| RE: MAP_SET error: linux [message #9201] |
Sat, 07 June 1997 00:00  |
mallozzi
Messages: 60
Registered: August 1994
|
Member |
|
|
David Fanning wrote:
>
> [A copy of this article was sent to the cited author.]
>
> Andy Loughe writes:
>
>> David,
>>
>> It appears you don't rely heavily upon map_set when using IDL.
>> These "low priority" grid line issues are *very* important to
>> me. At our lab we are judged not only by the quality of our
>> research, but also by the quality of the presentation of that
>> research. Next week we will present work to ABC television,
>> and I will use map_set to illustrate development of the current
>> El Nino. If these maps (to be presented in a movie loop using
>> Xinteranimate) all have a missing grid lines, to me and other geography
>> types it will make the presentation look second rate. If the same
>> occurs when I submit graphics for final publication to a journal,
>> the presentation will look second rate to thousands of journal
>> subscribers. I hope you see my point. Sure, there may be simple
>> work arounds to this, but have fun applying those work arounds to all
>> your IDL code which uses map_set. Software is supposed to make your
>> life simpler and more enjoyable... to increase your productivity, not
>> the other way around!
>>
>> I agree with Bob that it is frustrating to have a bug identified,
>> but not fixed. I've been there before, I'm there now. Certainly
>> I underestimate the magnitude of the task being undertaken by RSI,
>> but I hope they can be responsive, especially to those who put forth
>> the effort (at no charge) to beta test their proposed new releases.
>>
>> Andy
>
> My apologies to both Andy and Bob (who, BTW, admits that he *was*
> having a bad day in a private e-mail to me) if I stepped on their
> toes. I know perfectly well that one person's minor problem is
> another person's show stopper, but that is my point, pretty much.
> RSI has their own priorities that (hopefully) have to do with
> business decisions and keeping the largest number of customers
> happy at any one time.
>
> Lord knows I'm frustrated too. Right now I'm in the middle of
> a problem where "persistent" object graphics ain't acting so
> persistent. But I know some of those folks at RSI, and I can
> assure you that the ones I know are not sitting on their hands
> up there. They are working long, long hours to get this code
> working. They are frustrated too, no doubt.
>
> I'm just not sure it's helpful for them or us to cast our
> complaints with so much heat. I think we should identify
> problems, be persistent in asking for them to be fixed, but do
> it in a way that recognizes that RSI is not the enemy.
>
> Save the heat for your internet service provider and the IRS.
>
> Cheers,
>
> David
>
> --
> David Fanning, Ph.D.
> Fanning Software Consulting
> Customizable IDL Programming Courses
> Phone: 970-221-0438 E-Mail: davidf@dfanning.com
> Coyote's Guide to IDL Programming: http://www.dfanning.com
> IDL 5 Reports: http://www.dfanning.com/documents/anomaly5.html
I agree with what you said about IDL not having unlimited resources, but the
errors I pointed out are not related to anything new that was introduced in
IDL5.0. The PLOT function has been around since day 1, and it, in my opinion,
produces an error when using log scales. It does not matter if one uses "a
procedure that has been retired 3 versions ago", or the /xlog, /ylog keywords.
Take a look the image
ftp://crazyhorse.msfc.nasa.gov/pub/idl_log_err.gif
This plot was produced with the command
Plot, FINDGEN(100)+1, FINDGEN(100)+1, XR = [2, 8], XSTY = 1,/XLog, /Ylog
on my linux PC. In answer to your question "What is wrong with this...", the
xticks are not labeled, rendering the plot useless. I realize this can be
fixed by labeling the ticks explicitly, but I still must insist that this is an
error in the PLOT routine that has been around forever.
These same arguments can be applied to the MAP_SET function. It is not a new
function. Do you not agree that errors in established functions should be
fixed before (or at least concurrent with) new features? I must say, David,
that I very much disagree with your statement "I don't find it hard to believe
a grid line, for goodness sake, gets low priority." Putting a plot like that
in a proposal or publication is not, I think, a good way to impress your
colleagues.
I guess little, persistent errors like these, combined with the very expensive
cost of IDL, has me a little upset. All in all, I think RSI is doing a good
job, and that IDL is one of the best languages around. I will certainly
continue to use it. Also, you are right that it's probably not helpful for
them or us to cast our complaints with so much heat. It's just very
frustrating to report the bug, RSI responds with "Ok, got it", and then nothing
happens. In fact, with the log plot error, it was (sort of) fixed in 5.0b5,
then in 5.0b6, it was back to the old error...
OK, rant is done, thanks for listening :-)
Sincerely,
bob mallozzi
--
Dr. Robert S. Mallozzi
University of Alabama in Huntsville
http://cspar.uah.edu/~mallozzir/
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
