| Re: Rumors about new plotting software in IDL [message #5222] |
Tue, 31 October 1995 00:00 |
landsman
Messages: 93
Registered: August 1991
|
Member |
|
|
In article <472svd$s0a@post.gsfc.nasa.gov>, thompson@achilles.nascom.nasa.gov (William Thompson) writes...
>
> I've heard a rumor that RSI is planning to change their plotting software in an
> upcoming release of IDL. From what I hear, the new software will not be
> downwardly compatible with the old software. Supposedly, user's will be
> initially given the option of using either the old or new methods, but
> eventually, the old software will be phased out.
>
> Does anybody know anything more definite about this? I'm very concerned about
> backwards compatibility. Making a new version of IDL that isn't compatible
> with all our current software only pours flames on the debate about whether we
> should look for something to replace it with.
>
> Of course, I may have this hopelessly garbled. I'm only repeating what I heard
> second hand. Thus, take what I say with a grain of salt. The point of this
> post is to try to get more detailed information.
Here are some of my notes on a talk given by David Stern at the Astronomical
Data Analysis and Software Systems (ADASS) meeting in Tuscon last week.
(1) IDL V5.0 (~Summer 96) will use new internal plotting software, but RSI
will make every effort to make the user interface backward compatible with
the current version of IDL.
(2) However, the new plotting software will not be 100% backward
compatible. Therefore, users will have the option using either the
new or the old internal plotting software. I am pretty sure that nothing
was said about eventually phasing out the old software, although that seems
a reasonable supposition.
So my impression was (1) it will not be very difficult to update existing
IDL code to run the new internal plotting software, certainly much easier than
the changes that were needed to go from IDL V1.0 to V2.0, and (2) existing IDL
code should work without any changes on the most current version of IDL
for at least the next ~2 years.
The above aproach seems reasonable to me *provided* that the new graphics
software provides significant improvements, and that detailed instructions
are given on how to update existing IDL code (perhaps it could be automated?)
I think backwards compatibility should be a very high priority for IDL, but
it should not be an inviolable rule (or as high a priority as it should be for,
say, FORTRAN), if there is no other way to add powerful new features to the
language. Projects can always freeze their IDL version, if absolute backward
compatibility is required.
I should emphasize that my impressions are based on a 30 minute talk, and not
on any detailed conversations or inside information, and that the release of
V5.0 is sufficiently far in the future, that RSI could easily change their
plans.
--Wayne Landsman landsman@stars.gsfc.nasa.gov
|
|
|
|
| Re: Rumors about new plotting software in IDL [message #5224 is a reply to message #5222] |
Tue, 31 October 1995 00:00  |
Ann_Mazuk
Messages: 5
Registered: November 1993
|
Junior Member |
|
|
In article <472svd$s0a@post.gsfc.nasa.gov>,
thompson@achilles.nascom.nasa.gov (William Thompson) wrote:
> I've heard a rumor that RSI is planning to change their plotting
software in an
> upcoming release of IDL. From what I hear, the new software will not be
> downwardly compatible with the old software. Supposedly, user's will be
> initially given the option of using either the old or new methods, but
> eventually, the old software will be phased out.
>
> Does anybody know anything more definite about this? I'm very concerned about
> backwards compatibility. Making a new version of IDL that isn't compatible
> with all our current software only pours flames on the debate about whether we
> should look for something to replace it with.
>
> Of course, I may have this hopelessly garbled. I'm only repeating what
I heard
> second hand. Thus, take what I say with a grain of salt. The point of this
> post is to try to get more detailed information.
>
I cannot confirm or deny the rumors, but the routines PLOT_IO, PLOT_OI and
PLOT_OO are now obsolete and undocumented, however they do still work.
Instead, the user is supposed to use the keywords XLOG, YLOG, and ZLOG in
the PLOT routine. I saw this in the IDL Online Help under New Keywords to
Existing Routines. When I first saw this, I assumed that RSI was going to
do away with the routines in future releases of IDL. We, too, have a lot
of software we have developed that uses the plot routines, so it would be
ashame if it would not be downwardly compatible with older versions.
> Bill Thompson
Ann
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ^^^^^^^^^^^^^^^^^^
Ann Mazuk Internet: Ann_Mazuk@qmail2.aero.org
El Segundo, CA SPAN: DIRAC2::AMAZUK
Boring Disclaimer: These opinions are my own!
Ford Motto: Quality is #1
Microsoft Motto: Quality is #1.1
Pentium Motto: Quality is #0.9994728
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
