Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > comp.lang.forth > #8331 > unrolled thread

CarrierIQ Software and Forth

Started byKrishna Myneni <krishna.myneni@ccreweb.org>
First post2011-12-23 19:03 -0800
Last post2012-01-05 17:17 -0800
Articles 20 on this page of 120 — 25 participants

Back to article view | Back to comp.lang.forth


Contents

  CarrierIQ Software and Forth Krishna Myneni <krishna.myneni@ccreweb.org> - 2011-12-23 19:03 -0800
    Re: CarrierIQ Software and Forth Krishna Myneni <krishna.myneni@ccreweb.org> - 2011-12-23 19:14 -0800
      Re: CarrierIQ Software and Forth Mark Wills <markrobertwills@yahoo.co.uk> - 2011-12-24 01:52 -0800
        Re: CarrierIQ Software and Forth Krishna Myneni <krishna.myneni@ccreweb.org> - 2011-12-24 04:06 -0800
          Re: CarrierIQ Software and Forth Bernd Paysan <bernd.paysan@gmx.de> - 2011-12-24 14:56 +0100
            Re: CarrierIQ Software and Forth mhx@iae.nl (Marcel Hendrix) - 2011-12-24 17:53 +0200
              Re: CarrierIQ Software and Forth Mark Wills <markrobertwills@yahoo.co.uk> - 2011-12-24 09:23 -0800
              Re: CarrierIQ Software and Forth BruceMcF <agila61@netscape.net> - 2011-12-24 12:14 -0800
              Re: CarrierIQ Software and Forth Bernd Paysan <bernd.paysan@gmx.de> - 2011-12-25 02:19 +0100
                Re: CarrierIQ Software and Forth Dennis Ruffer <daruffer@gmail.com> - 2011-12-25 12:58 -0800
                  Re: CarrierIQ Software and Forth Bernd Paysan <bernd.paysan@gmx.de> - 2011-12-25 23:53 +0100
                Re: CarrierIQ Software and Forth BruceMcF <agila61@netscape.net> - 2011-12-25 13:25 -0800
                  Re: CarrierIQ Software and Forth Bernd Paysan <bernd.paysan@gmx.de> - 2011-12-26 00:10 +0100
                Re: CarrierIQ Software and Forth mhx@iae.nl (Marcel Hendrix) - 2011-12-25 22:36 +0200
                  Re: CarrierIQ Software and Forth Bernd Paysan <bernd.paysan@gmx.de> - 2011-12-26 00:09 +0100
                    Re: CarrierIQ Software and Forth Krishna Myneni <krishna.myneni@ccreweb.org> - 2011-12-28 08:36 -0800
                Re: CarrierIQ Software and Forth John Passaniti <john.passaniti@gmail.com> - 2011-12-27 09:43 -0800
                  Re: CarrierIQ Software and Forth BruceMcF <agila61@netscape.net> - 2011-12-27 09:56 -0800
                    Re: CarrierIQ Software and Forth John Passaniti <john.passaniti@gmail.com> - 2011-12-27 11:08 -0800
                      Re: CarrierIQ Software and Forth Krishna Myneni <krishna.myneni@ccreweb.org> - 2011-12-27 22:38 -0800
                        Re: CarrierIQ Software and Forth "Elizabeth D. Rather" <erather@forth.com> - 2011-12-27 22:01 -1000
                          Re: CarrierIQ Software and Forth Paul Rubin <no.email@nospam.invalid> - 2011-12-28 02:20 -0800
                            Re: CarrierIQ Software and Forth John Passaniti <john.passaniti@gmail.com> - 2011-12-28 12:12 -0800
                              Re: CarrierIQ Software and Forth Alex McDonald <blog@rivadpm.com> - 2011-12-28 14:15 -0800
                                Re: CarrierIQ Software and Forth Bernd Paysan <bernd.paysan@gmx.de> - 2011-12-28 23:56 +0100
                                  Re: CarrierIQ Software and Forth Alex McDonald <blog@rivadpm.com> - 2011-12-28 15:07 -0800
                      Re: CarrierIQ Software and Forth BruceMcF <agila61@netscape.net> - 2011-12-28 11:11 -0800
                  Re: CarrierIQ Software and Forth Bernd Paysan <bernd.paysan@gmx.de> - 2011-12-27 22:01 +0100
                    Re: CarrierIQ Software and Forth Nomen Nescio <nobody@dizum.com> - 2011-12-28 11:11 +0100
                      Re: CarrierIQ Software and Forth Bernd Paysan <bernd.paysan@gmx.de> - 2011-12-28 13:57 +0100
                      Re: CarrierIQ Software and Forth Albert van der Horst <albert@spenarnc.xs4all.nl> - 2011-12-29 11:00 +0000
                    Re: CarrierIQ Software and Forth John Passaniti <john.passaniti@gmail.com> - 2011-12-28 08:39 -0800
                      GPL [Was: CarrierIQ Software and Forth] Andrew Haley <andrew29@littlepinkcloud.invalid> - 2011-12-28 12:32 -0600
                        Re: GPL [Was: CarrierIQ Software and Forth] John Passaniti <john.passaniti@gmail.com> - 2011-12-28 12:07 -0800
                          Re: GPL [Was: CarrierIQ Software and Forth] Bernd Paysan <bernd.paysan@gmx.de> - 2011-12-28 23:19 +0100
                          Re: GPL [Was: CarrierIQ Software and Forth] Arnold Doray <invalid@invalid.com> - 2011-12-29 08:38 +0000
                            Re: GPL [Was: CarrierIQ Software and Forth] "Rod Pemberton" <do_not_have@noavailemail.cmm> - 2011-12-29 04:59 -0500
                              Re: GPL [Was: CarrierIQ Software and Forth] Bernd Paysan <bernd.paysan@gmx.de> - 2011-12-29 14:04 +0100
                                Re: GPL Aleksej Saushev <asau@inbox.ru> - 2011-12-30 00:12 +0400
                                  Re: GPL cas_news@strotmann.de (Carsten Strotmann (Usenet)) - 2011-12-29 21:48 +0100
                                    Re: GPL Bernd Paysan <bernd.paysan@gmx.de> - 2011-12-29 23:50 +0100
                                Re: GPL [Was: CarrierIQ Software and Forth] "Rod Pemberton" <do_not_have@noavailemail.cmm> - 2011-12-29 18:00 -0500
                                  Re: GPL [Was: CarrierIQ Software and Forth] Bernd Paysan <bernd.paysan@gmx.de> - 2011-12-30 01:38 +0100
                                    Re: GPL [Was: CarrierIQ Software and Forth] "Rod Pemberton" <do_not_have@noavailemail.cmm> - 2011-12-30 13:46 -0500
                                      Re: GPL [Was: CarrierIQ Software and Forth] Bernd Paysan <bernd.paysan@gmx.de> - 2011-12-30 22:58 +0100
                                        Re: GPL [Was: CarrierIQ Software and Forth] "Rod Pemberton" <do_not_have@noavailemail.cmm> - 2011-12-31 21:20 -0500
                                          Re: GPL [Was: CarrierIQ Software and Forth] BruceMcF <agila61@netscape.net> - 2012-01-01 01:46 -0800
                                          Re: GPL [Was: CarrierIQ Software and Forth] Andrew Haley <andrew29@littlepinkcloud.invalid> - 2012-01-01 04:43 -0600
                                          Re: GPL [Was: CarrierIQ Software and Forth] Bernd Paysan <bernd.paysan@gmx.de> - 2012-01-01 16:45 +0100
                                          Re: GPL [Was: CarrierIQ Software and Forth] Brad <hwfwguy@gmail.com> - 2012-01-01 08:19 -0800
                                Re: GPL [Was: CarrierIQ Software and Forth] Andrew Haley <andrew29@littlepinkcloud.invalid> - 2011-12-30 04:13 -0600
                              Re: GPL [Was: CarrierIQ Software and Forth] BruceMcF <agila61@netscape.net> - 2011-12-29 11:15 -0800
                                Re: GPL [Was: CarrierIQ Software and Forth] "Rod Pemberton" <do_not_have@noavailemail.cmm> - 2011-12-29 18:07 -0500
                                  Re: GPL [Was: CarrierIQ Software and Forth] BruceMcF <agila61@netscape.net> - 2011-12-29 16:21 -0800
                                  Re: GPL [Was: CarrierIQ Software and Forth] Bernd Paysan <bernd.paysan@gmx.de> - 2011-12-30 01:43 +0100
                                    Re: GPL [Was: CarrierIQ Software and Forth] Mark Wills <markrobertwills@yahoo.co.uk> - 2011-12-29 23:55 -0800
                                      Re: GPL cas_news@strotmann.de (Carsten Strotmann (Usenet)) - 2011-12-30 10:38 +0100
                                      Re: GPL cas_news@strotmann.de (Carsten Strotmann (Usenet)) - 2011-12-30 11:12 +0100
                                        Re: GPL anton@mips.complang.tuwien.ac.at (Anton Ertl) - 2011-12-30 13:14 +0000
                                    Re: GPL [Was: CarrierIQ Software and Forth] "Rod Pemberton" <do_not_have@noavailemail.cmm> - 2011-12-30 13:40 -0500
                                      Re: GPL [Was: CarrierIQ Software and Forth] BruceMcF <agila61@netscape.net> - 2011-12-30 13:24 -0800
                                    Re: GPL [Was: CarrierIQ Software and Forth] Nomen Nescio <nobody@dizum.com> - 2012-01-02 08:19 +0100
                                      Re: GPL [Was: CarrierIQ Software and Forth] Albert van der Horst <albert@spenarnc.xs4all.nl> - 2012-01-02 15:46 +0000
                                  Re: GPL [Was: CarrierIQ Software and Forth] Fritz Wuehler <fritz@spamexpire-201112.rodent.frell.theremailer.net> - 2011-12-30 20:26 +0100
                                    Re: GPL [Was: CarrierIQ Software and Forth] Spam@ControlQ.com - 2012-01-01 14:45 -0500
                                      Re: GPL [Was: CarrierIQ Software and Forth] BruceMcF <agila61@netscape.net> - 2012-01-01 12:28 -0800
                                      Re: GPL [Was: CarrierIQ Software and Forth] "Rod Pemberton" <do_not_have@noavailemail.cmm> - 2012-01-02 02:03 -0500
                              Re: GPL [Was: CarrierIQ Software and Forth] anton@mips.complang.tuwien.ac.at (Anton Ertl) - 2011-12-30 12:33 +0000
                                Re: GPL [Was: CarrierIQ Software and Forth] Brad <hwfwguy@gmail.com> - 2011-12-30 17:30 -0800
                              Re: GPL [Was: CarrierIQ Software and Forth] Arnold Doray <invalid@invalid.com> - 2011-12-30 17:07 +0000
                          Re: GPL [Was: CarrierIQ Software and Forth] Andrew Haley <andrew29@littlepinkcloud.invalid> - 2011-12-29 04:26 -0600
                          Re: GPL [Was: CarrierIQ Software and Forth] BruceMcF <agila61@netscape.net> - 2011-12-29 11:00 -0800
                          Re: GPL [Was: CarrierIQ Software and Forth] anton@mips.complang.tuwien.ac.at (Anton Ertl) - 2011-12-30 12:18 +0000
                        Re: GPL [Was: CarrierIQ Software and Forth] BruceMcF <agila61@netscape.net> - 2011-12-28 12:49 -0800
                          Re: GPL [Was: CarrierIQ Software and Forth] Andrew Haley <andrew29@littlepinkcloud.invalid> - 2011-12-29 04:37 -0600
                            Re: GPL [Was: CarrierIQ Software and Forth] BruceMcF <agila61@netscape.net> - 2011-12-29 11:09 -0800
                              Re: GPL [Was: CarrierIQ Software and Forth] Andrew Haley <andrew29@littlepinkcloud.invalid> - 2011-12-30 04:21 -0600
                        Re: GPL [Was: CarrierIQ Software and Forth] Albert van der Horst <albert@spenarnc.xs4all.nl> - 2011-12-29 13:53 +0000
                          Re: GPL [Was: CarrierIQ Software and Forth] BruceMcF <agila61@netscape.net> - 2011-12-29 11:58 -0800
                            Re: GPL [Was: CarrierIQ Software and Forth] Albert van der Horst <albert@spenarnc.xs4all.nl> - 2011-12-30 11:05 +0000
                              Re: GPL [Was: CarrierIQ Software and Forth] BruceMcF <agila61@netscape.net> - 2011-12-30 13:54 -0800
                                Re: GPL [Was: CarrierIQ Software and Forth] Albert van der Horst <albert@spenarnc.xs4all.nl> - 2012-01-01 11:51 +0000
                                  Re: GPL [Was: CarrierIQ Software and Forth] BruceMcF <agila61@netscape.net> - 2012-01-01 12:33 -0800
                                    Re: GPL [Was: CarrierIQ Software and Forth] Albert van der Horst <albert@spenarnc.xs4all.nl> - 2012-01-02 10:33 +0000
                                      Re: GPL [Was: CarrierIQ Software and Forth] BruceMcF <agila61@netscape.net> - 2012-01-02 11:40 -0800
                                  Re: GPL [Was: CarrierIQ Software and Forth] BruceMcF <agila61@netscape.net> - 2012-01-01 12:37 -0800
                                    Re: GPL [Was: CarrierIQ Software and Forth] Albert van der Horst <albert@spenarnc.xs4all.nl> - 2012-01-02 10:36 +0000
                                    Re: GPL [Was: CarrierIQ Software and Forth] Bernd Paysan <bernd.paysan@gmx.de> - 2012-01-02 14:27 +0100
                                      Re: GPL [Was: CarrierIQ Software and Forth] Alex McDonald <blog@rivadpm.com> - 2012-01-05 04:08 -0800
                                        Re: GPL [Was: CarrierIQ Software and Forth] Bernd Paysan <bernd.paysan@gmx.de> - 2012-01-05 17:40 +0100
                                      Re: GPL [Was: CarrierIQ Software and Forth] BruceMcF <agila61@netscape.net> - 2012-01-05 04:59 -0800
                                      Re: GPL [Was: CarrierIQ Software and Forth] Andrew Haley <andrew29@littlepinkcloud.invalid> - 2012-01-05 09:49 -0600
                                        Re: GPL [Was: CarrierIQ Software and Forth] BruceMcF <agila61@netscape.net> - 2012-01-05 07:56 -0800
                              Re: GPL [Was: CarrierIQ Software and Forth] BruceMcF <agila61@netscape.net> - 2011-12-30 14:00 -0800
                    Re: CarrierIQ Software and Forth John Passaniti <john.passaniti@gmail.com> - 2011-12-28 10:33 -0800
                      Re: CarrierIQ Software and Forth BruceMcF <agila61@netscape.net> - 2011-12-28 11:30 -0800
                        Re: CarrierIQ Software and Forth John Passaniti <john.passaniti@gmail.com> - 2011-12-28 15:28 -0800
                          Re: CarrierIQ Software and Forth BruceMcF <agila61@netscape.net> - 2011-12-29 11:49 -0800
                          Re: CarrierIQ Software and Forth jyf <jyf1987@gmail.com> - 2012-01-02 08:32 -0800
                      Re: CarrierIQ Software and Forth Bernd Paysan <bernd.paysan@gmx.de> - 2011-12-28 22:26 +0100
                  Re: CarrierIQ Software and Forth Paul Rubin <no.email@nospam.invalid> - 2011-12-27 20:22 -0800
            Re: CarrierIQ Software and Forth BruceMcF <agila61@netscape.net> - 2011-12-24 09:11 -0800
              Re: CarrierIQ Software and Forth Mark Wills <markrobertwills@yahoo.co.uk> - 2011-12-24 09:29 -0800
      Re: CarrierIQ Software and Forth Aleksej Saushev <asau@inbox.ru> - 2011-12-26 00:29 +0400
      Re: CarrierIQ Software and Forth Ian Osgood <iano@quirkster.com> - 2011-12-27 14:15 -0800
        Re: CarrierIQ Software and Forth Richard Owlett <rowlett@pcnetinc.com> - 2011-12-27 16:29 -0600
        Re: CarrierIQ Software and Forth Krishna Myneni <krishna.myneni@ccreweb.org> - 2011-12-27 21:10 -0800
      Re: CarrierIQ Software and Forth Mark Wills <markrobertwills@yahoo.co.uk> - 2011-12-29 23:46 -0800
        Re: CarrierIQ Software and Forth Krishna Myneni <krishna.myneni@ccreweb.org> - 2011-12-31 16:41 -0800
    Re: CarrierIQ Software and Forth Trey Boudreau <trey.boudreau@gmail.com> - 2011-12-31 20:50 -0800
      Re: CarrierIQ Software and Forth Krishna Myneni <krishna.myneni@ccreweb.org> - 2012-01-01 18:14 -0800
        Re: CarrierIQ Software and Forth Trey Boudreau <trey.boudreau@gmail.com> - 2012-01-01 21:33 -0800
          Re: CarrierIQ Software and Forth "Rod Pemberton" <do_not_have@noavailemail.cmm> - 2012-01-02 01:57 -0500
            Re: CarrierIQ Software and Forth Krishna Myneni <krishna.myneni@ccreweb.org> - 2012-01-02 04:45 -0800
            Re: CarrierIQ Software and Forth Trey Boudreau <trey.boudreau@gmail.com> - 2012-01-02 07:36 -0800
              Re: CarrierIQ Software and Forth Bernd Paysan <bernd.paysan@gmx.de> - 2012-01-05 15:13 +0100
                Re: CarrierIQ Software and Forth Trey Boudreau <trey.boudreau@gmail.com> - 2012-01-05 13:15 -0800
                  Re: CarrierIQ Software and Forth Krishna Myneni <krishna.myneni@ccreweb.org> - 2012-01-05 16:19 -0800
                    Re: CarrierIQ Software and Forth Krishna Myneni <krishna.myneni@ccreweb.org> - 2012-01-05 16:22 -0800
                    Re: CarrierIQ Software and Forth Trey Boudreau <trey.boudreau@gmail.com> - 2012-01-05 17:17 -0800

Page 2 of 6 — ← Prev page 1 [2] 3 4 5 6  Next page →


#8389

From"Elizabeth D. Rather" <erather@forth.com>
Date2011-12-27 22:01 -1000
Message-ID<WY2dnX3ITIRXUWfTnZ2dnUVZ_jWdnZ2d@supernews.com>
In reply to#8387
On 12/27/11 8:38 PM, Krishna Myneni wrote:
> On Dec 27, 1:08 pm, John Passaniti<john.passan...@gmail.com>  wrote:
>> On Dec 27, 12:56 pm, BruceMcF<agil...@netscape.net>  wrote:
>>
>>> Its not entirely clear why developing in Forth with the
>>> specified set of extensions at hand is not "Forth
>>> development for mobile devices".
>>
>> I believe that at some point after extending Forth with extensions and
>> application-specific words, you are no longer programming in Forth.
>> And in this particular case, the set of extensions and application-
>> specific words one would need to add to make it a *competitive*
>> language for mobile app development would make the resulting language
>> different from Forth in a number of important ways.
>>
>> In order for Forth to be competitive as a language in the mobile
>> space, I presume it would have a number of features that "raw" Forth
>> does not.  These include:
>>
>> 1.  An object system that respected the platform's object model.
>> 2.  Direct support for creating asynchronous, event-driven apps.
>> 3.  Garbage collection (primarily to support non-sequential,
>> asynchronous flows).
>> 4.  Dynamic language features like typed values and generic data
>> structures.
>>
>> The end result of what I consider a Forth-based competitive language
>> for mobile development would look and work a lot less like traditional
>> Forth and would be closer to Postscript or Factor.  You are certainly
>> free to call such a language Forth if you like, but I'd consider that
>> next to meaningless.

Forths adapt to the environment in which they're used. SwiftForth (and 
other Windows Forths) have #1 above, and every Forth I've worked with, 
dating back to the 70's has had #2.  I'm not convinced as to the 
necessity for #s 3 & $, but I freely admit I haven't programmed mobile aps.

I would call such a thing "a version of Forth adapted for ...".

...
> Also, you seem to imply that Forth isn't well suited for use in highly
> asynchronous, event-driven environments. I guess that would rule out
> writing Windows and X-Windows apps in Forth... wait, that's been done
> already (e.g., see Bernd's GUI development interface).

"Highly asynchronous, event-driven environments" accurately describes 
the kind of applications for which Forth was originally developed, way 
back when.

Cheers,
Elizabeth

-- 
==================================================
Elizabeth D. Rather   (US & Canada)   800-55-FORTH
FORTH Inc.                         +1 310.999.6784
5959 West Century Blvd. Suite 700
Los Angeles, CA 90045
http://www.forth.com

"Forth-based products and Services for real-time
applications since 1973."
==================================================

[toc] | [prev] | [next] | [standalone]


#8390

FromPaul Rubin <no.email@nospam.invalid>
Date2011-12-28 02:20 -0800
Message-ID<7x8vlxq9p0.fsf@ruckus.brouhaha.com>
In reply to#8389
"Elizabeth D. Rather" <erather@forth.com> writes:
>>> 1.  An object system that respected the platform's object model.
>>> 2.  Direct support for creating asynchronous, event-driven apps.
>>> 3.  Garbage collection (primarily to support non-sequential,
>>> asynchronous flows).
>>> 4.  Dynamic language features like typed values and generic data
>>> structures.
>
> Forths adapt to the environment in which they're used. SwiftForth (and
> other Windows Forths) have #1 above, and every Forth I've worked with,
> dating back to the 70's has had #2.  I'm not convinced as to the
> necessity for #s 3 & $, but I freely admit I haven't programmed mobile
> aps....
>
> "Highly asynchronous, event-driven environments" accurately describes
> the kind of applications for which Forth was originally developed, way
> back when.

There's tons of Forth OO libraries that handle #4, I'd expect.  Wrapping
them around mobile OS GUI widgets would be some work, of course.

#3 (GC) is convenient but I've written enough desktop GUI code in C to
know one can usually write in a style that doesn't need it.  Does
Objective-C even have it?  I thought Objective-C was basically regular C
plus a simple OO library wrapped with some syntax sugar.  I'd think that
by programming a multi-tasking Forth in an Erlang-like fashion
(i.e. accepting the overhead of copying messages around instead of
passing pointers) one could keep the memory allocation patterns simple
enough to manage manually without much fuss.  Also, I know that at least
for gforth, there is a Boehm-style GC available

[toc] | [prev] | [next] | [standalone]


#8408

FromJohn Passaniti <john.passaniti@gmail.com>
Date2011-12-28 12:12 -0800
Message-ID<00d2cb8d-3c1a-4191-897d-baee911b3fd2@z1g2000vbx.googlegroups.com>
In reply to#8390
On Dec 28, 5:20 am, Paul Rubin <no.em...@nospam.invalid> wrote:
> There's tons of Forth OO libraries that handle #4, I'd expect.
> Wrapping them around mobile OS GUI widgets would be some work,
> of course.

Of course.  And that "some work" is the point.  To be *competitive* as
a development platform, that "some work" has to be done.

> #3 (GC) is convenient but I've written enough desktop GUI code
> in C to know one can usually write in a style that doesn't
> need it.

Yes, one can always manually manage memory, track exceptions, unwind
call hierarchies, and with some amount of effort ensure that garbage
is collected and finalizers are called.  One can also defer all that
to the runtime and not have to spend time working at such a low level
and refocusing effort where it is better spent.

> Does Objective-C even have it?

Yes and no.  Apple has it for desktop systems, but not for iOS
devices.

Writing iOS apps in Objective C is popular, but it isn't the only
option.  Our app is in Corona (which is based on Lua), and other
popular choices for apps are Javascript and C# (using Mono).  All
three are garbage collected languages, and far more pleasant to work
with.

> I thought Objective-C was basically regular C plus a simple
> OO library wrapped with some syntax sugar.

Mostly.  The extensions to C are simple to understand and useful.  But
it's really the libraries that provide the power and that is what you
would want full access to for iOS development.  For Android, you would
want the same for Java.

> Nobody I'd think that by programming a multi-tasking Forth in
> an Erlang-like fashion (i.e. accepting the overhead of
> copying messages around instead of passing pointers) one
> could keep the memory allocation patterns simple enough to
> manage manually without much fuss.  [...]

Depends on the complexity of the app.  If you're writing the next fart
app, you're all set.  If your app involves complex mixtures of network
messaging, timers, and random user interaction, the effort you'll
spend on ensuring there won't be "much fuss" using manual methods is
non-zero.  Using a garbage collecting language means you focus on the
parts of the app that matter and that differentiate you, not the
plumbing.

The question I keep raising here isn't if X is possible in Forth, but
if X is *competitive* with other development tools.  Answering that
question is likely to lead to more insight into what mobile app
developers want and need than saying, "it's possible in Forth."

[toc] | [prev] | [next] | [standalone]


#8411

FromAlex McDonald <blog@rivadpm.com>
Date2011-12-28 14:15 -0800
Message-ID<be1f340e-e0bf-447e-bdda-572482b6dee0@l24g2000yqm.googlegroups.com>
In reply to#8408
On Dec 28, 8:12 pm, John Passaniti <john.passan...@gmail.com> wrote:

>
> Depends on the complexity of the app.  If you're writing the next fart
> app, you're all set.

*Fart* app? I've scanned my keyboard for possible key fumbles, but
nothing seems to fit...

[toc] | [prev] | [next] | [standalone]


#8413

FromBernd Paysan <bernd.paysan@gmx.de>
Date2011-12-28 23:56 +0100
Message-ID<jdg6rg$iav$1@online.de>
In reply to#8411
Alex McDonald wrote:

> On Dec 28, 8:12 pm, John Passaniti <john.passan...@gmail.com> wrote:
> 
>>
>> Depends on the complexity of the app.  If you're writing the next
>> fart app, you're all set.
> 
> *Fart* app? I've scanned my keyboard for possible key fumbles, but
> nothing seems to fit...

Don't search your keyboard, search the app store.  Or the Android 
market, I don't know how to search the app store... (at least 1000 
search results):

https://market.android.com/search?q=fart&c=apps

I'd buy some onions and beans, and tell people that my new fart app even 
smells ;-).

-- 
Bernd Paysan
"If you want it done right, you have to do it yourself"
http://bernd-paysan.de/

[toc] | [prev] | [next] | [standalone]


#8414

FromAlex McDonald <blog@rivadpm.com>
Date2011-12-28 15:07 -0800
Message-ID<b4a2dcbe-82aa-4a78-9db5-80ea9b92bb97@q17g2000yqh.googlegroups.com>
In reply to#8413
On Dec 28, 10:56 pm, Bernd Paysan <bernd.pay...@gmx.de> wrote:
> Alex McDonald wrote:
> > On Dec 28, 8:12 pm, John Passaniti <john.passan...@gmail.com> wrote:
>
> >> Depends on the complexity of the app.  If you're writing the next
> >> fart app, you're all set.
>
> > *Fart* app? I've scanned my keyboard for possible key fumbles, but
> > nothing seems to fit...
>
> Don't search your keyboard, search the app store.  Or the Android
> market, I don't know how to search the app store... (at least 1000
> search results):
>
> https://market.android.com/search?q=fart&c=apps
>
> I'd buy some onions and beans, and tell people that my new fart app even
> smells ;-).
>
> --
> Bernd Paysan
> "If you want it done right, you have to do it yourself"http://bernd-paysan.de/

Dear god, I thought I'd seen (and heard) it all. Interestingly, the
character limit and use of ellipsis has given the second offering on
that search some unintended humour;

The Whoopee releases a horde of realistic fart sounds at the touch of
a butt...

Debugging one of these apps must be (and I really do apologise for the
horrible double pun) a gas.

[toc] | [prev] | [next] | [standalone]


#8405

FromBruceMcF <agila61@netscape.net>
Date2011-12-28 11:11 -0800
Message-ID<2a2e5946-73f1-4abf-a738-f1bf75eba975@o9g2000yqa.googlegroups.com>
In reply to#8370
On Dec 27, 2:08 pm, John Passaniti <john.passan...@gmail.com> wrote:

> 1.  An object system that respected the platform's object model.
> 2.  Direct support for creating asynchronous, event-driven apps.
> 3.  Garbage collection (primarily to support non-sequential,
> asynchronous flows).
> 4.  Dynamic language features like typed values and generic data
> structures.

"When you've seen one Forth object system, you've just got started",
but most of the existing models seem to me to be programming in Forth,
and the fact that the object model is not built into the core model
itself suggests the straightforward approach of respecting the
platform's object model by loading an object system that respects the
platform's object model.

Direct support for creating asynchronous event-drive apps have been a
feature of Forth implementations since the 1980's. Whether they have
had to include garbage collection for some data structures seems to
have depended on whether the application as written leaves garbage to
be collected.

As far as the benefit of (4) in Rapid Application Development, I'll
let more experienced programmers argue through ~ for the kind of stuff
where I'd be programming a back-end in Forth, its not a big issue.

[toc] | [prev] | [next] | [standalone]


#8376

FromBernd Paysan <bernd.paysan@gmx.de>
Date2011-12-27 22:01 +0100
Message-ID<jddbni$nk$1@online.de>
In reply to#8366
John Passaniti wrote:
> But to be fair, Apple doesn't have a problem with all free software--
> they have a problem with free software with viral licenses that would
> cause them to have to release source for their proprietary code.

Oh, John, that's bullshit, and you know it.  No "viral license" free 
software requires to release the source of independently developed 
software, as the "viral license" uses copyright to get into that 
position.  The GPL, one particular prominent example of a so-called 
"viral license" (a political term from Microsoft, which nobody in their 
right mind should use without quotation marks) is even very explicit 
that you are allowed to link your GPL program to libraries and other 
parts that are commonly shipped with the operating system to run on, 
even when those libraries are not open sourced.

Apple actually uses GPL'd software in their products, including a very 
central part in iOS, that's webkit.  Ok, it's LGPL, because webkit is a 
library, but inside that library, the LGPL is just as "viral" as any 
other GPL.

I don't like the term "viral" at all.  A virus, in software terminology, 
is a kind of malware, attaching itself to other, independent programs.  
No license can do that.  It's attached to the original program, and it 
can only give permissions.  When something is viral, i.e. has malware-
like characteristics, then it is copyright itself.  The proprietary 
licenses usually deny third parties all rights except running the 
program, and therefore, they must by definition be worse than any GPL'd 
software.  The GPL gives you rights, but when you use these rights, it 
requires you that you release the source code of the additions and 
modifications you made to that program, i.e. these rights are 
transitive, you can't take them away by modifying the program.

Also, what pisses Apple off is that they want monopoly power over 
distributing apps to iOS devices (their App store).  This conflicts with 
free software, as anybody is allowed to redistribute it (regardless if 
"viral" or not).

What they might be concerned about is that as app store provider, they 
are redistributing the software, and therefore have to care about 
availability of the source code, maybe in the outdated way the GPLv2 
describes it.  However, nobody today wants a CD or tape with the 
software, everybody wants a link to some version control repository.  
Therefore, nobody will complain if they can provide that link.  They can 
check if the source code provided by the maintainers actually is the 
source code they ship, and remove the app from the store once that is no 
longer the case.

> Apple *was* unnecessarily grandiose and had said that applications
> couldn't use interpreters (other than their own Javascript
> implementation in WebKit).  But that ended shortly after it was
> pointed out that a best-selling game for iOS that Steve Jobs held up
> at a keynote address was written in Lua.  And not long after that,
> other developers of best-selling apps who were using a variety of
> other interpreted languages asked for clarification.  It was kind of
> hard for Apple to keep their no-interpreter-besides-ours stance after
> apps they showcased were found to be written in interpreted languages.

But that takes out most of the fun of having an interpreter.

Let's take my net2o as example, which I'm working on currently:  The 
basic network layer contains an interpreter where you send commands back 
and forth to make the communication work (flow control, retransmission, 
acknowledges, file requests: all those are commands of a Forth-like 
virtual machine there).  The commands are constructed in an ad-hoc 
manner.  So even without the higher level parts of net2o (which aren't 
written yet, but the plan certainly is to have a dynamic interpretive 
language in its core), this is a no-go for iOS devices.

Of course, I'm using Gforth to write this, i.e. one of these dreadful 
viral licenses, which brings you the smallpox or at least lets you grow 
a beard and long hair like RMS (Andrew Haley and Peter Knaggs are also 
known to be infected ;-).

> Regardless, anyone who thinks that Forth development for mobile
> devices is a good idea probably hasn't written a non-trivial app for a
> mobile device.  What would make more sense is to develop a variant of
> Forth that embraced the fundamental qualities of mobile platforms such
> as being dynamic, event-based, highly asynchronous in their
> structure.  And you want to be able to access native libraries
> (especially the UI libraries) in some sensible way.  I can easily see
> an augmented Forth for such platforms working well for developers.

Hm, Forth can be quite dynamic, and wherever I used in embedded 
programming or GUI programming, I always resorted to event-based 
execution, which is highly asynchronous.  Using native libraries is a 
little problem in Gforth on these platforms, as there is no gcc to 
compile the glue code - but that's just a little problem, not a big one.

-- 
Bernd Paysan
"If you want it done right, you have to do it yourself"
http://bernd-paysan.de/

[toc] | [prev] | [next] | [standalone]


#8392

FromNomen Nescio <nobody@dizum.com>
Date2011-12-28 11:11 +0100
Message-ID<2910db6c25777c2d0e3ce3eb9bc6d9fe@dizum.com>
In reply to#8376
Bernd Paysan <bernd.paysan@gmx.de> wrote:

> position.  The GPL, one particular prominent example of a so-called 
> "viral license" (a political term from Microsoft, which nobody in their 
> right mind should use without quotation marks)

Why not? That's exactly what the GPL is. It infects whatever it touches to
be GPL. It is a viral license indeed that's a very apt description.

> is even very explicit that you are allowed to link your GPL program to
> libraries and other parts that are commonly shipped with the operating
> system to run on, even when those libraries are not open sourced.

That is only for LGPL, not GPL.

> I don't like the term "viral" at all.  A virus, in software terminology, 
> is a kind of malware, attaching itself to other, independent programs.  
> No license can do that.

What? That's exactly what the GPL does! It's viral and it's malware.

> software.  The GPL gives you rights, but when you use these rights, it 
> requires you that you release the source code of the additions and 
> modifications you made to that program, i.e. these rights are 
> transitive, you can't take them away by modifying the program.

Nonsense, the GPL doesn't give anyone rights. It simply forces you to open
source whatever GPL infects.

[toc] | [prev] | [next] | [standalone]


#8394

FromBernd Paysan <bernd.paysan@gmx.de>
Date2011-12-28 13:57 +0100
Message-ID<jdf3no$m9e$1@online.de>
In reply to#8392
Nomen Nescio wrote:

> Bernd Paysan <bernd.paysan@gmx.de> wrote:
> 
>> position.  The GPL, one particular prominent example of a so-called
>> "viral license" (a political term from Microsoft, which nobody in
>> their right mind should use without quotation marks)
> 
> Why not? That's exactly what the GPL is. It infects whatever it
> touches to be GPL. It is a viral license indeed that's a very apt
> description.

Nonononono.  Copyright is viral.  Whenever you combine copyrighted work, 
you need permission.  It's not about touching, it's about combining.  
You plan and simply can't do that under copyright law, so you need 
permission.  The permission of transitive free software licenses (there 
are several) is that you can do that when you put the combined work 
under the original license of the parts.  Apparently it is too 
complicated to make these licenses compatible with each others, because 
anglo-saxonian contract law is broken by design.

The more premissive licenses (i.e. the ones who basically abandon 
copyright at all) allow you to relicense your work under something else.  
People use that to take it out of the free software world into the 
proprietary, secretive world.  Now, that's what I would say is "viral".  
The healthy sharing and improving ability dies, and the software gets 
the smallpox - look at the Carrier IQ example ;-).

>> is even very explicit that you are allowed to link your GPL program
>> to libraries and other parts that are commonly shipped with the
>> operating system to run on, even when those libraries are not open
>> sourced.
> 
> That is only for LGPL, not GPL.

No, that's the GPL.  Read it, search for "system libraries".  The LGPL 
additionally allows applications to link against it.

> What? That's exactly what the GPL does! It's viral and it's malware.

And you are a troll.

> Nonsense, the GPL doesn't give anyone rights. It simply forces you to
> open source whatever GPL infects.

The GPL gives you four freedoms, freedoms you don't have if it was plain 
simply copyrighted like all the fine proprietary software trolls like 
you love and cheerish.  I understand that selfish dumbasses don't 
understand the golden rule, and therefore can't comprehend that it's not 
all their own, and that they can't take without giving back.  That's why 
the GPL is necessary.

If you want to be a slave, you probably love the chains they put on your 
ankles.

-- 
Bernd Paysan
"If you want it done right, you have to do it yourself"
http://bernd-paysan.de/

[toc] | [prev] | [next] | [standalone]


#8421

FromAlbert van der Horst <albert@spenarnc.xs4all.nl>
Date2011-12-29 11:00 +0000
Message-ID<lwyp8h.fnb@spenarnc.xs4all.nl>
In reply to#8392
In article <2910db6c25777c2d0e3ce3eb9bc6d9fe@dizum.com>,
Nomen Nescio  <nobody@dizum.com> wrote:
>Bernd Paysan <bernd.paysan@gmx.de> wrote:
>
>> position.  The GPL, one particular prominent example of a so-called
>> "viral license" (a political term from Microsoft, which nobody in their
>> right mind should use without quotation marks)
>
>Why not? That's exactly what the GPL is. It infects whatever it touches to
>be GPL. It is a viral license indeed that's a very apt description.

No it isn't. That is the intention of the gut feeling you're supposed
to get by the spin doctors of Microsoft. Note that the
propaganda budget of Microsoft is of the same order of magnitude as
the government propaganda spent to promote the Third Reich.
Now I leave it to you to judge if such budget is sufficient to infect
people with lies. The budget is more effectively spent if accompagnied
with some arm twisting.

You better do some soul searching to find out whether you have
been infected with lies. For example try to imagine where the
hell that billion dollar budget of last year have flown.
Try to answer the question, has the propaganda been accompagnied
by some arm twisting?

<SNIP>

>Nonsense, the GPL doesn't give anyone rights. It simply forces you to open
>source whatever GPL infects.

You forget the IANAL. This is technically so far from the truth
that you better not leave the IANAL out.

Groetjes Albert

--
-- 
Albert van der Horst, UTRECHT,THE NETHERLANDS
Economic growth -- being exponential -- ultimately falters.
albert@spe&ar&c.xs4all.nl &=n http://home.hccnet.nl/a.w.m.van.der.horst

[toc] | [prev] | [next] | [standalone]


#8397

FromJohn Passaniti <john.passaniti@gmail.com>
Date2011-12-28 08:39 -0800
Message-ID<1c5e333c-d04f-4492-bc3c-116532c7706c@z19g2000vbe.googlegroups.com>
In reply to#8376
On Dec 27, 4:01 pm, Bernd Paysan <bernd.pay...@gmx.de> wrote:
> John Passaniti wrote:
> > But to be fair, Apple doesn't have a problem with all
> > free software-- they have a problem with free software
> > with viral licenses that would cause them to have to
> > release source for their proprietary code.
>
> Oh, John, that's bullshit, and you know it.  No "viral license"
> free software requires to release the source of independently
> developed software, as the "viral license" uses copyright to
> get into that position.  [...]

Regardless of your knee-jerk issues with the phrase "viral license,"
the fact is that Apple's "iOS Developer Program License Agreement"
makes the following statements about FOSS used as part of the
applications.

1.  You as a developer have to comply with the FOSS license terms.
2.  You will not use FOSS in such a way that would cause non-FOSS code
from Apple to be subject to any FOSS licensing terms or obligations.
3.  Your application will be consistent with Apple's digital signing
and content protection (that is, DRM) procedures and technologies.
4.  Any pre-release software from Apple is considered confidential,
except any FOSS Apple itself includes that does not impose additional
confidentiality obligations regarding use or disclosure of that FOSS.

That's it.  Item 1 is boilerplate and non-controversial.  Item 2 is
the issue we're talking about, and if you honestly believe there
doesn't exist now (or ever would exist) any FOSS licensing terms on
the planet that would provoke the situation, then it's a non-issue
that we can all ignore.  Item 3 is likely provoked by GPLv3, which has
(or at least had) placed restrictions regarding the use of DRM-- an
issue that Linux Torvalds explicitly cited as *one* of the reasons why
he isn't moving the Linux kernel to GPLv3.  And Item 4... I'm not
entirely sure what the concern is there, but again, if you honestly
believe it could never happen, then it can safely be ignored.

> The GPL, one particular prominent example of a so-called
> "viral license" (a political term from Microsoft, which
> nobody in their right mind should use without quotation
> marks) is even very explicit that you are allowed to link
> your GPL program to libraries and other parts that are
> commonly shipped with the operating system to run on,
> even when those libraries are not open sourced.

Bzzzt.  First, talking about the GPL as a single thing is sloppy.
There are a family of licenses that are collectively called "GPL" and
these have a variety of different restrictions and obligations.  In
particular, the relevant restrictions and obligations in question
occur when non-FOSS is compiled within or statically linked to the
executable itself and those are different between GPLv2 and GPLv3.

iOS developers can choose to statically link Apple's non-FOSS
libraries to their application.  In that case, depending on which GPL
license you're using, I can certainly see how both GPLv2 and GPLv3
could compel Apple to release source code if Apple didn't explicitly
forbid such in their developer license agreement.  If you disagree,
you are invited to point me to any authoritative statement from the
GPL folk (not your interpretation, but their words) which says this
can't happen.  You're also free to show me that no future version of
the GPL or other FOSS license could ever place such obligations on
Apple.

But really, it boils down to this:  You made the statement that Apple
has a "knee-jerk" policy regarding FOSS.  I don't exactly know what
precisely "knee-jerk" means, but if it means that Apple disallows or
discourages FOSS from being used in their iOS products, I have yet to
see it.  It may be fair to say that Apple (and Microsoft and others)
have a big problem with *specific* FOSS licenses.  It is not fair to
say that Apple (and Microsoft and others) have issues with *all* FOSS
licenses.  In particular, they don't seem to have any problem-- as
reflected in policy or license restriction-- with the so-called
"permissive FOSS licenses" like MIT and BSD.

> I don't like the term "viral" at all.

That's nice.  I'd get more out of this discussion if instead of a
discourse on phrases you don't like, you stick to both your original
claims and the substance of my reply.

> Also, what pisses Apple off is that they want monopoly power
> over distributing apps to iOS devices (their App store).  This
> conflicts with free software, as anybody is allowed to
> redistribute it (regardless if "viral" or not).

Yep, and I certainly don't defend Apple for that.  The restrictions
they place even on developers regarding distribution of applications
used internally for testing are painful and in my opinion, stupid.

> What they might be concerned about is that as app store provider,
> they are redistributing the software, and therefore have to care
> about availability of the source code, maybe in the outdated
> way the GPLv2 describes it.  However, nobody today wants a CD or
> tape with the software, everybody wants a link to some version
> control repository.  Therefore, nobody will complain if they
> can provide that link.  [...]

For GPLv3, yep-- and the GNU FAQ talks about this.  For GPLv2, the
"outdated way they describe it" still applies for GPLv2-licensed
FOSS.  But again, if you disagree, please provide a link to some
authoritative source (not your opinion).

> But that takes out most of the fun of having an interpreter.

It's worse than taking fun out; it potentially affects usability and
correctness.  In the apps that the company I work for is interested in
developing, we have a set of data structures (represented as Lua
tables) that describe our systems' capabilities and the communications
protocol used to talk to them.  Given that we're human and make
mistakes, we thought it would be cool if when we found problems in
those data structures, we could have the iOS device download from our
web site updated capability and protocol description data structures.
But Apple forbids this; we're free to interpret these data structures
at run-time, but they have to be part of the app.  Grrr.

[toc] | [prev] | [next] | [standalone]


#8400 — GPL [Was: CarrierIQ Software and Forth]

FromAndrew Haley <andrew29@littlepinkcloud.invalid>
Date2011-12-28 12:32 -0600
SubjectGPL [Was: CarrierIQ Software and Forth]
Message-ID<ybednQU2Y7JF_WbTnZ2dnUVZ_rKdnZ2d@supernews.com>
In reply to#8397
John Passaniti <john.passaniti@gmail.com> wrote:

> iOS developers can choose to statically link Apple's non-FOSS
> libraries to their application.  In that case, depending on which
> GPL license you're using, I can certainly see how both GPLv2 and
> GPLv3 could compel Apple to release source code if Apple didn't
> explicitly forbid such in their developer license agreement.

If you can see how a third party linking Apple's libraries to their
own code would cause Apple to have to release source code, please tell
us.  I can't see it at all.

> If you disagree, you are invited to point me to any authoritative
> statement from the GPL folk (not your interpretation, but their
> words) which says this can't happen.

And you are invited to explain how it *can* happen.  As far as I can
see the third party would be in breach of the GPL (2 or 3), and would
have to desist from shipping their software.  Apple wouldn't be a
party to the GPL violation.  How is it possible for the actions of a
third party to compel you to release your code?

Andrew.

[toc] | [prev] | [next] | [standalone]


#8407 — Re: GPL [Was: CarrierIQ Software and Forth]

FromJohn Passaniti <john.passaniti@gmail.com>
Date2011-12-28 12:07 -0800
SubjectRe: GPL [Was: CarrierIQ Software and Forth]
Message-ID<2aefef68-2340-48b5-920f-f71d5c10a2ce@o14g2000vbo.googlegroups.com>
In reply to#8400
On Dec 28, 1:32 pm, Andrew Haley <andre...@littlepinkcloud.invalid>
wrote:
> If you can see how a third party linking Apple's libraries
> to their own code would cause Apple to have to release
> source code, please tell us.  I can't see it at all.

http://www.gnu.org/licenses/gpl-faq.html#GPLIncompatibleLibs

I interpret that as saying that if I create an executable that
includes GPL code (my own code and/or other's code) *and* also
statically links to non-FOSS code (such as system libraries from
Apple), I need to get permission to release the source code from each
of the copyright holders of the non-FOSS code.  Apple isn't going to
provide that permission.

One central question here is if application libraries that are
statically linked to an executable qualify as "system libraries."  I
don't know the answer to that, but implicitly it seems that Apple
believes it is possible that such code (and possibly other code) might
not qualify.

> > If you disagree, you are invited to point me to any authoritative
> > statement from the GPL folk (not your interpretation, but their
> > words) which says this can't happen.
>
> And you are invited to explain how it *can* happen.

No, it doesn't work that way.  I'm not responsible for reaching into
the mind of Richard Stallman and friends and working out whatever
rational, subjective, reasoned, or arcane thought processes go into
their licenses.  They are the ones with the license and they are the
ones with legal muscle behind them who will defend the terms of their
license if they feel it is violated.  I am not a lawyer, not an expert
in copyright law, and not someone who even gives a damn.  I typically
run screaming away from all this monstrous complexity and
interpretation and worry and run into the warm and welcoming arms of
permissive licenses that are simpler, far less subjective, and let me
spend my time developing software and not reading licenses.

> As far as I can see the third party would be in breach of the
> GPL (2 or 3), and would have to desist from shipping their
> software.  Apple wouldn't be a party to the GPL violation.
> How is it possible for the actions of a third party to compel
> you to release your code?

I don't know.  That is the claim that has been made by others who
study this more seriously than I do.  Merely saying that you don't see
the mechanism doesn't mean it doesn't exist or couldn't exist.  If the
GPL (pick a version) can't possibly be used in situations like the one
being discussed (GPL'ed code statically linked to non-FOSS application
libraries), then there should be a clear and objective statement that
(1) makes unambiguous if application libraries fall under the "system
library exception" and (2) that such a situation is nonsense and
doesn't apply.  Alternatively, if there is an existing decision by the
GNU folk that is substantially similar to the situation described here
(or other related situations) that says there is no worry, then that
would also be acceptable.

[toc] | [prev] | [next] | [standalone]


#8412 — Re: GPL [Was: CarrierIQ Software and Forth]

FromBernd Paysan <bernd.paysan@gmx.de>
Date2011-12-28 23:19 +0100
SubjectRe: GPL [Was: CarrierIQ Software and Forth]
Message-ID<jdg4la$gq0$1@online.de>
In reply to#8407
John Passaniti wrote:
> http://www.gnu.org/licenses/gpl-faq.html#GPLIncompatibleLibs
> 
> I interpret that as saying that if I create an executable that
> includes GPL code (my own code and/or other's code) *and* also
> statically links to non-FOSS code (such as system libraries from
> Apple), I need to get permission to release the source code from each
> of the copyright holders of the non-FOSS code.  Apple isn't going to
> provide that permission.
> 
> One central question here is if application libraries that are
> statically linked to an executable qualify as "system libraries."  I
> don't know the answer to that, but implicitly it seems that Apple
> believes it is possible that such code (and possibly other code) might
> not qualify.

The answer to the question is fairly trivial.  Is the library part of 
iOS, as installed on the device (it then is probably a shared library)?  
Is it part of Xcode, the development environemnt used to create iOS Apps 
(you then probably do a static link)?  If yes, it is a system library.  
If Apple fears that some library or other might not qualify, it's 
because their lawyers are busy photoshopping competitors' products so 
that they look more similar to iOS devices, instead of giving advice to 
their team.

> I typically
> run screaming away from all this monstrous complexity and
> interpretation and worry and run into the warm and welcoming arms of
> permissive licenses that are simpler, far less subjective, and let me
> spend my time developing software and not reading licenses.

The GPL is rather simple (and yes, there is only one GPL, the versions 
do not change the intention, they only change the legal text).  The law 
is complex, as it is anglo-saxonian law, which does not give a shit 
about intent and principles, but wants it expressed like a computer 
program.  The GPL is the four freedoms.  That's it.  It's a single 
sentence in the preamble.  That's the GPL, that's all you as non-lawyer 
have to obey and think about.  Can the receiver of the code 
redistribute, study and modify your program, and does he know about 
this?  If he can, you follow the GPL.  If he can't, you don't.  Or you 
are a lawyer and found a loophole in the long legal part of the GPL.

Ah, and you are victim of FUD sprayed against the GPL.  The MS EULA is a 
monstrous complex beast where you should worry about your firstborn 
child if you don't read it.  The GPL has no such nonsense.  As long as 
you enable your receivers to have the four freedoms, there is nothing to 
worry about.  It's the cases where you want to take these freedoms away 
where you have to worry, because usually the GPL says "no", and maybe 
there's some loophole, because the GPL legal text is too old to cover 
that abuse (like DRM and signatures and such).

I suppose Apple uses something like signatures to prevent people 
installing apps from outside the app store.  And a developer account can 
only install self-compiled apps on the devices associated with that 
developer account.  It's not like Android, where you have your checkbox 
in the settings which activate third-party apps.  You have to jailbreak 
your iPhone to get there.  Given how popular jailbroken iPhones are, I'd 
probably say that it's ok for an open source app to say "your phone must 
be jailbroken".  Quite a number of open source Android apps require a 
rooted device, which is more or less the same.

> I don't know.  That is the claim that has been made by others who
> study this more seriously than I do.  Merely saying that you don't see
> the mechanism doesn't mean it doesn't exist or couldn't exist.  If the
> GPL (pick a version) can't possibly be used in situations like the one
> being discussed (GPL'ed code statically linked to non-FOSS application
> libraries), then there should be a clear and objective statement that
> (1) makes unambiguous if application libraries fall under the "system
> library exception" and (2) that such a situation is nonsense and
> doesn't apply.  Alternatively, if there is an existing decision by the
> GNU folk that is substantially similar to the situation described here
> (or other related situations) that says there is no worry, then that
> would also be acceptable.

There is no easy answer to that, because the world is complicated.  
Remember Windows 95?  When you installed MS Office on it, a whole bunch 
of system libraries were updated by the Office install, because they 
were so buggy that Office wouldn't run on an unpatched Windows 95.  
Microsoft however did not bother to fix Windows 95 itself with a service 
pack or such, Office was that service pack.  Would Windows 95 with MS 
Office installed qualify as "operating system"?  I would have some 
doubts, especially when you do this static linking thing, i.e. you as 
developer have such a system with Office installed, and you ship your 
binary plus source to some third party who does not have Office 
installed, and thinks Windows 95 is the operating system, and can't 
rebuild your program from the sources you delivered.

The easy part is: if your static library comes with either the OS or the 
development system as is, without anything strange, then it is a system 
library.

I must say that in the beginning (with the App store), Apple was very 
strongly opposed both against interpreters and against free software 
(copylefted software).  They had to backpedal almost completely on the 
interpreter issue, and they also had to significantly backpedal on the 
free software front.¹  I have the same feelings here as you have towards 
GPL'd software, I simply dare not to touch this ecosystem, this is too 
unfriendly for me, too dictatoric, too much at whim of whoever in the 
App store system decides.

¹) Note: Most of 2010, I was working for a supplyer for Apple, working 
on a battery monitoring system.  I had a bigForth program to visualize 
the battery charging and discharging cycles.  The people at Apple liked 
this program, and they said "we would like to have that as an app".  My 
response was that a) it was written in an interpreter, and b) the 
interpreter was free software, so nooooo way.  Soon after I said that, 
the app store restrictions on interpreters as such were lifted, and also 
the free software ban was converted into the restrictions as above.  
Still too strict for my taste, but a step in the right direction.

-- 
Bernd Paysan
"If you want it done right, you have to do it yourself"
http://bernd-paysan.de/

[toc] | [prev] | [next] | [standalone]


#8418 — Re: GPL [Was: CarrierIQ Software and Forth]

FromArnold Doray <invalid@invalid.com>
Date2011-12-29 08:38 +0000
SubjectRe: GPL [Was: CarrierIQ Software and Forth]
Message-ID<jdh8u3$ec1$1@dont-email.me>
In reply to#8407
On Wed, 28 Dec 2011 12:07:43 -0800, John Passaniti wrote:

> I interpret that as saying that if I create an executable that includes
> GPL code (my own code and/or other's code) *and* also statically links
> to non-FOSS code (such as system libraries from Apple), I need to get
> permission to release the source code from each of the copyright holders
> of the non-FOSS code.  Apple isn't going to provide that permission.

If this were true, it would be impossible to use the GPL on Windows 
applications.  

I believe the reverse is true -- you can't distribute closed-source 
excutables/libraries which have static links to GPL'd binaries. If you 
link to GPL software, your software must also be distributed under the 
GPL. This is precisely why the LGPL was created. 

Cheers,
Arnold

[toc] | [prev] | [next] | [standalone]


#8419 — Re: GPL [Was: CarrierIQ Software and Forth]

From"Rod Pemberton" <do_not_have@noavailemail.cmm>
Date2011-12-29 04:59 -0500
SubjectRe: GPL [Was: CarrierIQ Software and Forth]
Message-ID<jdhdiq$78l$1@speranza.aioe.org>
In reply to#8418
"Arnold Doray" <invalid@invalid.com> wrote in message
news:jdh8u3$ec1$1@dont-email.me...
> On Wed, 28 Dec 2011 12:07:43 -0800, John Passaniti wrote:
>
> > I interpret that as saying that if I create an executable that includes
> > GPL code (my own code and/or other's code) *and* also statically links
> > to non-FOSS code (such as system libraries from Apple), I need to get
> > permission to release the source code from each of the copyright holders
> > of the non-FOSS code.  Apple isn't going to provide that permission.
>
> If this were true, it would be impossible to use the GPL on Windows
> applications.
>

I'm not sure of the correct answer, but it may be dynamic linking, i.e., no
static linking between GPL and proprietary operating system code or
libraries ...

> I believe the reverse is true -- you can't distribute closed-source
> excutables/libraries which have static links to GPL'd binaries. If you
> link to GPL software, your software must also be distributed under the
> GPL. This is precisely why the LGPL was created.
>

LGPL is always touted as the solution to these issues.  History has proven
otherwise.  The basic premise of the LGPL is sound except for one critical
license term which nullifies it's premise entirely.  Let's disregard the
fact that Stallman and/or the FSF is trying to kill off all LGPL code with
GPLv3.  Let's also disregard Stallman's encoding of his clearly socialist
political beliefs in a FOSS license.  Let's also disregard the fact that the
FOSS is killing the commercial market for programmers and leaving them
without jobs or food.  A huge part of the problem with the LGPL is that LGPL
code can be converted to GPL code at will.  That's the critical license term
that is killing the LGPL.  I.e., the license terms for the LGPL'd source can
be changed by a third party without the original copyright holder's
permission.  The third party can change the license terms to the more
restrictive GPL license terms.  License "rebranding" is a huge negative for
a supposedly FOSS license.  Another part of the problem with the LGPL is
that the biggest library claiming to be LGPL, i.e., GNU's own GLIBC, is not
purely LGPL, so the entire GLIBC library must be treated as GPL'd, instead
of LGPL'd, to prevent license non-compliance.  There is a bunch of strictly
GPLv2 code in the supposedly LGPL'd GLIBC.  There is also a bunch of
non-GPL, copyrighted and licensed, FOSS code in there too.  That's why there
are three (3) license files are included with GLIBC: COPYING, COPYING.LIB,
and LICENSES.  COPYING.LIB is the LGPL for much of the code in the library.
COPYING is for the strictly GPL files in the library.  LICENSES includes the
FOSS licenses for the six (6) other copyright holders besides GNU ...  You
read that correctly: SIX!  This is major FOSS license "pollution".  AIUI,
because GLIBC is not purely LGPL, all current non-Unix/Linux/POSIX ports of
the GCC compiler use their own, non-GPL, custom C library, e.g., DJGPP,
MinGW, CygWin.  AIR, (Hey, I looked up GLIBC ...) the GCC compiler has some
license exceptions preventing violations of it's GPL terms in non-GPL
situations.  A huge part of the problem with GPL is that it does not have
far less rectrictive license terms like other major FOSS licenses, e.g., BSD
or MIT.  Another huge part of the problem with the GPL is that the original
copyright holder is not required to forfit their copyrights, or is not
prohibited from claiming copyrights at a later time.  The original copyright
holder can close source the code at a later date.  The only real solution to
most of the FOSS license issues is to eliminate the issue of copyright
entirely, i.e., Public Domain, but that apparently is not recognized as a
legal solution everywhere and is supposedly legally untested even in the
USA.


Rod Pemberton


[toc] | [prev] | [next] | [standalone]


#8424 — Re: GPL [Was: CarrierIQ Software and Forth]

FromBernd Paysan <bernd.paysan@gmx.de>
Date2011-12-29 14:04 +0100
SubjectRe: GPL [Was: CarrierIQ Software and Forth]
Message-ID<jdhog6$umt$1@online.de>
In reply to#8419
Rod Pemberton wrote:
> Let's also disregard the fact that the
> FOSS is killing the commercial market for programmers and leaving them
> without jobs or food.

Let's disregard that RedHat managed to cross the 1 billion dollar 
revenue in 2011.  No job, no food?  I'm sure with a 1 billion dollar 
revenue, food supply for people who work at RedHat can't be that bad; 
the last RH employee I saw, appeared to be well fed (that was Andrew 
Haley).  They do that entirely with open source software, of which a 
significant portion and the core of RedHat's business model is GPL 
software.

What's true is that the LGPL provides an "additional permission" which 
can be taken away by the next contributor (it cannot be retroactively be 
taken away from the original distribution, if you fear that - you may 
pass it on without the LGPL permissions, but anybody who wants still can 
get the LGPL'd sources from the original author, and the GPL is very 
clear that the license is between you and the original authors, not with 
some man-in-the-middle).  This possible removal is necessary to make it 
possible to mix LGPL with GPL programs, it's a compatibility issue - if 
it wasn't there, the LGPL would be GPL-incompatible, and that would be 
really, really bad.

-- 
Bernd Paysan
"If you want it done right, you have to do it yourself"
http://bernd-paysan.de/

[toc] | [prev] | [next] | [standalone]


#8437 — Re: GPL

FromAleksej Saushev <asau@inbox.ru>
Date2011-12-30 00:12 +0400
SubjectRe: GPL
Message-ID<87pqf7p279.fsf@inbox.ru>
In reply to#8424
Bernd Paysan <bernd.paysan@gmx.de> writes:

> Rod Pemberton wrote:
>> Let's also disregard the fact that the
>> FOSS is killing the commercial market for programmers and leaving them
>> without jobs or food.
>
> Let's disregard that RedHat managed to cross the 1 billion dollar 
> revenue in 2011.

Have you got any other example besides RedHat? How many of them?
Bringing extremely rare exception to example doesn't prove anything.

Besides, what is the structure of revenue? Does it come from programming
or from system administration?


-- 
HE CE3OH...

[toc] | [prev] | [next] | [standalone]


#8442 — Re: GPL

Fromcas_news@strotmann.de (Carsten Strotmann (Usenet))
Date2011-12-29 21:48 +0100
SubjectRe: GPL
Message-ID<87oburxfxb.fsf@csgate4.strotmann.de>
In reply to#8437
Aleksej Saushev <asau@inbox.ru> writes:

> Bernd Paysan <bernd.paysan@gmx.de> writes:
>
>> Rod Pemberton wrote:
>>> Let's also disregard the fact that the
>>> FOSS is killing the commercial market for programmers and leaving them
>>> without jobs or food.
>>
>> Let's disregard that RedHat managed to cross the 1 billion dollar 
>> revenue in 2011.
>
> Have you got any other example besides RedHat? How many of them?
> Bringing extremely rare exception to example doesn't prove anything.
>
> Besides, what is the structure of revenue? Does it come from programming
> or from system administration?

Well, I make a good living around open source software with training and
some programming.

I also know a good number of companies that also make good money with
open source software:

* Univention: http://www.univention.de/
* MariaDB (and previously MySQL DB): http://askmonty.org/
* Freiheit.com: http://www.freiheit.com/ (with bookzilla.de one of the
 main sponsors of FSFe)
* LinuxHotel: linuxhotel.de (Open Source Training)
* PowerDNS - got successful after open-sourcing a former closed source
product (www.powerdns.com)
* Internet Systems Consortium - BIND and ISC DHCPd, among other software
* NlNetLabs 
* Zimbra 
* Mozilla
* Canonical (Ubuntu Linux)
* Digium (Asterisk PBX)
.....

the list goes on and on. There is as much money on open source software
than there is in closed source. It just needs some nderstanding of the
customers needs and a slightly different mindset.

-- Carsten

[toc] | [prev] | [next] | [standalone]


Page 2 of 6 — ← Prev page 1 [2] 3 4 5 6  Next page →

Back to top | Article view | comp.lang.forth


csiph-web