public inbox for cygwin@cygwin.com
 help / color / mirror / Atom feed
* RE: To whom do we contribute API header/
@ 1998-07-15  8:29 Marcus Brown
  0 siblings, 0 replies; 9+ messages in thread
From: Marcus Brown @ 1998-07-15  8:29 UTC (permalink / raw)
  To: Larry Hall, Geoffrey Noer, Robin Crampton, phawkins, gnu-win32

Isn't there a way (or a process that was proposed) that would allow you
to "patch" or "massage" the MS Platform SDK header files to use with
cygwin32?  AFAIK the Platform SDK is freely available from Microsoft,
and this would seem to be a logical way to go instead of recreating each
header file.  I could see problems with import libraries, but maybe it's
possible to incorporate the ms-compat binutils or use another method.
I'm not certain, maybe this is more on the mingw32 side of cygwin32...

Marcus

-----Original Message-----
From: Larry Hall [ mailto:lhall@rfk.com ]
Sent: Tuesday, July 14, 1998 8:48 AM
To: Geoffrey Noer; Robin Crampton; phawkins@dynamite.com.au;
gnu-win32@cygnus.com
Subject: Re: To whom do we contribute API header/


At 07:48 AM 7/13/98 +0000, Geoffrey Noer wrote:
>On Fri, Jul 10, 1998 at 11:05:58AM +0000, Robin Crampton wrote:
>[...]
>> How different do the cygwin32 Windows header files have to be from   
>> Microsofts to not infringe copyright, or whatever?  Can we just strip
out   
>> all the comments and expand all the typedefs (well, obviously not all
the   
>> typedefs, just the simple ones such as DWORD) ?  Do we even need to
do   
>> that?  I imagine it's hard to enforce copyright on data structure   
>> definitions and function prototypes.
>
>To the best of my knowledge, all of our Win32 headers were generated
>by looking at documentation and then adding equivalent definitions in
>ours.  It is not ok to copy sections from Microsoft headers into ours.
>I'll try to investigate this issue some more and send additional mail
>if there's anything more to say...
>
>> I've got my-mmsystem.h which I'll gladly donate, but at the moment
it's   
>> just Microsofts with my hacks.  
>
>That's not going to work I'm afraid.
>

Perhaps this does need some clarification then.  You mentioned that the
Win32
headers were generated by looking at "documentation" and then adding 
"equivalent" definitions.  What documentation is this and how
"equivalent"
is equivalent?  I don't want to get into too much legalese but if I were

going to try to do some of this, I would probably consult MSDN for 
documentation to start or some other good Win32 source, assuming I was 
going to ignore the headers themselves in VC++.  It seems highly likely
to 
me that any source I choose is going to be from MS.  So is it OK to look
at
documentation like this?  Personally, I don't see that much difference 
between the "cut-and-paste" operations that come from this activity and 
that which would be involved with using the headers "directly" but I get

the impression that I'm missing some important point in this regard.
Geoff,
can you clarify that point for me/us?

--
-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request@cygnus.com" with one line of text: "help".

^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: To whom do we contribute API header/
  1998-07-15 23:41       ` Geoffrey Noer
@ 1998-07-18  0:01         ` Marty Leisner
  0 siblings, 0 replies; 9+ messages in thread
From: Marty Leisner @ 1998-07-18  0:01 UTC (permalink / raw)
  To: Geoffrey Noer; +Cc: root, gnu-win32

From what I've seen (except for a few proprietary products
like source navigator) Cygnus sells service and support, the
software is still GPL.



-- 
marty
leisner@sdsp.mc.xerox.com  
The Feynman problem solving Algorithm
        1) Write down the problem
        2) Think real hard
        3) Write down the answer
                Murray Gell-mann in the NY Times




-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request@cygnus.com" with one line of text: "help".

^ permalink raw reply	[flat|nested] 9+ messages in thread

* RE: To whom do we contribute API header/
@ 1998-07-16  1:49 Robin Crampton
  0 siblings, 0 replies; 9+ messages in thread
From: Robin Crampton @ 1998-07-16  1:49 UTC (permalink / raw)
  To: gnu-win32

>Attention: the copyright is now mine. If cygnus wants to include this   
header
>files with their CDK they should pay me money for it, since they sell   
their
>stuff too.

Storm clouds gathering....  

-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request@cygnus.com" with one line of text: "help".

^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: To whom do we contribute API header/
  1998-07-15  2:38     ` root
@ 1998-07-15 23:41       ` Geoffrey Noer
  1998-07-18  0:01         ` Marty Leisner
  0 siblings, 1 reply; 9+ messages in thread
From: Geoffrey Noer @ 1998-07-15 23:41 UTC (permalink / raw)
  To: root, gnu-win32

> You never answered anything about my work with the headers, ignoring all the
> work I do as a matter of course.

Actually, I did look into it a bit.  I forwarded your mail to a couple
of gcc engineers who thought that your changes were probably good
ones to make.

> I have been working two years now in this headers. They can be
> downloaded with the lcc system from:
> 
> http://www.cs.virginia.edu/~lcc-win32 .

Done.  I will look into it again.  I guess someone else could always
rewrite them in a similar manner.

> Attention: the copyright is now mine. If Cygnus wants to include
> this header files with their CDK they should pay me money for it,
> since they sell their stuff too.

Aha!  And that's why those changes never made it anywhere...  But it's
certainly your choice to make.

Personally, I do not have a problem with Cygnus charging money for
some things if that means that Cygnus can hire more engineers to do
work that ends up benefitting the free software community.  (Two
thirds of our employees are software engineers, many of whom spend a
substantial percentage of their working hours improving free
software).  But don't take my word for it.  Take a look at a ChangeLog
file from any of the GNU development tools and look for "@cygnus.com".

Altruistic motives aside, profit is also a goal but it's definitely
not the only one.

That's my outlook.  You're certainly welcome to see things
differently (and do as far as I can tell)...

Regards,

Geoffrey Noer
noer@cygnus.com
-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request@cygnus.com" with one line of text: "help".

^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: To whom do we contribute API header/
@ 1998-07-15  6:41 Fred Reimer
  0 siblings, 0 replies; 9+ messages in thread
From: Fred Reimer @ 1998-07-15  6:41 UTC (permalink / raw)
  To: Geoffrey Noer, Robin Crampton, phawkins, gnu-win32, Larry Hall

I'm a little confused why we can't just use the include files that come with
the Platform SDK from Microsoft.  They seem to work fine using Mikey's
link-compat distribution, which includes patches to get them to work.  Since
everyone can download the Platform SDK from MS directly, there would be no
reason to redistribute MS's "intellectual property."  All we would have to
do is distribute the patch files to make them work with Cygwin32.  If anyone
is concerned that "standard" patch files contain a certain amount of the
original files I'm sure we could come up with a patch program that does not
rely on ANY original text but rather uses offsets to delete and add text.
We could also include a checksum on the original in the patch to make sure
we are patching the right "version."  Am I missing something here?  (I admit
that I did not read MS's "license" in the Platform SDK so don't know if they
prohibit this, so take it easy on me!).

Fred Reimer
Eclipsys Corporation


-----Original Message-----
From: Larry Hall <lhall@rfk.com>
To: Geoffrey Noer <noer@cygnus.com>; Robin Crampton
<rcrampton@armature.com>; <phawkins@dynamite.com.au>; <gnu-win32@cygnus.com>
Date: Wednesday, July 15, 1998 2:12 AM
Subject: Re: To whom do we contribute API header/


>At 07:48 AM 7/13/98 +0000, Geoffrey Noer wrote:
>>On Fri, Jul 10, 1998 at 11:05:58AM +0000, Robin Crampton wrote:
>>[...]
>>> How different do the cygwin32 Windows header files have to be from
>>> Microsofts to not infringe copyright, or whatever?  Can we just strip
out
>>> all the comments and expand all the typedefs (well, obviously not all
the
>>> typedefs, just the simple ones such as DWORD) ?  Do we even need to do
>>> that?  I imagine it's hard to enforce copyright on data structure
>>> definitions and function prototypes.
>>
>>To the best of my knowledge, all of our Win32 headers were generated
>>by looking at documentation and then adding equivalent definitions in
>>ours.  It is not ok to copy sections from Microsoft headers into ours.
>>I'll try to investigate this issue some more and send additional mail
>>if there's anything more to say...
>>
>>> I've got my-mmsystem.h which I'll gladly donate, but at the moment it's
>>> just Microsofts with my hacks.
>>
>>That's not going to work I'm afraid.
>>
>
>Perhaps this does need some clarification then.  You mentioned that the
Win32
>headers were generated by looking at "documentation" and then adding
>"equivalent" definitions.  What documentation is this and how "equivalent"
>is equivalent?  I don't want to get into too much legalese but if I were
>going to try to do some of this, I would probably consult MSDN for
>documentation to start or some other good Win32 source, assuming I was
>going to ignore the headers themselves in VC++.  It seems highly likely to
>me that any source I choose is going to be from MS.  So is it OK to look at
>documentation like this?  Personally, I don't see that much difference
>between the "cut-and-paste" operations that come from this activity and
>that which would be involved with using the headers "directly" but I get
>the impression that I'm missing some important point in this regard.
Geoff,
>can you clarify that point for me/us?
>
>
>Larry Hall                              lhall@rfk.com
>RFK Partners, Inc.                      (781) 239-1053
>8 Grove Street                          (781) 239-1655 - FAX
>Wellesley, MA  02181                    http://www.rfk.com
>-
>For help on using this list (especially unsubscribing), send a message to
>"gnu-win32-request@cygnus.com" with one line of text: "help".
>

-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request@cygnus.com" with one line of text: "help".

^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: To whom do we contribute API header/
  1998-07-14  8:29   ` Larry Hall
@ 1998-07-15  2:38     ` root
  1998-07-15 23:41       ` Geoffrey Noer
  0 siblings, 1 reply; 9+ messages in thread
From: root @ 1998-07-15  2:38 UTC (permalink / raw)
  To: gnu-win32

Mr Noer: 			RE: Header files:

I am the author of lcc-win32, and I have been working with the header files for
windows since two years now.

At the beginning, I mailed to you and to Scott Christley my modificatkions, and
sent regular update messages to this list.

But then, hearing the silence that always followed up my posts, I stopped that
activity. Scott was (and is probably still) very busy with his Gnustep project, and he didn't have the time to pursue this.

You never answered anything about my work with the headers, ignoring all the
work I do as a matter of course.

Nobody in this list answered anything either, so I just continued working and
didn't give a damm about crying to people that do not want to hear what I say.

This more than two years work with the headers have produced a very complete
set of headers for windows. I have written all of them, since the format Scott 
used has proved inadequate for my needs.

What are my needs?

I need compiler speed. THEN I must minimize I/O, i.e. reduce the size of the
headers to a bare minimum. I have rewritten all Scott's headers, The speedup
in compilation is ENORMOUS. This means in fact, that if I compile the following
file:

#include <windows.h>

If I do then:

lcc  test.c

lcc compiles this file in 1.09 seconds.

If I call Microsoft compiler and compile this file
cl -c test.c

It will take 1.6 seconds, i.e. almost 50% more. Is it a slower compiler then?
Is lcc really so fast?

If I do
cl -c -I\lcc\include test.c

I.e. if I direct Microsoft's compiler to use lcc's header files, the speed
of cl is incresed by more than 400% falling down to 0.4 seconds!!!!!!!

I have produced a set of header files that increases compilation speed by an 
enormous factor. This will surely be the case with gcc, since gcc is much
slower than lcc or cl.


I have been working two years now in this headers. They can be downloaded with
the lcc system from:

http://www.cs.virginia.edu/~lcc-win32 .

Attention: the copyright is now mine. If cygnus wants to include this header
files with their CDK they should pay me money for it, since they sell their
stuff too.

For individual users however, there are absolutely NO charges to pay. Neither 
do I gain any rights to the software compiled using lcc or the headers. For
individual users there are NO strings attached. This is the spirit of free 
software.

The license is not GPL, neither LGPL, but JPL: Jacob's public license. :-)

-- 
Jacob Navia	Logiciels/Informatique
41 rue Maurice Ravel			Tel 01 48.23.51.44
93430 Villetaneuse 			Fax 01 48.23.95.39
France
-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request@cygnus.com" with one line of text: "help".

^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: To whom do we contribute API header/
  1998-07-13  3:13 ` Geoffrey Noer
@ 1998-07-14  8:29   ` Larry Hall
  1998-07-15  2:38     ` root
  0 siblings, 1 reply; 9+ messages in thread
From: Larry Hall @ 1998-07-14  8:29 UTC (permalink / raw)
  To: Geoffrey Noer, Robin Crampton, phawkins, gnu-win32

At 07:48 AM 7/13/98 +0000, Geoffrey Noer wrote:
>On Fri, Jul 10, 1998 at 11:05:58AM +0000, Robin Crampton wrote:
>[...]
>> How different do the cygwin32 Windows header files have to be from   
>> Microsofts to not infringe copyright, or whatever?  Can we just strip out   
>> all the comments and expand all the typedefs (well, obviously not all the   
>> typedefs, just the simple ones such as DWORD) ?  Do we even need to do   
>> that?  I imagine it's hard to enforce copyright on data structure   
>> definitions and function prototypes.
>
>To the best of my knowledge, all of our Win32 headers were generated
>by looking at documentation and then adding equivalent definitions in
>ours.  It is not ok to copy sections from Microsoft headers into ours.
>I'll try to investigate this issue some more and send additional mail
>if there's anything more to say...
>
>> I've got my-mmsystem.h which I'll gladly donate, but at the moment it's   
>> just Microsofts with my hacks.  
>
>That's not going to work I'm afraid.
>

Perhaps this does need some clarification then.  You mentioned that the Win32
headers were generated by looking at "documentation" and then adding 
"equivalent" definitions.  What documentation is this and how "equivalent"
is equivalent?  I don't want to get into too much legalese but if I were 
going to try to do some of this, I would probably consult MSDN for 
documentation to start or some other good Win32 source, assuming I was 
going to ignore the headers themselves in VC++.  It seems highly likely to 
me that any source I choose is going to be from MS.  So is it OK to look at
documentation like this?  Personally, I don't see that much difference 
between the "cut-and-paste" operations that come from this activity and 
that which would be involved with using the headers "directly" but I get 
the impression that I'm missing some important point in this regard.  Geoff,
can you clarify that point for me/us?


Larry Hall                              lhall@rfk.com
RFK Partners, Inc.                      (781) 239-1053
8 Grove Street                          (781) 239-1655 - FAX
Wellesley, MA  02181                    http://www.rfk.com
-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request@cygnus.com" with one line of text: "help".

^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: To whom do we contribute API header/
  1998-07-10 12:03 Robin Crampton
@ 1998-07-13  3:13 ` Geoffrey Noer
  1998-07-14  8:29   ` Larry Hall
  0 siblings, 1 reply; 9+ messages in thread
From: Geoffrey Noer @ 1998-07-13  3:13 UTC (permalink / raw)
  To: Robin Crampton, phawkins, gnu-win32

On Fri, Jul 10, 1998 at 11:05:58AM +0000, Robin Crampton wrote:
[...]
> How different do the cygwin32 Windows header files have to be from   
> Microsofts to not infringe copyright, or whatever?  Can we just strip out   
> all the comments and expand all the typedefs (well, obviously not all the   
> typedefs, just the simple ones such as DWORD) ?  Do we even need to do   
> that?  I imagine it's hard to enforce copyright on data structure   
> definitions and function prototypes.

To the best of my knowledge, all of our Win32 headers were generated
by looking at documentation and then adding equivalent definitions in
ours.  It is not ok to copy sections from Microsoft headers into ours.
I'll try to investigate this issue some more and send additional mail
if there's anything more to say...

> I've got my-mmsystem.h which I'll gladly donate, but at the moment it's   
> just Microsofts with my hacks.  

That's not going to work I'm afraid.

-- 
Geoffrey Noer
noer@cygnus.com
-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request@cygnus.com" with one line of text: "help".

^ permalink raw reply	[flat|nested] 9+ messages in thread

* RE: To whom do we contribute API header/
@ 1998-07-10 12:03 Robin Crampton
  1998-07-13  3:13 ` Geoffrey Noer
  0 siblings, 1 reply; 9+ messages in thread
From: Robin Crampton @ 1998-07-10 12:03 UTC (permalink / raw)
  To: phawkins, gnu-win32, noer

>Such changes should be sent to cygwin32-developers@cygnus.com.

What's the legal position?

How different do the cygwin32 Windows header files have to be from   
Microsofts to not infringe copyright, or whatever?  Can we just strip out   
all the comments and expand all the typedefs (well, obviously not all the   
typedefs, just the simple ones such as DWORD) ?  Do we even need to do   
that?  I imagine it's hard to enforce copyright on data structure   
definitions and function prototypes.

I've got my-mmsystem.h which I'll gladly donate, but at the moment it's   
just Microsofts with my hacks.  

-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request@cygnus.com" with one line of text: "help".

^ permalink raw reply	[flat|nested] 9+ messages in thread

end of thread, other threads:[~1998-07-18  0:01 UTC | newest]

Thread overview: 9+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
1998-07-15  8:29 To whom do we contribute API header/ Marcus Brown
  -- strict thread matches above, loose matches on Subject: below --
1998-07-16  1:49 Robin Crampton
1998-07-15  6:41 Fred Reimer
1998-07-10 12:03 Robin Crampton
1998-07-13  3:13 ` Geoffrey Noer
1998-07-14  8:29   ` Larry Hall
1998-07-15  2:38     ` root
1998-07-15 23:41       ` Geoffrey Noer
1998-07-18  0:01         ` Marty Leisner

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).