| Re: IDL, GDL, copyright, EULAs and such [message #44687] |
Thu, 07 July 2005 00:41  |
Chris Lee
Messages: 101
Registered: August 2003
|
Senior Member |
|
|
In article <1120669967.155972.166240@g14g2000cwa.googlegroups.com>, "Y.T."
<ytyourclothes@p.zapto.org> wrote:
> ....
> essential tool like "linfit" is unavailable. So I copy it from my
There's a free version of linfit at
http://www.astro.washington.edu/deutsch-bin/getpro/library37 .html?LINFIT
it's not the same as the IDL version, so you'd need to work it into a
function, and add the keywords that the IDL version makes available
> an open-ended question -- can I continue to use basic, trivial functions
> like "poly.pro" or "factorial.pro" with GDL after IDL has expired? How
factorial is written (I know, I wrote it :) It should be fully compatible
with the IDL version. It, and a few other library functions live in
the gdl/src/pro directory, in case you haven't found them.
POLY is something like
function poly,x,c
nc=n_elements(c)
p=0*x
for i=nc-1,1L,-1 do p=(p+c[i])*x
p=p+c[0]
return p
end
Chris.
|
|
|
|
| Re: IDL, GDL, copyright, EULAs and such [message #44689 is a reply to message #44687] |
Wed, 06 July 2005 17:21  |
Ken Mankoff
Messages: 158
Registered: February 2000
|
Senior Member |
|
|
On Wed, 6 Jul 2005, Y.T. wrote:
> ; Copyright (c) 1994-2001, Research Systems, Inc. All rights reserved.
> ; Unauthorized reproduction prohibited.
I wrote this newsgroup on April 8 of this year asking the exact same
question. The thread is here:
http://groups-beta.google.com/group/comp.lang.idl-pvwave/bro wse_thread/thread/c7427292e352938c/9df63979178d7896?q=mankof f&rnum=27#9df63979178d7896
I also sent the question to RSI and got a reply from them:
> FYI, Just in case anyone else has a similar question:
>
> I got a reply from RSI. They said that as an RSI customer with a
> valid/current license, pretty much anything I do with their demo
> is authorized.
While this does not answer your question, their wording implies that
once your license expires you do not maintain the rights to their
libraries.
But I think if you look around
http://www.astro.washington.edu/deutsch/idl/htmlhelp/
You'll find that almost all the IDL code has been duplicated by
someone else somewhere else. You are probably allowed to use that.
And the above site has posted the IDL 4.0.1 libraries on the web for
anyone to see...
-k.
--
http://spacebit.dyndns.org/
|
|
|
|
| Re: IDL, GDL, copyright, EULAs and such [message #44690 is a reply to message #44689] |
Wed, 06 July 2005 15:19 |
Paul Van Delst[1]
Messages: 1157
Registered: April 2002
|
Senior Member |
|
|
Y.T. wrote:
> A linear fit, for example, is something I've written back on my
> Atari-800: it's not exactly a patentable algorithm. NR published a
> fortran version of the algorithm. RSI implemented it in IDL, cited NR
> in the header as a source and then slapped the following two lines on
> it:
>
> ; Copyright (c) 1994-2001, Research Systems, Inc. All rights reserved.
> ; Unauthorized reproduction prohibited.
I think the business end of the above two lines is
"Unauthorized reproduction prohibited"
That's a pretty generic statement and one could make the argument that if you have the source code
on a machine that doesn't contain an IDL license you have reproduced it with no authorization from
RSI. Why not ask them? At least then you'll no longer have to rationalise -- you'll know what the
company line is.
From the "Installing and Licensing IDL 6.0" pdf:
<quote>
The IDL�, ION Script", and ION Java" software programs and the accompanying procedures, functions,
and documentation described herein are sold under license agreement. Their use, duplication, and
disclosure are subject to the restrictions stated in the license agreement. Research Systems, Inc.,
reserves the right to make changes to this document at any time and without notice.
</quote>
Note the terms "accompanying procedures, functions, and documentation". That seems pretty clear to
me (although I couldn't find a copy of the license agreement itself.)
> But if I were to implement my linear fit by myself from scratch, it
> would look very much like RSI's, of course. Especially now that I've
> seen their code. How many different ways are there, really, to do this?
But would you do it in the IDL/GDL language? Why not link in a precompiled library and use a GDL
wrapper that has the same interface as the IDL version? It'll probably be faster and there are scads
of available libraries/source codes to pick from.... just visit www.netlib.org. Bypass the IDL
library code altogether.
> Then again, maybe there's no problem with the library code as long as I
> didn't use their actual IDL binary(?)
That's not how I read the copyright statement above.
> Ah, the vagaries of writing code in the 21st century...
But, you wouldn't be writing the code. You'd be copying it. (And RSI might equate "copying" with
"stealing" -- especially when the IDL product ship quantities start to drop..... :o)
paulv
--
Paul van Delst
CIMSS @ NOAA/NCEP/EMC
|
|
|
|
| Re: IDL, GDL, copyright, EULAs and such [message #44691 is a reply to message #44690] |
Wed, 06 July 2005 14:02 |
Edd Edmondson
Messages: 50
Registered: January 2003
|
Member |
|
|
Haje Korth <haje.korth@jhuapl.edu> wrote:
> I don't get the point: Why use GDL if you already have a licensed copy of
> IDL. I see GDL as an alternative solutions for people who cannot cough up
> the $$$ for the in my opinion overprized IDL license fees. Since you already
> own IDL, just ignore the fact that it does way more than your brain can
> handle and live happily ever after. :-)
That is not the point of the G in GDL, even if the project was created
because some group of people could not afford the $$$.
I could sell GDL for more money than IDL sells for, although I'd
likely not sell many copies as anyone else able to get their hands on
it, including my own customers could resell it at any price. It is
that which makes GDL inexpensive above anything else, and that's a
side effect of the real purpose of GPL'd software.
The real strength of GDL being free is that I can change it. I could
fix any bugs in it, and I could add features as I wish and do almost
anything with it I choose. So, whilst there may not currently be
reasons to use GDL over IDL if you own the latter, that isn't
necessarily true in the future, and indeed in rare cases may not even
be true now.
--
Edd
|
|
|
|
| Re: IDL, GDL, copyright, EULAs and such [message #44692 is a reply to message #44691] |
Wed, 06 July 2005 13:45 |
Y.T.
Messages: 25
Registered: December 2004
|
Junior Member |
|
|
Haje Korth wrote:
> I don't get the point: Why use GDL if you already have a licensed copy of
> IDL. I see GDL as an alternative solutions for people who cannot cough up
> the $$$ for the in my opinion overprized IDL license fees.
Exactly. :)
This includes people who have a valid license now which is bound to
expire at some point in the near future; people who are about to change
employers and aren't sure the new employer is going to cough up the
dough for IDL; and people who have no chance of being able to afford a
license privately for themselves but would like to tinker with data at
home. And even people who are tired of having to convince their IT
people that they should pay money every couple years to renew their IDL
license "when you can do all this in PAW"...
> PS: You could ask a lawyer to check into that for you, but they are
> overprized too and the IDL license may seem cheap compared to your legal
> fees.
I'm not really under the impression that lawyers understand the online
intellectual property jungle any better than laymen.
At this point I wasn't really looking for an expensive certifiable
answer anyways - I was trying to get a bit of a reading on how real
people feel about these things.
A linear fit, for example, is something I've written back on my
Atari-800: it's not exactly a patentable algorithm. NR published a
fortran version of the algorithm. RSI implemented it in IDL, cited NR
in the header as a source and then slapped the following two lines on
it:
; Copyright (c) 1994-2001, Research Systems, Inc. All rights reserved.
; Unauthorized reproduction prohibited.
But what does that really mean? I can't hand this to someone else?
Can't make it public? Can't copy it to a subdirectory of my own to use
in GDL? What's actually covered under "fair use" these days?
As long as nobody looks at my harddisk, obviously there's no problem
here anywhere. But lets say I tinker with GDL and write some kind of
groundbreaking stock-trading software or some kind of super-accurate
prediction code for earthquakes -- something that makes me rich enough
to become a worthy target for litigation and famous enough for all the
litigators to know that I exist. At that point I'd rather not have RSI
thugs kick down my door and demand all my money because I did my linear
fits with a routine that had their copyright in the header.
And if I then say "I was using a licensed copy of IDL on my employer's
computer", my employer is going to demand my money instead.
But if I were to implement my linear fit by myself from scratch, it
would look very much like RSI's, of course. Especially now that I've
seen their code. How many different ways are there, really, to do this?
Then again, maybe there's no problem with the library code as long as I
didn't use their actual IDL binary(?)
Ah, the vagaries of writing code in the 21st century...
cordially
Y.T.
--
Remove YourClothes before you email me.
|
|
|
|
| Re: IDL, GDL, copyright, EULAs and such [message #44693 is a reply to message #44692] |
Wed, 06 July 2005 12:50 |
Haje Korth
Messages: 651
Registered: May 1997
|
Senior Member |
|
|
I don't get the point: Why use GDL if you already have a licensed copy of
IDL. I see GDL as an alternative solutions for people who cannot cough up
the $$$ for the in my opinion overprized IDL license fees. Since you already
own IDL, just ignore the fact that it does way more than your brain can
handle and live happily ever after. :-)
Haje
PS: You could ask a lawyer to check into that for you, but they are
overprized too and the IDL license may seem cheap compared to your legal
fees.
"Y.T." <ytyourclothes@p.zapto.org> wrote in message
news:1120669967.155972.166240@g14g2000cwa.googlegroups.com.. .
> So I've been playing around with GDL, simply because I like IDL. I like
> GDL's focus on the essentials - I wouldn't mind widget-tools, but I
> consider them a luxury, really. And if GDL never acquires the
> incomprehensible object junk of IDL, it'll be too soon.
>
> However much of the power of IDL lies in the library, of course - and
> thus GDL's weakness lies in the lack of that library.
>
> For now, I figure, there shouldn't be a problem with my re-using the
> routines from my valid, legal IDL installation -- it is just a massive
> pain when such an essential tool like "linfit" is unavailable. So I
> copy it from my /usr/local/rsi to a local GDL-directory. No big deal.
>
> Or is it a big deal? The routine is copyrighted by RSI, no? Am I
> allowed to use it on a program like GDL that is quite blatantly
> intended to be a free replacement for IDL? Can I use my idl/lib
> routines as long as I have a functioning IDL installation? Do I have to
> stop using them as soon as my license expires? But the license is only
> for IDL, no? That's why the hasp-thingee only protects use of the
> binary, right? So the library should be considered "acquired" and still
> be allowed to be used with GDL even in the absence of a working IDL.
> No? Yes?
>
> The longer I think about it the less sure I am that running GDL might
> not be in violation of the "reverse engineering" clause in the IDL
> license agreement or some such -- except that that would only apply to
> people who actually have a valid installation of IDL since only those
> would ever have agreed to that license.
>
> And the IDL license actually never spells out what the "software"
> really is that is being licensed -- whether it is IDL itself or also
> the contents of idl/lib/*
>
> Since I'm already pretty confused by the whole notion of "intellectual
> property" (and much more so on the 'net) I figure I'll throw this out
> as an open-ended question -- can I continue to use basic, trivial
> functions like "poly.pro" or "factorial.pro" with GDL after IDL has
> expired? How about more complex stuff? Why or why not? What's the
> status of the routines that were originally from NR?
>
> I'm just baffled.
>
>
> cordially
>
>
> Y.T.
>
> --
> Remove YourClothes before you email me.
>
|
|
|
|
| Re: IDL, GDL, copyright, EULAs and such [message #44698 is a reply to message #44689] |
Wed, 13 July 2005 10:40 |
Y.T.
Messages: 25
Registered: December 2004
|
Junior Member |
|
|
> But I think if you look around
> http://www.astro.washington.ed u/deutsch/idl/htmlhelp/
>
> You'll find that almost all the IDL code has been duplicated by
> someone else somewhere else. You are probably allowed to use that.
Maybe someone needs to start collecting these things and compile them
into a "free idl-alike .pro library". The value of something like GDL
would really increase drastically through this.
And earlier, Christopher Lee wrote:
> There's a free version of linfit at
> http://www.astro.washington.ed u/deutsch-bin/getpro/library37 .html?LINFIT
>
> it's not the same as the IDL version, so you'd need to work it into a
> function, and add the keywords that the IDL version makes available
Not to sound ungrateful, but that's kinda pointless. I can write my own
version of linfit (and everything else in the IDL library). If I have
to write scads of code to make a trivial algorithm work, then I might
as well implement the whole thing myself.
The point was that there's already an existing library that has it all
inside - the IDL library. Up until recently, RSI could be rather
cavalier about sharing this around as the library files are pretty nigh
useless without a working installation of IDL. So they put all their
license-enforcement efforts into the binary and didn't fuss when/where
people shared .pro-files around.
But these days, the library suddenly attains value by itself as the
binary core can be replaced (to better and better degree) with GDL.
Some of this is very intimately linked to IDL and it's workings -- like
all the windowing/widget stuff. But for myself I don't need those
things -- all *my* routines that need complex user-interactions have
been talking to ports for the last five years or so where the user can
pick things up with his/her favorite web-browser (which doesn't even
have to run on the same computer).
But there's other routines, and linfit is an obvious example, that are
not in the least married to IDL - but who's well-defined (by RSI)
interface still expedites software development. We can all start
writing our own versions, but then my programs would either become
non-portable or otherwise lumbering hunks of re-re-re-duplicated code.
The longer I think about it, the more I realize that the ability to
replicate (copy? re-write? cobble together from various places?
reverse-engineer? mimick? which of these is legal these days under the
RSI license?) the library is going to be the thing that'll make GDL
stand or fall.
just some thoughts...
cordially
Y.T.
--
Remove YourClothes before you email me.
|
|
|
|
| Re: IDL, GDL, copyright, EULAs and such [message #44770 is a reply to message #44698] |
Thu, 14 July 2005 08:00 |
m_schellens
Messages: 31
Registered: February 2005
|
Member |
|
|
Y.T. wrote:
>> But I think if you look around
>> http://www.astro.washington.ed u/deutsch/idl/htmlhelp/
>>
>> You'll find that almost all the IDL code has been duplicated by
>> someone else somewhere else. You are probably allowed to use that.
>
> Maybe someone needs to start collecting these things and compile them
> into a "free idl-alike .pro library". The value of something like GDL
> would really increase drastically through this.
Well, that is already on the way. As Chris already poined out:
Severall routines which come with GDL are written in GDL (src/pro
subdirectory).
> And earlier, Christopher Lee wrote:
>> There's a free version of linfit at
>> http://www.astro.washington.ed u/deutsch-bin/getpro/library37 .html?LINFIT
>>
>> it's not the same as the IDL version, so you'd need to work it into a
>> function, and add the keywords that the IDL version makes available
>
> Not to sound ungrateful, but that's kinda pointless. I can write my own
> version of linfit (and everything else in the IDL library). If I have
> to write scads of code to make a trivial algorithm work, then I might
> as well implement the whole thing myself.
>
> The point was that there's already an existing library that has it all
> inside - the IDL library. Up until recently, RSI could be rather
> cavalier about sharing this around as the library files are pretty nigh
> useless without a working installation of IDL. So they put all their
> license-enforcement efforts into the binary and didn't fuss when/where
> people shared .pro-files around.
It is their code. If visible (.pro) or not (IDL binary).
I would treat it the same way.
from another post:
> As long as nobody looks at my harddisk, obviously there's no problem
> here anywhere. But lets say I tinker with GDL and write some kind of
> groundbreaking stock-trading software or some kind of super-accurate
> prediction code for earthquakes -- something that makes me rich enough
> to become a worthy target for litigation and famous enough for all the
> litigators to know that I exist. At that point I'd rather not have RSI
> thugs kick down my door and demand all my money because I did my linear
> fits with a routine that had their copyright in the header.
Not being an expert for law as well to me it is pretty obvious that
you cannot use IDL's code and distribute it with GDL applications.
But if an algorithm is that obvious you can use it of course.
Just write it yourself and don't copy it.
It will look different. Even if it is similar: I think nobody could
convince a judge that you should pay anything because you used
*similar* code, especially if a subroutine is almost trivial.
But if you use the *same* code it might be different.
> But these days, the library suddenly attains value by itself as the
> binary core can be replaced (to better and better degree) with GDL.
>
> Some of this is very intimately linked to IDL and it's workings -- like
> all the windowing/widget stuff. But for myself I don't need those
> things -- all *my* routines that need complex user-interactions have
> been talking to ports for the last five years or so where the user can
> pick things up with his/her favorite web-browser (which doesn't even
> have to run on the same computer).
>
> But there's other routines, and linfit is an obvious example, that are
> not in the least married to IDL - but who's well-defined (by RSI)
> interface still expedites software development. We can all start
> writing our own versions, but then my programs would either become
> non-portable or otherwise lumbering hunks of re-re-re-duplicated code.
Therefore put the routines you wrote in/for GDL under the GPL and
contribute them.
The more people do the quicker the library gets filled.
> The longer I think about it, the more I realize that the ability to
> replicate (copy? re-write? cobble together from various places?
> reverse-engineer? mimick? which of these is legal these days under the
> RSI license?) the library is going to be the thing that'll make GDL
> stand or fall.
You are right. But it is not about the IDL library only but the
internal (C++) routines as well or even more. And thats is what the
developers and me mainly work on.
Join in!
Cheers,
marc
|
|
|
|
| Re: IDL, GDL, copyright, EULAs and such [message #44781 is a reply to message #44698] |
Wed, 13 July 2005 12:26 |
Ken Mankoff
Messages: 158
Registered: February 2000
|
Senior Member |
|
|
On Wed, 13 Jul 2005, Y.T. wrote:
>> But I think if you look around
>> http://www.astro.washington.edu/deutsch/idl/htmlhelp/
>>
>> You'll find that almost all the IDL code has been duplicated by
>> someone else somewhere else. You are probably allowed to use
>> that.
>
> Maybe someone needs to start collecting these things and compile
> them into a "free idl-alike .pro library". The value of something
> like GDL would really increase drastically through this.
Umm... did you follow the above link? Doesn't that constitute a
"collection"?
-k.
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
