| Message From RSI VP of Engineering [message #27225] |
Tue, 16 October 2001 10:36  |
Matthew Powell
Messages: 6
Registered: October 2001
|
Junior Member |
|
|
As VP of Engineering and IT at Research Systems, I would like to elaborate
on the posting from our CEO. I have been monitoring the newsgroups and
understand that this is an issue that many of you are very passionate about.
So, I felt it important to post this message to all of you.
As with most issues there is more than meets the eye with our recent
announcement regarding our decision not to support Mac OS X. The most
important one being that in our desire to be the best scientific
visualization package we can be, we had to shed some unprofitable platforms
in order to focus on the core platforms which make up the bulk of IDL sales.
Our initial desire was to deliver a fully native OS X solution that included
native widgets and a redesigned Aqua user interface. Once we realized the
costs associated with doing that, we knew we could not support the Macintosh
platform profitably. Whether we like it or not, in this day and age,
profitability is a requirement for which we are held accountable.
Now for the good news. In subsequent discussions with Apple they have made
us aware of a commercial X-Windows library for OS X. We are in the process
of evaluating it for use in a native Unix/X-Windows implementation of IDL
and ENVI for the Mac OS X platform. This would solve many technical issues
for us and allow us to continue to support the Macintosh platform both
natively and profitably, as it would leverage off our other Unix/X
platforms. The only thing this does not accomplish is providing IDL with a
new Aqua UI and widget set.
While I cannot promise this will be successful and that we can take it to
market, we are very encouraged by our early results working with the
library. We hope to have some solid results upon which we can base a solid
commitment to our customers in the very near future. We certainly apologize
for any inconvenience this may have caused you and your colleagues and ask
that you give us an opportunity to complete our investigation and try to
come up with a solution that will meet your needs and retain you as a
satisfied customer of RSI.
Thank you for your time and consideration.
Richard W. Cooke
Vice President, Engineering and IT
Research Systems, Inc. - A Kodak Company
|
|
|
|
|
|
| Re: Message From RSI VP of Engineering [message #27275 is a reply to message #27225] |
Wed, 17 October 2001 09:55  |
Bernard K.
Messages: 11
Registered: October 2001
|
Junior Member |
|
|
I am not familiar with command line IDL on Unix implementations. Could
someone elaborate a bit more on what an Aqua-less IDL would probably
look like? Just a command line tool that would ressemble using the
actual command line in IDL for OS9? Apart from loosing the IDL
developper environment and its associated menus (file edit search run
project macros etc) what else would be missing?
Thanks,
Bernard.
In article <3BCDAF3E.5B496216@astro.cornell.edu>, JD Smith
<jdsmith@astro.cornell.edu> wrote:
> "Pavel A. Romashkin" wrote:
>>
>> I am certainly for this. Mainly because I am sure that if the support
>> for OSX continues in any form, it will eventually include all the native
>> features of OSX, including Aqua. I am *speculating* that many of these
>> features are there already, and taking them out would be more of the
>> effort than leaving them in. For one, I expect graphics already being
>> (mostly) ready in the native OSX form. So, using existing Unix libraries
>> to cut down on the cost of implementation might sustain the OSX version
>> long enough to make it after a while what we all thought it will be.
>
> I think the real situation is this: RSI ran into a lot of work porting
> all of the drawing/widget code to Aqua, which is an entirely different
> API than either earlier MacOS, or Unix/Motif. It offers some
> significant advantages, but represents a pretty major display code fork
> (although with the option of eventually dropping MacOS <=9). So I doubt
> RSI would start with the abandoned OSX fork, try to strip out the Aqua
> interface stuff, and replace it with a X11/Motif subsystem. Instead,
> they'd start with their Linux or Solaris port, and try to migrate it
> over to the BSD layer of OSX + a proprietary X-server, without much more
> than a recompile and some fiddling.
>
> Although obviously something is better than nothing, people should
> realize what they'd be getting with such a product: an undersupported
> and "ugly duckling" IDL knock-off, made explicity to silence the
> critical masses. Their basic lack of commitment to the Mac platform
> would probably remain. On the other hand, I, like Pavel, am hopeful
> this attitude might change with time, and having some Mac development
> remaining would of course help.
>
> JD
|
|
|
|
| Re: Message From RSI VP of Engineering [message #27277 is a reply to message #27225] |
Wed, 17 October 2001 09:18 |
John-David T. Smith
Messages: 384
Registered: January 2000
|
Senior Member |
|
|
"Pavel A. Romashkin" wrote:
>
> I am certainly for this. Mainly because I am sure that if the support
> for OSX continues in any form, it will eventually include all the native
> features of OSX, including Aqua. I am *speculating* that many of these
> features are there already, and taking them out would be more of the
> effort than leaving them in. For one, I expect graphics already being
> (mostly) ready in the native OSX form. So, using existing Unix libraries
> to cut down on the cost of implementation might sustain the OSX version
> long enough to make it after a while what we all thought it will be.
I think the real situation is this: RSI ran into a lot of work porting
all of the drawing/widget code to Aqua, which is an entirely different
API than either earlier MacOS, or Unix/Motif. It offers some
significant advantages, but represents a pretty major display code fork
(although with the option of eventually dropping MacOS <=9). So I doubt
RSI would start with the abandoned OSX fork, try to strip out the Aqua
interface stuff, and replace it with a X11/Motif subsystem. Instead,
they'd start with their Linux or Solaris port, and try to migrate it
over to the BSD layer of OSX + a proprietary X-server, without much more
than a recompile and some fiddling.
Although obviously something is better than nothing, people should
realize what they'd be getting with such a product: an undersupported
and "ugly duckling" IDL knock-off, made explicity to silence the
critical masses. Their basic lack of commitment to the Mac platform
would probably remain. On the other hand, I, like Pavel, am hopeful
this attitude might change with time, and having some Mac development
remaining would of course help.
JD
|
|
|
|
|
|
| Re: Message From RSI VP of Engineering [message #27284 is a reply to message #27225] |
Wed, 17 October 2001 06:20 |
btt
Messages: 345
Registered: December 2000
|
Senior Member |
|
|
Hello,
My understanding of the issues is limited when it gets down to the details
like X-Windows and OS X. The technicalities whiz over my head; I wish I
could comment on this matter thoughtfully.
The overriding issue for me is access to the continuing new feature
implementations in IDL. There are some features in the next release of IDL
that we have been very excited about and anxiously awaiting for use in our
Mac only lab facility. So far, every new IDL release has introduced
features to our lab software that we have exploited as much as we can. That
ever expanding capability for our platform is what we want out of IDL. I
won't know what I'm missing from an OS X native IDL versus a UNIX like
implementation...a price of ignorance, I guess. Unless the next
implementation is a step down in performance from the classic Mac
implementation, how could I argue against ANY continuation of Mac support?
A fella who never had peppermint stick ice cream thinks that plain vanilla
is ambrosia.
Ben
PS Rum raisin ice cream is my favorite.
----- Original Message -----
From: "Liam Gumley" <Liam.Gumley@ssec.wisc.edu>
Newsgroups: comp.lang.idl-pvwave
Sent: Tuesday, October 16, 2001 10:26 PM
Subject: Re: Message From RSI VP of Engineering
>
> Any comment from the community on whether this suggestion would meet the
> needs of Mac OS X users? (That is, an OS X version of IDL that uses
> X-Windows for direct/object graphics and user interface). Granted, you
don't
> get the fancy Aqua user interface, but do IDL programmers really care, as
> long IDL is running on their Mac? For cross-platform development purposes,
I
> would think a more UNIX-like version for the Mac would be a good thing.
>
> Cheers,
> Liam.
> Practical IDL Programming
> http://www.gumley.com/
>
>
>
|
|
|
|
|
|
| Re: Message From RSI VP of Engineering [message #27298 is a reply to message #27225] |
Tue, 16 October 2001 21:25 |
John-David T. Smith
Messages: 384
Registered: January 2000
|
Senior Member |
|
|
Liam Gumley wrote:
>
> "Matthew Powell" <mpowell@rsinc.com> wrote in message
> news:9qhr6c$3to$1@news.rsinc.com...
>> As VP of Engineering and IT at Research Systems, I would like to elaborate
>> on the posting from our CEO. I have been monitoring the newsgroups and
>> understand that this is an issue that many of you are very passionate
> about.
>> So, I felt it important to post this message to all of you.
>>
>> As with most issues there is more than meets the eye with our recent
>> announcement regarding our decision not to support Mac OS X. The most
>> important one being that in our desire to be the best scientific
>> visualization package we can be, we had to shed some unprofitable
> platforms
>> in order to focus on the core platforms which make up the bulk of IDL
> sales.
>> Our initial desire was to deliver a fully native OS X solution that
> included
>> native widgets and a redesigned Aqua user interface. Once we realized the
>> costs associated with doing that, we knew we could not support the
> Macintosh
>> platform profitably. Whether we like it or not, in this day and age,
>> profitability is a requirement for which we are held accountable.
>>
>> Now for the good news. In subsequent discussions with Apple they have made
>> us aware of a commercial X-Windows library for OS X. We are in the
> process
>> of evaluating it for use in a native Unix/X-Windows implementation of IDL
>> and ENVI for the Mac OS X platform. This would solve many technical
> issues
>> for us and allow us to continue to support the Macintosh platform both
>> natively and profitably, as it would leverage off our other Unix/X
>> platforms. The only thing this does not accomplish is providing IDL with a
>> new Aqua UI and widget set.
>>
>> While I cannot promise this will be successful and that we can take it to
>> market, we are very encouraged by our early results working with the
>> library. We hope to have some solid results upon which we can base a solid
>> commitment to our customers in the very near future. We certainly
> apologize
>> for any inconvenience this may have caused you and your colleagues and ask
>> that you give us an opportunity to complete our investigation and try to
>> come up with a solution that will meet your needs and retain you as a
>> satisfied customer of RSI.
>
> Any comment from the community on whether this suggestion would meet the
> needs of Mac OS X users? (That is, an OS X version of IDL that uses
> X-Windows for direct/object graphics and user interface). Granted, you don't
> get the fancy Aqua user interface, but do IDL programmers really care, as
> long IDL is running on their Mac? For cross-platform development purposes, I
> would think a more UNIX-like version for the Mac would be a good thing.
Though not an IDL mac-user per say, I would regard this as only a partial
solution, not necessarily because the loss of an Aqua interface would inhibit
sales (which of course it would) and confuse some users (imagine a Mac person
looking at a Unix file-finding dialog -- filter, what filter?), but more
importantly, X11 is not a standard component of OSX, but instead has been
grafted on after the fact in various ways. RSI would need to bundle X11 with
IDL and require running some X11 server, along with a Motif toolkit. This
coexistence of the native graphics subsystem with X would probably severely
hamper draw speed, both for widgets, and even for images/plots. It also might
have some big consequences for OpenGL driven Object Graphics (support for which,
as RSI previously bellowed from the hallways, was one of the striking new
benefits available with OSX).
So while I think this would be better than *nothing* (and while they're at it,
they might as well recompile for LinuxPPC), I believe it falls far short of a
drop-in replacement, and would probably end up looking a little crufty even for
us cruft-hardened IDL wonks.
JD
|
|
|
|
| Re: Message From RSI VP of Engineering [message #27307 is a reply to message #27225] |
Tue, 16 October 2001 19:26 |
Liam E. Gumley
Messages: 378
Registered: January 2000
|
Senior Member |
|
|
"Matthew Powell" <mpowell@rsinc.com> wrote in message
news:9qhr6c$3to$1@news.rsinc.com...
> As VP of Engineering and IT at Research Systems, I would like to elaborate
> on the posting from our CEO. I have been monitoring the newsgroups and
> understand that this is an issue that many of you are very passionate
about.
> So, I felt it important to post this message to all of you.
>
> As with most issues there is more than meets the eye with our recent
> announcement regarding our decision not to support Mac OS X. The most
> important one being that in our desire to be the best scientific
> visualization package we can be, we had to shed some unprofitable
platforms
> in order to focus on the core platforms which make up the bulk of IDL
sales.
> Our initial desire was to deliver a fully native OS X solution that
included
> native widgets and a redesigned Aqua user interface. Once we realized the
> costs associated with doing that, we knew we could not support the
Macintosh
> platform profitably. Whether we like it or not, in this day and age,
> profitability is a requirement for which we are held accountable.
>
> Now for the good news. In subsequent discussions with Apple they have made
> us aware of a commercial X-Windows library for OS X. We are in the
process
> of evaluating it for use in a native Unix/X-Windows implementation of IDL
> and ENVI for the Mac OS X platform. This would solve many technical
issues
> for us and allow us to continue to support the Macintosh platform both
> natively and profitably, as it would leverage off our other Unix/X
> platforms. The only thing this does not accomplish is providing IDL with a
> new Aqua UI and widget set.
>
> While I cannot promise this will be successful and that we can take it to
> market, we are very encouraged by our early results working with the
> library. We hope to have some solid results upon which we can base a solid
> commitment to our customers in the very near future. We certainly
apologize
> for any inconvenience this may have caused you and your colleagues and ask
> that you give us an opportunity to complete our investigation and try to
> come up with a solution that will meet your needs and retain you as a
> satisfied customer of RSI.
Any comment from the community on whether this suggestion would meet the
needs of Mac OS X users? (That is, an OS X version of IDL that uses
X-Windows for direct/object graphics and user interface). Granted, you don't
get the fancy Aqua user interface, but do IDL programmers really care, as
long IDL is running on their Mac? For cross-platform development purposes, I
would think a more UNIX-like version for the Mac would be a good thing.
Cheers,
Liam.
Practical IDL Programming
http://www.gumley.com/
|
|
|
|
|
|
| Re: Message From RSI VP of Engineering [message #27345 is a reply to message #27225] |
Thu, 18 October 2001 07:43 |
John-David T. Smith
Messages: 384
Registered: January 2000
|
Senior Member |
|
|
Stein Vidar Hagfors Haugan wrote:
>
> JD Smith <jdsmith@astro.cornell.edu> writes:
>
>> "Bernard K." wrote:
>>>
>>> I am not familiar with command line IDL on Unix implementations. Could
>>> someone elaborate a bit more on what an Aqua-less IDL would probably
>>> look like? Just a command line tool that would ressemble using the
>>> actual command line in IDL for OS9? Apart from loosing the IDL
>>> developper environment and its associated menus (file edit search run
>>> project macros etc) what else would be missing?
>>
>> A Unix version for OSX would certainly not be command-line only. It
>> would probably even have an IDLDE... only it would be a Unix/X-Windows
>> program, so it would look more like a program running on Solaris or
>> Linux, and would otherwise be dissimilar from programs running under
>> Aqua.
>
> But on the other hand, you'd be able to take advantage of the best IDL
> development environment available anywhere, called idlwave, no??
I didn't want to browbeat, but, yes. Traditionally, the Mac market
would probably reject something with an Emacs interface, but then again,
wasn't I just arguing that the traditional Mac market is not what would
fuel IDL MacOSX sales?
As a side note, I long ago asked RSI developers if the OSX version being
developed would have a command-line only version available, and the
answer was yes. So, it could have been the best of both worlds.
JD
|
|
|
|
| Re: Message From RSI VP of Engineering [message #27347 is a reply to message #27225] |
Thu, 18 October 2001 06:56 |
Stein Vidar Hagfors H[1]
Messages: 56
Registered: February 2000
|
Member |
|
|
JD Smith <jdsmith@astro.cornell.edu> writes:
> "Bernard K." wrote:
>>
>> I am not familiar with command line IDL on Unix implementations. Could
>> someone elaborate a bit more on what an Aqua-less IDL would probably
>> look like? Just a command line tool that would ressemble using the
>> actual command line in IDL for OS9? Apart from loosing the IDL
>> developper environment and its associated menus (file edit search run
>> project macros etc) what else would be missing?
>
> A Unix version for OSX would certainly not be command-line only. It
> would probably even have an IDLDE... only it would be a Unix/X-Windows
> program, so it would look more like a program running on Solaris or
> Linux, and would otherwise be dissimilar from programs running under
> Aqua.
But on the other hand, you'd be able to take advantage of the best IDL
development environment available anywhere, called idlwave, no??
--
------------------------------------------------------------ --------------
Stein Vidar Hagfors Haugan
ESA SOHO SOC/European Space Agency Science Operations Coordinator for SOHO
NASA Goddard Space Flight Center, Email: shaugan@esa.nascom.nasa.gov
Mail Code 682.3, Bld. 26, Room G-1, Tel.: 1-301-286-9028/240-354-6066
Greenbelt, Maryland 20771, USA. Fax: 1-301-286-0264
------------------------------------------------------------ --------------
|
|
|
|
| Re: Message From RSI VP of Engineering [message #27361 is a reply to message #27307] |
Wed, 17 October 2001 22:38 |
Dennis Boccippio
Messages: 23
Registered: July 2000
|
Junior Member |
|
|
In article <9qiq2d$qjg$1@news.doit.wisc.edu>,
"Liam Gumley" <Liam.Gumley@ssec.wisc.edu> wrote:
> Any comment from the community on whether this suggestion would meet the
> needs of Mac OS X users? (That is, an OS X version of IDL that uses
> X-Windows for direct/object graphics and user interface). Granted, you don't
> get the fancy Aqua user interface, but do IDL programmers really care, as
> long IDL is running on their Mac? For cross-platform development purposes, I
> would think a more UNIX-like version for the Mac would be a good thing.
>
> Cheers,
> Liam.
> Practical IDL Programming
> http://www.gumley.com/
>
It'd be fine by me. I don't need Aqua buttons and interfaces, but I do
want IDL humming along doing batch jobs on my dual-800 G4 while I work
on something else, so I can finally retire my e$pensive SGI. Not to
mention a Powerbook license or two for field work....
- Dennis Boccippio
NASA/MSFC SD-60
|
|
|
|
|
|
| Re: Message From RSI VP of Engineering [message #27372 is a reply to message #27307] |
Wed, 17 October 2001 12:32 |
K. Bowman
Messages: 330
Registered: May 2000
|
Senior Member |
|
|
In article <9qiq2d$qjg$1@news.doit.wisc.edu>, Liam Gumley
<Liam.Gumley@ssec.wisc.edu> wrote:
> Any comment from the community on whether this suggestion would meet the
> needs of Mac OS X users?
This would work fine for me, especially since I don't use the
development environment. (I also like to think it was my idea ... ;-)
) I'm happy with BBEdit. BBEdit has good hooks into other development
tools. It would be nice if RSI could get Barebones to add IDL syntax
coloring to BBEdit and possibly breakpoints, automatic jump-to-error
features, etc.
A semi-native Mac version will require an X Windows server. We have
tried several under OS X. Our experience so far:
1. XFree86. Works, but to a Mac user feels very kludgy.
2. Xtools from Tenon. Very buggy. Not really usable, at least in the
release we tried a few months ago.
3. Exodus. New OS X version seems to work well. Provides Aqua window
manager (as well as Motif, etc.). I've used the OS 9 version for a
long time. OS 9 version had some serious bugs, but recent releases
have been stable and capable. Supports OpenGL, etc. We are planning
to go with this.
Ken Bowman
|
|
|
|
| Re: Message From RSI VP of Engineering [message #27409 is a reply to message #27307] |
Sat, 20 October 2001 19:56 |
Paul Woodford
Messages: 43
Registered: June 2000
|
Member |
|
|
In article <9qiq2d$qjg$1@news.doit.wisc.edu>,
"Liam Gumley" <Liam.Gumley@ssec.wisc.edu> wrote:
> Any comment from the community on whether this suggestion would meet
> the needs of Mac OS X users? (That is, an OS X version of IDL that
> uses X-Windows for direct/object graphics and user interface).
I haven't yet decided.
> Granted, you don't get the fancy Aqua user interface, but do IDL
> programmers really care, as long IDL is running on their Mac?
Yes. Or at least I do.
> For cross-platform development purposes, I would think a more
> UNIX-like version for the Mac would be a good thing.
No, it wouldn't. For cross-platform development purposes, RSI needs to
support the native interface of each system. In my opinion, they have
done a fairly good job of making sure widgets look alright on all the
platforms. IDLDE is another story. The Linux IDLDE has some annoying
bugs and is kind of clunky. I appreciate the power of idlwave, but I
greatly dislike emacs - the finger gymnastics, ugly GUI, etc.
The Macintosh IDLDE was obviously written by someone who cared about the
interface and understood Mac users. That's one of the reasons why RSI's
decision is so disappointing. The most persuasive reason for accepting
an ugly hybrid is to keep the camel's nose in the tent.
I'm holding onto my maintenance dollars for now.
Paul
|
|
|
|
| Re: Message From RSI VP of Engineering [message #27468 is a reply to message #27343] |
Tue, 23 October 2001 00:40 |
Struan Gray
Messages: 178
Registered: December 1995
|
Senior Member |
|
|
Paul Woodford, cpwoodford@spamcop.net writes:
> The Macintosh IDLDE was obviously written by
> someone who cared about the interface and
> understood Mac users. That's one of the
> reasons why RSI's decision is so disappointing.
I agree. IDL on the Mac has its quirks, but it didn't feel like
an ugly port, which is important for some of the people who use my
programs.
I've written my letter to RSI, re-hashing many of the points made
in this and other threads. I'm not a big user, but I have been an IDL
evangelist and I still think IDL has advantages over the mish-mash of
programs and code libraries I'll probably replace it with. I'm sad to
have to swap, but for me and my users cross-platform compatibility was
the only thing that made IDL truly stand out from its competitors.
Those competitors are actively courting me. Perhaps I just feel
extra warm and fuzzy because I'm on the IDL rebound, but significant
price reductions for academic users and help with porting from
IDL-trained technicians are not to be sniffed at.
> I'm holding onto my maintenance dollars for now.
Me too. I will use IDL 5.5 with pleasure, but active programming
is suspended for the duration, and the ION-based teaching project I
just finished the grant application for is headed for the dustbin. My
impression is that RSI is more interested in servicing its few large
customers than using people like me to kick-start interest in IDL
among new users. So be it.
Struan
|
|
|
|
| Re: Message From RSI VP of Engineering [message #27490 is a reply to message #27225] |
Wed, 24 October 2001 14:21 |
stevens
Messages: 7
Registered: July 1995
|
Junior Member |
|
|
In <9r73m1$bck$1@sulawesi-fi.lerc.nasa.gov>, Logan Lindquist:
[Apologies for a lot of intertwined snippage...]
> Interesting! I didn't know which Linux Release the X platform was based on.
> It took me a little while to figure out your *nix termenology. You are
> referring to the word, Linux?
Sorry; I was misleading and presumptive. Actually, "*nix" is kinda shorthand
for any variant/clone of Unix (Irix, AIX, Linux) and not just Linux alone.
As for X, it's not really Unix in that it developed alongside Unix, but from
an independent source ("Open Group" at one time, IIRC). It's just the GUI or
windowing system that grew up with the variants of Unix, and so the de facto
standard in that regard.
X and *nix ain't always pretty; it's ABM, and it gets the job done for me.
> Yes hopefully RSI will release IDL for X. :)
As I said elsewhere, for the realists, it's lemonade from lemons time. :)
> Not that the port of IDL will sell more new Apple stuff, but it will prevent
> IDL users who are not too pissed off and still want to use the language,
I dunno. Seems like a lot of IDL-Aqua or Bust sentiment lately. :)
Seriously, I think it would be a shame if RSI drops OS-X ports, but for very
different reasons than just: "RSI is greedy/stupid/incompetent/malicious" or
all of that. We do have Macroslop, afterall. :)
> populace, but maybe marketing doesn't matter too much when looking for any
> sales that would allow for development to continue].
This is my point: I'm an engineer. The marketeers can toast in Hades, for my
part. I want reliable, speedy, affordable visualization tools. Even if it is
back to Gnuplot for me, as Macroslop tries to jam XP, .NET and Passport down
our collective throats. There is a bigger issue here than our pet OS/GUI.
> Apple should provide the utility Development utilities to a company as large
> as Kodak/RSI for FREE.
Easy for you to say. Own any Kodak common stock? :)
> I would like to thank you for you intelligent, thoughtful response. Not like
> other people that use this group.
Give 'em a chance; there are a lot of very nice, helpful folks here. Really.
--
Regards, Weird (Harold Stevens) * IMPORTANT EMAIL INFO FOLLOWS *
Pardon the bogus email domain (dseg etc.) in place for spambots.
Stevens at sc45[.]rsc{period}raytheon(dot)com. *DO NOT SPAM IT.*
Standard Disclaimer: These are my opinions not Raytheon Company.
|
|
|
|
| Re: Message From RSI VP of Engineering [message #27491 is a reply to message #27225] |
Wed, 24 October 2001 13:46 |
John-David T. Smith
Messages: 384
Registered: January 2000
|
Senior Member |
|
|
"Liam E. Gumley" wrote:
>
> JD Smith wrote:
> [stuff deleted]
>> The bigger trouble lies under the hood. IDL for MacOSX had some
>> significant optimizations for display and within the core engine itself
>> which are being tossed out with the bath water. The display speed will
>> suffer, since in effect you're running through *two* levels of display
>> (the X level, which translates drawing commands into the native display
>> level). Any use of the much-improved OpenGL OS/hardware support will be
>> impossible. The powerful AltiVec tuning already accomplished or planned
>> for the OSX version will not be included.
>>
>> Here's a small sampling of a feature table comparison, far from
>> complete:
>>
>> +=========================================================== ======+
>> IDL feature comparison OSX Native OSX Straight Unix Port
>> +=========================================================== ======+
>> Interface Aqua X/Motif (server required)
>> Display Speed Fast Slow
>> 3D/OpenGL Optimization Yes No
>> Altivec Vectorization Complete None, or limited
>> Separate Core/IDE Threads Yes No
>> Pervasive PDF Output Yes No
>> +=========================================================== ======+
>
> Please correct me if I'm wrong, but I don't see why the display speed
> would suffer.
I'll put a fine point on it: running the RSI-supplied graphics_times3
benchmark on a native OSX vs. a X11-based IDL would reveal the former to
be much faster than the latter. This is a direct result of layering two
display devices one atop another (which is different from your SGI, for
which X11 is the native drawing layer).
You could demonstrate this to yourself quite convincingly by running
graphics_times3 in your IDL version running in an X-emulator under
Windows, and on the Windows version directly. I think you'll find the
latter to be a good deal faster.
This may not be a *practical* limit for what you do, but certainly could
impact others with more display-taxing applications.
A similar story could be told for core routine performance and lack of
Altivec tuning. Unless RSI is hiding a miracle up their sleeve, "IDL
OSX--" will be noticeably slower in both display and computation than
the aborted IDL OSX. Of course, we may never know the difference.
JD
|
|
|
|
|
|
| Re: Message From RSI VP of Engineering [message #27494 is a reply to message #27225] |
Wed, 24 October 2001 13:01 |
Liam E. Gumley
Messages: 378
Registered: January 2000
|
Senior Member |
|
|
JD Smith wrote:
[stuff deleted]
> The bigger trouble lies under the hood. IDL for MacOSX had some
> significant optimizations for display and within the core engine itself
> which are being tossed out with the bath water. The display speed will
> suffer, since in effect you're running through *two* levels of display
> (the X level, which translates drawing commands into the native display
> level). Any use of the much-improved OpenGL OS/hardware support will be
> impossible. The powerful AltiVec tuning already accomplished or planned
> for the OSX version will not be included.
>
> Here's a small sampling of a feature table comparison, far from
> complete:
>
> +=========================================================== ======+
> IDL feature comparison OSX Native OSX Straight Unix Port
> +=========================================================== ======+
> Interface Aqua X/Motif (server required)
> Display Speed Fast Slow
> 3D/OpenGL Optimization Yes No
> Altivec Vectorization Complete None, or limited
> Separate Core/IDE Threads Yes No
> Pervasive PDF Output Yes No
> +=========================================================== ======+
Please correct me if I'm wrong, but I don't see why the display speed
would suffer.
If I run IDL locally on an SGI console then it's the same issue, right?
That is, IDL direct graphics are translated from the X level into the
native display level. I've never had any complaints with display speed
in this mode. In fact, most of the time I run IDL on a remote UNIX host
while sitting in front of a PC running Exceed, and even then I don't
have major complaints about the display speed of direct graphics (which
admittedly is network bandwidth limited).
Cheers,
Liam.
Practical IDL Programming
http://www.gumley.com/
|
|
|
|
| Re: Message From RSI VP of Engineering [message #27495 is a reply to message #27225] |
Wed, 24 October 2001 13:00 |
stevens
Messages: 7
Registered: July 1995
|
Junior Member |
|
|
In <3BD70EEA.ECF66A21@astro.cornell.edu>, JD Smith:
[Snip...]
> I really do like the analogy made by Ron Syml: how would you Windows
> users
Correction: I do not use a single piece of Macroslop ever, with the single
exception of my yearly review. The pointyhairs INSIST it be done with that
satanic abomination called MS-Word, and so I oblige them, only as it seems
a condition of my employment. I'm composing this on an NCD Xterm with slrn
on a Sun Enterprise, manage the department's Turbolinux webserver, and use
a SuSE Linux PC on the table next to me. Windows user? Hah! NEVER! :)
> like it if RSI told you they didn't want to support your system,
> and the "solution" was to run an X server with IDL for Unix; how would
> Unix users like it if they were dropped and told to run a bundled
> Windows emulation program on top of IDL for Windows?
As long as I have some options other than Windows? Like Linux or Mac? Then
I don't care. Really. I go find the tools to get the job done, and (as the
point above) select them whenever possible mostly on the basis of personal
utility and a sense of principles. It involves very little misplaced ardor
or cowed unthinking acceptance, I believe. Yearly review excepted. :)
That is NOT to say I like the RSI headfake recently on IDL OS-X ports, and
I actually looked forward to the Aqua port. Seriously.
But that's just not gonna happen, so I gotta make lemonade from lemons.
> Somehow, I doubt
> you'd chalk it up to "it's just another GUI, what do I care?"
Actually, yes, unless *speed* is absolutely crippled. Then I get bent. I'm
a radar simulations kinda guy, not a wiredup eyecandy goober. If RSI ports
IDL to OS-X using something X, I'm won't throw baby out with bathwater, on
the sole basis of a disputed pedigree.
[Snip...]
--
Regards, Weird (Harold Stevens) * IMPORTANT EMAIL INFO FOLLOWS *
Pardon the bogus email domain (dseg etc.) in place for spambots.
Stevens at sc45[.]rsc{period}raytheon(dot)com. *DO NOT SPAM IT.*
Standard Disclaimer: These are my opinions not Raytheon Company.
|
|
|
|
|
|
| Re: Message From RSI VP of Engineering [message #27498 is a reply to message #27225] |
Wed, 24 October 2001 12:16 |
Logan Lindquist
Messages: 50
Registered: October 2001
|
Member |
|
|
Harold,
> Not Linux...FreeBSD. Both *nix, of course, but ports 'twixt *nix is not
> a walk in the park, generally. Then there's the whole zealotry thing in
Interesting! I didn't know which Linux Release the X platform was based on.
It took me a little while to figure out your *nix termenology. You are
referring to the word, Linux?
> Speaking for myself Aqua's just another GUI. If RSI would rather modify
> standard X toolkits and build a low(er)cost IDL port, go for it. I will
Yes hopefully RSI will release IDL for X. :)
> There's two sets of cashflow interest here: RSI, and Apple. But there's
Not that the port of IDL will sell more new Apple stuff, but it will prevent
IDL users who are not too pissed off and still want to use the language,
from buying x86. RSI has to figure out cost of development divided by the
project sales to X users [targeted at the correct market, not the general
populace, but maybe marketing doesn't matter too much when looking for any
sales that would allow for development to continue].
> If it won't include Aqua (i.e., Apple proprietary) there's hardly a way
> Apple can make money. Gnu is mainly about choice, and only peripherally
Apple should provide the utility Development utilities to a company as large
as Kodak/RSI for FREE. Also look to my paragraph above for further
motivation for Apple to keep its users from investing in x86.
I would like to thank you for you intelligent, thoughtful response. Not like
other people that use this group.
Logan Lindquist
|
|
|
|
| Re: Message From RSI VP of Engineering [message #27501 is a reply to message #27225] |
Wed, 24 October 2001 11:56 |
John-David T. Smith
Messages: 384
Registered: January 2000
|
Senior Member |
|
|
"Harold Stevens US.972.952.3293" wrote:
>
> In <9r6ob0$6ot$1@sulawesi-fi.lerc.nasa.gov>, Logan Lindquist:
>
> [Snip...]
>
>> I do not know the details of porting stuff that has been written for Linux
>> over to OS X but I wouldn't image that they would be much different, since
>> OS X is based on a Linux kernel.
>
> Not Linux...FreeBSD. Both *nix, of course, but ports 'twixt *nix is not
> a walk in the park, generally. Then there's the whole zealotry thing in
> porting GUIs between *nixen and their distribution sets. For a while it
> seemed as if some Linux diehards would go to war over Gnome vs. KDE for
> crying out loud. Enough of this GUI jihad especially proprietary; where
> the heck is the science and visualization stuff?
>
> Actually, these kernels don't need any GUI (even *nix's X) to operate.
>
> Now, it's true there is considerable experience porting X among *nix.
>
> Speaking for myself Aqua's just another GUI. If RSI would rather modify
> standard X toolkits and build a low(er)cost IDL port, go for it. I will
> use any X interface they might finally offer mainly since beggars can't
> be choosers. Having cut my teeth on Linux in a world infested with dumb
> ideas like proprietary Winmodems, I'm happy when any vendor makes these
> efforts to port functionality to "niche" environments like OS-X.
I think the most important side-effects of dropping the native OSX
version are being completely missed here. I am perfectly comfortable
with the X/Motif widgets, which I use everyday... especially since I
don't use IDLDE. I suspect most IDL users on OSX would not balk
terribly at the foreign-looking interface (inferior as it may be).
The bigger trouble lies under the hood. IDL for MacOSX had some
significant optimizations for display and within the core engine itself
which are being tossed out with the bath water. The display speed will
suffer, since in effect you're running through *two* levels of display
(the X level, which translates drawing commands into the native display
level). Any use of the much-improved OpenGL OS/hardware support will be
impossible. The powerful AltiVec tuning already accomplished or planned
for the OSX version will not be included.
Here's a small sampling of a feature table comparison, far from
complete:
+=========================================================== ======+
IDL feature comparison OSX Native OSX Straight Unix Port
+=========================================================== ======+
Interface Aqua X/Motif (server required)
Display Speed Fast Slow
3D/OpenGL Optimization Yes No
Altivec Vectorization Complete None, or limited
Separate Core/IDE Threads Yes No
Pervasive PDF Output Yes No
+=========================================================== ======+
The bottom line? If you don't care about a non-native interface, your
worries aren't over. I welcome corrections to this table (especially by
RSI).
I really do like the analogy made by Ron Syml: how would you Windows
users like it if RSI told you they didn't want to support your system,
and the "solution" was to run an X server with IDL for Unix; how would
Unix users like it if they were dropped and told to run a bundled
Windows emulation program on top of IDL for Windows? Somehow, I doubt
you'd chalk it up to "it's just another GUI, what do I care?"
JD
|
|
|
|
| Re: Message From RSI VP of Engineering [message #27505 is a reply to message #27225] |
Wed, 24 October 2001 11:02 |
stevens
Messages: 7
Registered: July 1995
|
Junior Member |
|
|
In <9r6ob0$6ot$1@sulawesi-fi.lerc.nasa.gov>, Logan Lindquist:
[Snip...]
> I do not know the details of porting stuff that has been written for Linux
> over to OS X but I wouldn't image that they would be much different, since
> OS X is based on a Linux kernel.
Not Linux...FreeBSD. Both *nix, of course, but ports 'twixt *nix is not
a walk in the park, generally. Then there's the whole zealotry thing in
porting GUIs between *nixen and their distribution sets. For a while it
seemed as if some Linux diehards would go to war over Gnome vs. KDE for
crying out loud. Enough of this GUI jihad especially proprietary; where
the heck is the science and visualization stuff?
Actually, these kernels don't need any GUI (even *nix's X) to operate.
Now, it's true there is considerable experience porting X among *nix.
Speaking for myself Aqua's just another GUI. If RSI would rather modify
standard X toolkits and build a low(er)cost IDL port, go for it. I will
use any X interface they might finally offer mainly since beggars can't
be choosers. Having cut my teeth on Linux in a world infested with dumb
ideas like proprietary Winmodems, I'm happy when any vendor makes these
efforts to port functionality to "niche" environments like OS-X.
[Snip...]
> seems to be a preliminary one. Business people will change their mind if it
> is deemed profitable for the company.
There's two sets of cashflow interest here: RSI, and Apple. But there's
not a lot RSI can do in pitching itself as a Gnu charity to Apple. They
are on their own here, and Apple certainly has bigger fish to fry.
If it won't include Aqua (i.e., Apple proprietary) there's hardly a way
Apple can make money. Gnu is mainly about choice, and only peripherally
somebody's cashflow problems. Then there's the whole FUD scene in which
Macroslop screams bloody murder about Gnu "contamination" when all that
proprietary codebase (like a lot of Apple) "comingles" with Gnu. That's
enough reason to keep legions of attack lawyers licking their chops, if
only in the service of ignorant, unrestrained avarice. Get the pix?
Stallman might say, Gnu is not here to make Apple and RSI (etc.) money.
> About the pricing. If everyone would remember back to economics, the
> quantity/demand curves and the price/cost curves will give us some useful
> tool to analyze their decisions.
[Snip...]
Then please consider X toolkits and Gnu may not be Aqua, but just about
the only way RSI can afford IDL-anything for OS-X (or Macs, period) and
"keep the dream alive" as some might say.
--
Regards, Weird (Harold Stevens) * IMPORTANT EMAIL INFO FOLLOWS *
Pardon the bogus email domain (dseg etc.) in place for spambots.
Stevens at sc45[.]rsc{period}raytheon(dot)com. *DO NOT SPAM IT.*
Standard Disclaimer: These are my opinions not Raytheon Company.
|
|
|
|
| Re: Message From RSI VP of Engineering [message #27511 is a reply to message #27298] |
Wed, 24 October 2001 09:02 |
Logan Lindquist
Messages: 50
Registered: October 2001
|
Member |
|
|
"Wolf Schweitzer" <wuff@swisswuff.ch> wrote in message
news:3BD678DC.9050505@swisswuff.ch...
> To me, there is a misconception that some people assume we are using
> Macs because of their "cutsy" interface and that can be "over" now.
The graphic user interface between different operating systems is only a
tool used to created by different companies to allow you to perform work and
create things using the hardware that is available to that particular
system. The major debate over the GUI's does not really matter when the
hardware aspect is not considered. THE CAUSE OF THE MAC OS VS PC OS DEBATE
IS BASED SOLELY ON WHAT GUI YOU ARE MOST FAMILAR WITH. That is all that
should be considered. How quickly can you get the work you need to get done
on a particluar operating system is based on previous amount of time spent
using that operating system IN ADDITION TO WHAT HARDWARE YOU MACHINE HAS.
> An important advantage for IDL on Macintosh is its ability to do
> parameter-passing with Applescript. In order to understand the
> usefulness of that you need to know what other applications also do
> Applescript on a Mac OS.
This is a true statement, as you can see...
*****
IDL 5.5 Functional Summary
Development & Programming Tools
Macintosh AppleScript support
*****
This is also supported under a windows enviroment.
******
IDL 5.5 Functional Summary
Development & Programming Tools
Callable Windows DLL
ActiveX control (dual interface)
******
> As Applescript would not sell without Mac OS and we are all happy it's
> also part of Mac OS X, I think that IDL would need to be shipped with
> Mac OS X - it is just an essential ingredient for the scientific
> Macintosh community.
I do not know the details of porting stuff that has been written for Linux
over to OS X but I wouldn't image that they would be much different, since
OS X is based on a Linux kernel. IDL already supports Linux on Alpha's and
x86. So the real question is if they already have a compilation that is
somewhat similar, and there is enough support to figure out the details of
porting the x86 Linux or the Alpha Linux over to OS X, why not start an open
sourced development of such?
I just went and reviewed what the VP of RSI said about this issue.
"> > Now for the good news. In subsequent discussions with Apple they have
made
>> us aware of a commercial X-Windows library for OS X. We are in the
> process
>> of evaluating it for use in a native Unix/X-Windows implementation of
IDL
>> and ENVI for the Mac OS X platform. This would solve many technical
> issues
>> for us and allow us to continue to support the Macintosh platform both
>> natively and profitably, as it would leverage off our other Unix/X
>> platforms. The only thing this does not accomplish is providing IDL with
a
>> new Aqua UI and widget set."
So it looks like the big gripe that many of you have is unfounded, because
RSI can't afford to pay a graphic artist/computer scientist to redesign the
GUI of IDL OR they don't want to change the look because of IDL is a
professional product. Thus the redesign would make the interface less
professional looking. At least they are considering porting to OS X. It all
then comes down to a usability issue. OS X users would have to get used to a
slightly different interface. I suggest stop complaining and wait to see
they decide to support it.
Their business people are looking at the financial information related to
how many Mac users buy or renew licenses each year. The decision so far
seems to be a preliminary one. Business people will change their mind if it
is deemed profitable for the company.
About the pricing. If everyone would remember back to economics, the
quantity/demand curves and the price/cost curves will give us some useful
tool to analyze their decisions.
Since it's software the quantity/demand curve operate a bit differently.
Easy to produce once the code exists. The amount of demand gives us some
idea of why it costs so much. I agree that they should reconsider their
pricing structure for educational software[increase demand - lower prices],
but I also think they are doing a good job of targeting the specific group
of students that is mostly likely to use the language in the business
environment. When compared to mathematical programs such as MatLab or
Mathmatica, I think that those are better targeted towards Math majors. I
haven't used either extensively. The trick is to get future users to become
familiar with the language. You do not do this by limiting the number of
copies that an institution can buy. These future users will hopefully equal
future dollars spent once they graduate and get a job. I am an example! It
would be helpful if someone who actually bought an educational version to
contribute to the price range we are talking about.
That's all I have to say for now. It's lunch time and I'm hungry,
Logan Lindquist
|
|
|
|
| Re: Message From RSI VP of Engineering [message #27611 is a reply to message #27225] |
Thu, 25 October 2001 09:11 |
Pavel A. Romashkin
Messages: 531
Registered: November 2000
|
Senior Member |
|
|
I am sure you will miss the DE that you are used to. I can't stand the
DE under Unix, sorry. If the performance will be what it is like on our
HP-UX, I will not even try the OSX version.
Pavel
Ben Tupper wrote:
>
> JD
>
> Thanks for assembling this comparison. This is the kind of meat-and-potatoes
> information I have been lacking. The numerical performance issue is a weighty
> one; that's not to diminish the importance of the native interface and display
> speed/rendering. I don't want to ask you to compare apples and oranges, but (I
> will anyway) how do you think the numerical performance of a Unix port of IDL to
> OS X will compare to that we currently see in IDL on the G4 under OS 9? When we
> switched from Unix IDL to Mac, I was blown away by the performance increase...
> will I be blown 'back' to the slower performance?
>
> Ben
|
|
|
|
| Re: Message From RSI VP of Engineering [message #27612 is a reply to message #27225] |
Thu, 25 October 2001 08:55 |
Dennis Boccippio
Messages: 23
Registered: July 2000
|
Junior Member |
|
|
In article <3BD826E9.F60CB501@bigelow.org>,
Ben Tupper <btupper@bigelow.org> wrote:
> JD
>
> Thanks for assembling this comparison. This is the kind of meat-and-potatoes
> information I have been lacking. The numerical performance issue is a
> weighty
> one; that's not to diminish the importance of the native interface and
> display
> speed/rendering. I don't want to ask you to compare apples and oranges, but
> (I
> will anyway) how do you think the numerical performance of a Unix port of IDL
> to
> OS X will compare to that we currently see in IDL on the G4 under OS 9? When
> we
> switched from Unix IDL to Mac, I was blown away by the performance
> increase...
> will I be blown 'back' to the slower performance?
>
> Ben
>
Numerical performance aside, from the 'beggars can't be choosers' front,
compare the following:
+=========================================================== ============+
large jobs monopolize CPU
-------------------------
OS-9: almost entirely (despite decent RSI and MacOS progress)
OS-X: almost negligibly
+=========================================================== ============+
+=========================================================== ============+
CLI/shells allow efficient management/preprocessing of data files
------------------------------------------------------------ ---
OS-9: difficult (Applescript or 3rd party shell tools required)
OS-X: trivial (and flexible) Un*x
+=========================================================== ============+
These two issues alone are the reason my MacOS license is used only for
short/interactive jobs and code development, while my meaty jobs are run
on IRIX licenses. They have nothing to do with OS-X nativization of the
IDL code itself.
The short form: what I *want* is:
- The ability to store, manage and preprocess many large data files
*inexpensively* on one machine. OS-X/shells with banks of Firewire
drives give me this.
(other Un*x solutions give me shells but storage and maintenance are
expensive, Linux aside)
- The ability to develop analysis code efficiently and comfortably.
IDL/OS-X with the DE and profiler give me this.
(any IDL platform gives me this)
- The ability to batch process these data files without dragging the
machine to a halt. IDL/OS-X gives me this.
(any IDL platform but OS-9 gives me this. [Windoze?])
- The ability to postprocess the results for visualization and sharing
efficiently on the same machine. (For me, Noesys, Quicktime,
MediaCleaner, Acrobat and/or <gasp> even iMovie/iDVD give me this)
(no other platform gives me this functionality the way I want it)
- The ability to create presentations, slides, papers, multimedia demos,
etc on the same machine using the tools I want (Director, Powerpoint,
TeX, Acrobat, Illustrator, etc.)
(Windoze could give me this, I guess.)
- The ability to serve out results from the same machine using standard/
"safe" tools (like Apache) and access them from home (like ssh)
(any Un*x solution could give me this)
Yes, I know, I can (and do) do all this by using multiple platforms. I
can even do some of this less effectively using a variety of shareware
or GNU tools on one platform (appease the Linux folks). I've tried that
option and it always comes up short; the tools just aren't up to snuff
compared with commercial solutions in the Mac/Windoze world. None of my
needs (not necessarily representative, granted) require significant
G4/OS-X/Aqua nativization of IDL. Would it be nice, yes, but critical,
no. YMMV. I'm almost certain that any cost or productivity losses
incurrent by non-G4-optimized or non-Aquafied OS-X/IDL would be more
than offset by consolidating my current multi-platform approach to a
single-platform solution...
- Dennis Boccippio, NASA/MSFC SD-60
|
|
|
|
| Re: Message From RSI VP of Engineering [message #27619 is a reply to message #27491] |
Thu, 25 October 2001 07:51 |
btt
Messages: 345
Registered: December 2000
|
Senior Member |
|
|
JD
Thanks for assembling this comparison. This is the kind of meat-and-potatoes
information I have been lacking. The numerical performance issue is a weighty
one; that's not to diminish the importance of the native interface and display
speed/rendering. I don't want to ask you to compare apples and oranges, but (I
will anyway) how do you think the numerical performance of a Unix port of IDL to
OS X will compare to that we currently see in IDL on the G4 under OS 9? When we
switched from Unix IDL to Mac, I was blown away by the performance increase...
will I be blown 'back' to the slower performance?
Ben
JD Smith wrote:
> "Liam E. Gumley" wrote:
>>
>> JD Smith wrote:
>> [stuff deleted]
>>> The bigger trouble lies under the hood. IDL for MacOSX had some
>>> significant optimizations for display and within the core engine itself
>>> which are being tossed out with the bath water. The display speed will
>>> suffer, since in effect you're running through *two* levels of display
>>> (the X level, which translates drawing commands into the native display
>>> level). Any use of the much-improved OpenGL OS/hardware support will be
>>> impossible. The powerful AltiVec tuning already accomplished or planned
>>> for the OSX version will not be included.
>>>
>>> Here's a small sampling of a feature table comparison, far from
>>> complete:
>>>
>>> +=========================================================== ======+
>>> IDL feature comparison OSX Native OSX Straight Unix Port
>>> +=========================================================== ======+
>>> Interface Aqua X/Motif (server required)
>>> Display Speed Fast Slow
>>> 3D/OpenGL Optimization Yes No
>>> Altivec Vectorization Complete None, or limited
>>> Separate Core/IDE Threads Yes No
>>> Pervasive PDF Output Yes No
>>> +=========================================================== ======+
>>
>> Please correct me if I'm wrong, but I don't see why the display speed
>> would suffer.
>
> I'll put a fine point on it: running the RSI-supplied graphics_times3
> benchmark on a native OSX vs. a X11-based IDL would reveal the former to
> be much faster than the latter. This is a direct result of layering two
> display devices one atop another (which is different from your SGI, for
> which X11 is the native drawing layer).
>
> You could demonstrate this to yourself quite convincingly by running
> graphics_times3 in your IDL version running in an X-emulator under
> Windows, and on the Windows version directly. I think you'll find the
> latter to be a good deal faster.
>
> This may not be a *practical* limit for what you do, but certainly could
> impact others with more display-taxing applications.
>
> A similar story could be told for core routine performance and lack of
> Altivec tuning. Unless RSI is hiding a miracle up their sleeve, "IDL
> OSX--" will be noticeably slower in both display and computation than
> the aborted IDL OSX. Of course, we may never know the difference.
>
> JD
--
Ben Tupper
180 Mckown Point Road
West Boothbay Harbor, ME 04575
email: btupper@bigelow.org
telephone: (207) 633-9600
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
