anyone who may have used both systems
: might have to say about comparing the two. We use PV-WAVE
: but I've been told that IDL is
: a) cheaper, and,
: b) more robust and bug free in usage.
: Would anyone have any comments to make on this?
The "cheaper" seems to apply to the US. In the UK, at education prices,
they cost exactly the same. Both can be evaluated free of charge, so you
can check for yourself on robustness.
I evaluated both side by side, and went for PV~Wave because *at the time*
I couldn't get the Solaris 2 IDL version working properly under the then
just released Sol 2.2. Were it not for that, I think I'd have tossed a
coin to decide, and that reason no longer applies... Both products are
good, and I've been impressed with both companies' backup and support too.
The "Advantage" version of Wave costs a bit more, but has more in the way
of maths and stats functionality than the basic version. Also, the help
is better (entire manual set online with hypertext links!) than Wave CL
and, unless it's improved from when I saw it last, IDL.
As suggested before, both companies will give you free demos so you can
compare them side by side. If you want a fiddle with IDL their UK
distributor is Floating Point on 01344 56921.
Pete.
--
Peter Clinch Dundee Teaching Hospitals NHS Trust
voice: 44 1382 660111 x 3637 snail: Directorate of Medical Physics
fax: 44 1382 640177 Ninewells Hospital
email: p.j.clinch@dundee.ac.uk Dundee DD1 9SY Scotland UK
|
|
|
|
| Re: Which is better WAVE or IDL? [message #3441 is a reply to message #3364] |
Mon, 23 January 1995 16:55 |
mahan
Messages: 17
Registered: October 1994
|
Junior Member |
|
|
Iarla Kilbane-Dawe (iarla@atm.ch.cam.ac.uk) wrote:
: I'm curious to know what anyone who may have used both systems
: might have to say about comparing the two. We use PV-WAVE
: but I've been told that IDL is
They have the same origin, but from my experience w/ running a couple of
pieces of computationally intensive code- RSI's IDL is about a factor
of 2 faster than PV-Wave on a SUN MP690-52 running Solaris 2.3. This
measure was done in the Summer of 1994. One of the tests that I ran was
a modified version of RSI's time_test code. The modification was
necessary for compatibility w/ VNI's pv-wave.
Regarding RSI's IDL and some benchmarks of IDL versions of Fortran 77
code and their Fortran 77 counterparts on several platforms, I'll
recommend the July/August 1994 issue of Computers in Physics, pg 439.
In this review article there are several platforms w/ corresponding
timings.
-Steve
------------------------------------------------------------ -----------
Stephen L. Mahan Telephone: 615-974-7847 Fax: 615-974-7843
Department of Physics and Astronomy, University of Tennessee, Knoxville
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
