public domain boot sector

[kerravon]

Hi.

As part of the public domain PDOS, I need to
have a public domain boot sector. I have
written one of those already, but it is bare
bones, and only works on floppy drive A. I'm
doing my work on Bochs hard disks now.

Can someone take a look at what I've got so
far and suggest what changes are required?
Or is there an alternate boot sector that is
public domain?

One thing I know is that I should be using DL
to get the drive letter instead of hardcoding
to 0 (A).

Thanks. Paul.



; pbootsec.asm - pdos boot sector
;
; This program written by Paul Edwards
; Released to the public domain

; This code will be loaded to location 07C00h by the BIOS
; We set the stack at 07C00h and make it grow down

; Registers at entry are:
; CS:IP = 07c00H, probably 0000:7c00
; DL = drive number
; nothing else guaranteed

; We should store the drive number in BPB[25], since
; pload is expecting it there. pload also expects
; that the CS:IP will have an IP of 0 when it is called.
; It also expects some sort of stack to be set up.

% .model memodel

_DATA segment word public 'DATA'
_DATA ends
_BSS segment word public 'BSS'
_BSS ends

_TEXT segment word public 'CODE'

org 0100h
top:

public start
start proc

; jump around buffer
jmp bypass

sysname db 'PDOS x.x'
bpb db 51 dup(?)

; new disk parameter table
dpt db 11 dup(?)


bypass:
mov ax, 0
mov ss, ax
mov sp, 07c00h

push cs
pop ds

; don't bother setting new dpt, it doesn't appear to be required
;call newdpt
;call reset
call loadsecs

;jmp 0070h:0000h
mov ax, 0070h
push ax
mov ax, 0
push ax
retf

start endp


; newdpt - set up new dpt
newdpt proc

; old dpt is located at interrupt 1E = 078h
push es
push bx
push ax
push cx
push dx
push ds

mov ax, 0
mov es, ax
mov bx, 078h
mov ax, es:[bx + 2]
mov ds, ax
mov ax, es:[bx]
mov dx, ax

mov ax, 0
mov es, ax
mov bx, 07C3Eh
mov cx, 0
newdpt_loop:
cmp cx, 11
jge newdpt_end
push bx
mov bx, dx
mov al, [bx]
pop bx
mov es:[bx], al
inc dx
inc bx
inc cx
jmp newdpt_loop
newdpt_end:

mov ax, 0
mov es, ax
mov bx, 078h
mov word ptr es:[bx + 2], 0
mov word ptr es:[bx], 07C3Eh
mov bx, 07C3Eh

; set offset 4 to 18 (should be number of sectors per track)
;mov byte ptr es:[bx + 4], 18

pop ds
pop dx
pop cx
pop ax
pop bx
pop es
ret

newdpt endp


; loadsecs - load sectors
loadsecs proc

; read 3 sectors
mov bx, 0700h
mov cl, 010h
call saferead
add bx, 0200h
add cl, 1
call saferead
add bx, 0200h
add cl, 1
call saferead
ret

loadsecs endp


; saferead
saferead proc

mov ax, 0
jmp first
retry:
call reset
inc ax
cmp ax, 05h
jge giveup
first:
call secread
jnz retry
giveup:
ret

saferead endp


; reset
reset proc

mov ah, 00h ; function
mov dl, 00h ; drive

int 013h

ret

reset endp


; secread - bx = offset, cl = sector
secread proc

push bx
push cx
push ax

mov ax, 0h
mov es, ax ; buffer segment
mov ah, 02h ; code
mov al, 01h ; 1 sector
mov ch, 00h ; track 0
mov dh, 01h ; head
mov dl, 00h ; drive

int 013h

pop ax
pop cx
pop bx
ret

secread endp


org 02feh
lastword dw 0aa55h

_TEXT ends

end top

[Alexei A. Frounze]

Do you want me to release my FAT12, FAT16 and FAT32 bootsectors under
FreeBSD license?

The FAT12 and FAT16 versions are available in http://alexfru.narod.ru/os/fat/fat-2006-12-03.zip
(FAT12.ASM and FAT16.ASM under _boot). The code is for NASM.

Alex

[kerravon]

On Tuesday, April 16, 2013 1:14:04 PM UTC+10, Alexei A. Frounze wrote:
> Do you want me to release my FAT12, FAT16 and FAT32 bootsectors under
> FreeBSD license?

Thanks for the offer, but I'm after public
domain, not FreeBSD.

Another thing I realized is that I'm hoping
to get a version of the boot sector that works
for all FAT types because it just needs to load
3 consecutive sectors. With 3 sectors loaded,
the C code is able to come into play.

Also, I am confused about something. The boot
loader is designed to use BosDiskSectorRead,
but the FAT routines are designed to use
BosDiskSectorRLBA:

/* BosDiskSectorRead - read using track etc Int 13h Function 02h */
/* BosDiskSectorRLBA - read using LBA Int 13h Function 42h */

I cannot remember why I ended up using both. How
should this ideally work? Should the boot sector
be using RLBA instead?

Thanks. Paul.

[Alexei A. Frounze]

On Apr 15, 10:17 pm, kerravon <kerra...@w3.to> wrote:
> On Tuesday, April 16, 2013 1:14:04 PM UTC+10, Alexei A. Frounze wrote:
> > Do you want me to release my FAT12, FAT16 and FAT32 bootsectors under
> > FreeBSD license?
>
> Thanks for the offer, but I'm after public
> domain, not FreeBSD.

Why?

> Another thing I realized is that I'm hoping
> to get a version of the boot sector that works
> for all FAT types because it just needs to load
> 3 consecutive sectors. With 3 sectors loaded,
> the C code is able to come into play.

My bootsectors properly read a file from the FAT's root irrespective
of the length (to a reasonable limit, of course, you wouldn't want to
try loading 500+ KB files because there may be no memory for that) and
irrespective of FAT fragmentation and support execution of .COM
and .EXE files directly, out of the box.

And I see no reason why one would want to have a single universal boot
sector supporting all three FAT formats. When you format the disk you
write the bootsector appropriate for the FAT format and forget about
it.

> Also, I am confused about something. The boot
> loader is designed to use BosDiskSectorRead,
> but the FAT routines are designed to use
> BosDiskSectorRLBA:
>
> /* BosDiskSectorRead - read using track etc Int 13h Function 02h */
> /* BosDiskSectorRLBA - read using LBA Int 13h Function 42h */
>
> I cannot remember why I ended up using both. How
> should this ideally work? Should the boot sector
> be using RLBA instead?
>
> Thanks.  Paul.

Function 2 has a limit on the size of the disk, the limit is baked
into the CHS encoding scheme.
Function 42h is what you want to use on large drives to avoid the CHS
limit.

Alex

[Rod Pemberton]

"kerravon" <kerr...@w3.to> wrote in message
news:e5d08e1b-cdc2-48c7...@googlegroups.com...

>
> Another thing I realized is that I'm hoping
> to get a version of the boot sector that works
> for all FAT types because it just needs to load
> 3 consecutive sectors. With 3 sectors loaded,
> the C code is able to come into play.
>
> Also, I am confused about something. The boot
> loader is designed to use BosDiskSectorRead,
> but the FAT routines are designed to use
> BosDiskSectorRLBA:
>

I'm assuming "Bos" is shorthand for "Bios" or a misspelling ...

> /* BosDiskSectorRead - read using track etc Int 13h Function 02h
> */
> /* BosDiskSectorRLBA - read using LBA Int 13h Function 42h */
>
> I cannot remember why I ended up using both.

...

> How should this ideally work?

No idea.

I've implemented some boot sectors, but not any boot sector
loaders.

> Should the boot sector be using RLBA instead?
>

Perhaps, you originally planned to check if LBA was available
first ... ?

Looking at RBIL's info on each function only tells me that one of
them seems to be a bit "easier" to use.

Obviously, there was a point in time when the MS Int 13h functions
weren't available, e.g., prior to IBM/MS Int 13h extensions of
1992.

You're using LBA which implies LBA is available. So, if LBA is
available, I don't see any reason to not use it...

LBA has been around for a long time now, i.e., at least since
ATA-2 in 1996 introduced 28-bit LBA. 48-bit LBA came with
ATA/ATAPI-6 in 2001. I'm not sure if an earlier version of LBA
was standardized prior to ATA-2.

Which of the two calls reduces overall code space, code setup, and
complexity? If you're only loading 3 sectors, I'd assume
compactness is a factor.


Rod Pemberton

[Rod Pemberton]

"kerravon" <kerr...@w3.to> wrote in message

news:3df10ff1-46d4-4fc2...@googlegroups.com...

>
> As part of the public domain PDOS, I need to
> have a public domain boot sector. I have
> written one of those already, but it is bare
> bones, and only works on floppy drive A. I'm
> doing my work on Bochs hard disks now.
>
> Can someone take a look at what I've got so
> far and suggest what changes are required?
> Or is there an alternate boot sector that is
> public domain?
>

Well, I've found three supposedly Public Domain boot sector's in
my collection of PD code. One is by you, probably the one you
just mentioned. One is by Kristof Benes (Cribix), and one in
Chris Giese's code collection. Chris' has code copyrighted by
Linus Torvalds, Bruce Evans etc, without mention of what license,
or if any of the code is actually Chris' or not... I.e., it might
just be a conversion from one assembler (GNU GAS) to another
(NASM). Kristof's has A20 enable code copyrighted by J. Andrew
McLaughlin (Visopsys) which is licensed as "free to use".
Otherwise, I haven't explicitly searched for true PD bootsectors
or bootloaders. I was mostly after C code at the time. There
might be a few in the archives Usenet or on the Internet.

> One thing I know is that I should be using DL
> to get the drive letter instead of hardcoding
> to 0 (A).
>
> Thanks. Paul.
>

...

> ; pbootsec.asm - pdos boot sector
> ;
> ; This program written by Paul Edwards
> ; Released to the public domain
>
> ; This code will be loaded to location 07C00h by the BIOS
> ; We set the stack at 07C00h and make it grow down
>
> ; Registers at entry are:
> ; CS:IP = 07c00H, probably 0000:7c00
> ; DL = drive number
> ; nothing else guaranteed
>
> ; We should store the drive number in BPB[25], since
> ; pload is expecting it there. pload also expects
> ; that the CS:IP will have an IP of 0 when it is called.
> ; It also expects some sort of stack to be set up.
>
> % .model memodel
>
> _DATA segment word public 'DATA'
> _DATA ends
> _BSS segment word public 'BSS'
> _BSS ends
>
> _TEXT segment word public 'CODE'
>
> org 0100h

I thought 100h was for DOS .com's ... ;-) Did you want 7C00h
here?

> top:
>
> public start
> start proc
>
> ; jump around buffer
> jmp bypass
>

It's my understanding that for compatibility with MS-DOS and Win98
etc, that the "jmp bypass" needs to be a short jump, i.e., two
byte encoding, followed by a NOP instruction, for a total of three
bytes. This appears to be for MASM. I'm not sure how to encode
the short jump with MASM.

> sysname db 'PDOS x.x'

Ok. Assuming that's 8 characters. It looks like there's one
space in there here.

> bpb db 51 dup(?)
>

Ok. You seem to be allocating the correct amount.

Why do you allocate a BPB without setting any values in the BPB?

Either you need the BPB and it's values, or you don't. AIUI, the
JMP instruction and BPB is for MS-DOS, or for compatibility with
MS-DOS and Windows.

"Detailed Explanation of FAT Boot Sector"
http://support.microsoft.com/kb/140418

There is also an in-depth FAT document from MS as well as an
international standard for FAT12/16.

> ; new disk parameter table
> dpt db 11 dup(?)
>

Sorry, I'm not familiar with the DPT.

>
> bypass:
> mov ax, 0
> mov ss, ax
> mov sp, 07c00h
>

Good. IIRC, AMD manuals said to either use LSS instruction, or to
load SS and SP together to perform instruction locking during
loading.

Of course, the stack is right at the start of your code, so you
definately don't want to have any errant pops and pushes... I've
seen some people put it at 600h but that's just above the end of
the IVT.

> push cs
> pop ds

Good. Stack is setup first. Sigh, I may have to review some of
my code on that... I usually PUSH CS, POP DS first.

You're not explicitly setting CS and IP via a far jump or RETF, so
it's possible that they're typically CS=0 and IP=7C00h, but
sometimes may be CS=07C0h and IP=0, or other combinations. But, I
think you're aware of that issue.

> ; don't bother setting new dpt, it doesn't appear to be required
> ;call newdpt
> ;call reset
> call loadsecs
>

...

> ;jmp 0070h:0000h
> mov ax, 0070h
> push ax
> mov ax, 0
> push ax
> retf

Why are you using RETF instead of the absolute JMP that's
commented out? Is this a MASM limitation? Isn't there a DB form
for MASM to do that? I might've used RETF earlier to set CS:IP...

> start endp
>
>
> ; newdpt - set up new dpt
> newdpt proc
>
> ; old dpt is located at interrupt 1E = 078h
> push es
> push bx
> push ax
> push cx
> push dx
> push ds
>

With that many pushes, I'd look into whether PUSHA was usable here
and POPA later.

> mov ax, 0
> mov es, ax
> mov bx, 078h

> ...
> [snip]

I'm assuming the snipped section is correct.

> ; secread - bx = offset, cl = sector
> secread proc
>
> push bx
> push cx
> push ax
>
> mov ax, 0h
> mov es, ax ; buffer segment
> mov ah, 02h ; code
> mov al, 01h ; 1 sector
> mov ch, 00h ; track 0
> mov dh, 01h ; head
> mov dl, 00h ; drive
>

If needed, you might be able to shave some bytes here. AX, AL, AH
are zero. They can be MOVed to other registers that take zero.
INC maybe be useful too. I'd probably look into XOR AX and then
MOVW to CX and DX and use INC to adjust to 1 or 2.

> int 013h
>
> pop ax
> pop cx
> pop bx
> ret
>
> secread endp
>
>
> org 02feh
> lastword dw 0aa55h
>
> _TEXT ends
>
> end top

HTH,


Rod Pemberton

[kerravon]

On Tuesday, April 16, 2013 5:53:17 PM UTC+10, Alexei A. Frounze wrote:
> > Thanks for the offer, but I'm after public
> > domain, not FreeBSD.
>
> Why?

Because I don't consider a body of work done
properly until it has reached the stage of
Shakespeare's work. ie belonging to the public,
the entire world, with no strings attached
whatsoever.

Also see this:
http://pdos.cvs.sourceforge.net/viewvc/pdos/pdos/pdpgoal.txt?view=markup

> And I see no reason why one would want to have a single universal boot
> sector supporting all three FAT formats. When you format the disk you
> write the bootsector appropriate for the FAT format and forget about
> it.

I haven't got to the point of writing a format
program. I was planning on using FreeDOS's
version. But then I need to replace FreeDOS's boot
sector. Which I can cobble together with existing
tools and install later.

> Function 2 has a limit on the size of the disk, the limit is baked
> into the CHS encoding scheme.
>
> Function 42h is what you want to use on large drives to avoid the CHS
> limit.

Ok, thanks.

BFN. Paul.

[Herbert Kleebauer]

On 16.04.2013 14:18, kerravon wrote:

> Because I don't consider a body of work done
> properly until it has reached the stage of
> Shakespeare's work. ie belonging to the public,
> the entire world, with no strings attached
> whatsoever.
>
> Also see this:
> http://pdos.cvs.sourceforge.net/viewvc/pdos/pdos/pdpgoal.txt?view=markup

Unless something is public domain, it essentially
hasn't been invented yet, as far as not needing to reinvent it is
concerned. It's like inventing the cure for cancer and keeping it a
secret.

Seems you don't understand "free software". To release software into
the Public Domain doesn't make it free, the opposite is true, everybody
has the freedom to enslave it.

If you find a medicine for cancer and release it into the Public
Domain, then a company is free to do a little modification and
patent the result. Then they only have to find a way to modify
cancer in way that only their patented but not your original
medicine does work. Your medicine (software) is only free if released
under GPL, which means that the modified medicine also has to be
free. FreeBSD license also doesn't make software "free".

[wolfgang kern]

Herbert Kleebauer replied to kerravon:

Fully agree Herbert,
If some of us find out something which can improve our code creation
then we can share it with our friends whithin the few remaining NGs or
just use it for our very own purpose and keep quite (iaw: dont tell) :)

Patents and Licences are made for the greedy and for the not enough rich...

__
wolfgang

[kerravon]

On Wednesday, April 17, 2013 3:58:47 AM UTC+10, Herbert Kleebauer wrote:
> Seems you don't understand "free software". To release software into
> the Public Domain doesn't make it free, the opposite is true, everybody
> has the freedom to enslave it.

That's the old communist trope, equating "freedom"
with "communist slavery".

There is nothing "enslaved" about Shakespeare's work.
Or my work (PDOS etc) for that matter.

BFN. Paul.

[Alexei A. Frounze]

On Apr 16, 5:18 am, kerravon <kerra...@w3.to> wrote:
> On Tuesday, April 16, 2013 5:53:17 PM UTC+10, Alexei A. Frounze wrote:
> > > Thanks for the offer, but I'm after public
> > > domain, not FreeBSD.
>
> > Why?
>
> Because I don't consider a body of work done
> properly until it has reached the stage of
> Shakespeare's work. ie belonging to the public,
> the entire world, with no strings attached
> whatsoever.

How many strings are in a FreeBSD license? Keeping a copy of the
license in the source code and/or documentation? Other than that you
can do anything with the code.

Alex

[Jean-Marc L.]

Alexei A. Frounze wrote:
> On Apr 16, 5:18 am, kerravon <kerra...@w3.to> wrote:

...

>> Because I don't consider a body of work done
>> properly until it has reached the stage of
>> Shakespeare's work. ie belonging to the public,
>> the entire world, with no strings attached
>> whatsoever.
>
> How many strings are in a FreeBSD license? Keeping a copy of the
> license in the source code and/or documentation? Other than that you
> can do anything with the code.

Yes, but if you put the license in the documentation and then somebody
redistribute the binaries without documentation... The binaries alone
are violating the license terms.

And note that if you use source code from different authors, you must
check every source files and reproduce every different copyright notices
in the documentation.

That's why I do prefer public domain. Note that there is two Open Source
licenses whose don't have the mentioned problems.
In my opinion, The "Boost Software License" and the "Zlib License" are
acceptable.

Jean-Marc

--
http://www.cod5.org/archive

[Rod Pemberton]

Alexei A. Frounze" <alexf...@gmail.com> wrote in message
news:04e2aad5-dbd1-42dc...@di5g2000pbc.googlegroups.com...

> On Apr 16, 5:18 am, kerravon <kerra...@w3.to> wrote:
> > On Tuesday, April 16, 2013 5:53:17 PM UTC+10, Alexei A.
Frounze wrote:

> > > Thanks for the offer, but I'm after public
> > > domain, not FreeBSD.

Sigh, I've just been involved in too many of these PD (US Public
Domain) versus FOSS copyright arguments arguing the same as you
Paul ... It's sad to find out that my other Usenet friends
Alexei, Herbert, and Wolfgang are on the "wrong" side here ...
Otherwise, Paul, I'd be supporting your other statements in this
thread in regards to this. Instead, I'll just (re)state my
perspective.

I believe in both PD and copyrighted code. I don't believe in
FOSS code like GPL or LGPL. Those licenses are too restrictive to
be called "open source", and they have an political agenda:
anti-company. They're designed to end code rights of
corporations. They're used by people who wish to "destroy" for
profit software companies. But, if you program for a company,
then that company needs to have ownership rights to make money,
and pay your salary as a programmer. Do you want a job as a
programmer? If you're an "artist" with code, you'd probably like
to release some your personal code as PD, or maybe MIT or BSD. I
also like MIT or BSD licensed code, but less so. They don't have
any prohibitions against using their code with GPL or LGPL code.
That's their weakness. The MIT or BSD licenses can't stop the GPL
from tainting the MIT or BSD code base. All MIT or BSD code
becomes mixed with GPL code over time.

What's interesting is that a non-Socialist country like the US
allows code without a copyright (not allowed by law, forfeited
copyright, or copyright term has expired, etc.) which allows for
code that is effectively owned by all Americans, or anyone else
for that matter. However, in a Socialist country like Germany,
all code is copyrighted, upon creation, forever, no matter who
created it, no matter who paid for it. If US tax payers paid for
the code, then there is no copyright on the code. It's owned by
all Americans. Does the German government retain copyright to
such code? I've conversed with more than a few Germans who can't
distinquish between BSD or MIT licensed code and PD code. They
see them as being equivalent when they're similar but not the
same. One is copyrighted, but all legal rights that can be
granted to others have been granted to others. Some rights are
not transferable. The other is not copyrighted, cannot be
copyrighted [uncopyrightable product of the US government, or a
previously copyrighted work from the legal perspective - forfeited
copyright or expired copyright], and all rights to it are owned by
everyone. Americans via our taxes have paid for much code within
many industries. That code is owned by all of us - not by the
corporations that use it.

> > > Why?
> >
> > Because I don't consider a body of work done
> > properly until it has reached the stage of
> > Shakespeare's work. ie belonging to the public,
> > the entire world, with no strings attached
> > whatsoever.
>
> How many strings are in a FreeBSD license?

Not enough, IMO. There are no prohibitions against combining
minimal term MIT or BSD licensed code with more restrictive GPL or
LGPL licensed code. There is no "string" in MIT or BSD licensed
code requiring that additions, changes, or modifications to the
code also fall under the original MIT or BSD copyright. I.e.,
there is no usage exclusion for abusive open source licenses like
the GPL.

> Keeping a copy of the license in the source
> code and/or documentation? Other than that you
> can do anything with the code.

That's part of the problem. GPL'd and LGPL'd licensed code can be
combined with other all other Free Open Source Software. The
GPL/LGPL licenses are seen as being equivalent or compatible open
source licenses to other licenses like MIT or BSD licenses.
Unfortunately, that's not true. The GPL and LGPL licenses are
far, far more restrictive than BSD or MIT licenses. Merging code
from with different licenses effectively places the resulting
combination under GPL, either since the GPL portion can't be
determined in order to be removed, or it enters into legal limbo.
Under US law, only *one* copyright applies to the *entire* work.
So, merged code - within one file - with multiple licenses may
actually be illegal in the US. I don't believe there has been a
court case testing this in the US. However, linking code, under
different licenses, in different files, keeps the licenses
separate and compatible.


Rod Pemberton

[Rod Pemberton]

"Jean-Marc L." <inv...@none.none> wrote in message
news:kklgot$os4$1...@dont-email.me...

> Alexei A. Frounze wrote:
> > On Apr 16, 5:18 am, kerravon <kerra...@w3.to> wrote:
> ...
> >> Because I don't consider a body of work done
> >> properly until it has reached the stage of
> >> Shakespeare's work. ie belonging to the public,
> >> the entire world, with no strings attached
> >> whatsoever.
> >
> > How many strings are in a FreeBSD license? Keeping a copy of
> > the license in the source code and/or documentation? Other
> > than that you can do anything with the code.
>
> Yes, but if you put the license in the documentation and then
> somebody redistribute the binaries without documentation...
> The binaries alone are violating the license terms.
>

How so? The license applies to the source code since that's what
the copyright protects. The binaries aren't protected by
copyright since they're not the source code. So, unless there is
a clause in the license that says the license applies to the
resulting binaries also, I don't see how distributing the binary
without a license violates the licensing terms. That's assuming
it's even legal to have a copyright license for the source code
apply rights and terms to something that's not copyrightable such
as the binary. It probably isn't. G.P.L. created F.U.D.


Rod Pemberton

[kerravon]

On Wednesday, April 17, 2013 10:57:05 AM UTC+10, Alexei A. Frounze wrote:
> On Apr 16, 5:18 am, kerravon <kerra...@w3.to> wrote:
>
> > Because I don't consider a body of work done
> > properly until it has reached the stage of
> > Shakespeare's work. ie belonging to the public,
> > the entire world, with no strings attached
> > whatsoever.
>
> How many strings are in a FreeBSD license?

While ever you baulk at making your code public
domain, I will baulk at accepting FreeBSD. Your
FreeBSD code will eventually become true public
domain (in 100 years or whatever). I want to
see that transition sped up.

> Keeping a copy of the
> license in the source code and/or documentation? Other than that you
> can do anything with the code.

I don't want any strings whatsoever. If people wish
to acknowledge other parties, that's their choice.
I don't wish to be a dog in a manger, and I don't
wish to force companies or others to do something
they don't really want to do.

BFN. Paul.

[Jean-Marc L.]

Rod Pemberton wrote:
> "Jean-Marc L." <inv...@none.none> wrote in message
> news:kklgot$os4$1...@dont-email.me...
>> Alexei A. Frounze wrote:
>>> On Apr 16, 5:18 am, kerravon <kerra...@w3.to> wrote:
>> ...
>>>> Because I don't consider a body of work done
>>>> properly until it has reached the stage of
>>>> Shakespeare's work. ie belonging to the public,
>>>> the entire world, with no strings attached
>>>> whatsoever.
>>>
>>> How many strings are in a FreeBSD license? Keeping a copy of
>>> the license in the source code and/or documentation? Other
>>> than that you can do anything with the code.
>>
>> Yes, but if you put the license in the documentation and then
>> somebody redistribute the binaries without documentation...
>> The binaries alone are violating the license terms.
>>
>
> How so? The license applies to the source code since that's what
> the copyright protects. The binaries aren't protected by
> copyright since they're not the source code.

That's true.

> So, unless there is
> a clause in the license that says the license applies to the
> resulting binaries also, I don't see how distributing the binary
> without a license violates the licensing terms.

The FreeBSD and BSD licenses have such clause.

MIT doesn't have a "binaries" clause, it has
"substantial portions of the Software".
Is a binary "a portion" of the source code ?

> That's assuming
> it's even legal to have a copyright license for the source code
> apply rights and terms to something that's not copyrightable such
> as the binary. It probably isn't. G.P.L. created F.U.D.

I think it is legal, such license is, in my opinion, a mix
of a copyright license and an End-User License Agreement.

But I'm not a lawyer, my opinion may change in future.

Jean-Marc

--
http://www.cod5.org/archive

[Alexei A. Frounze]

On Apr 17, 12:45 am, "Rod Pemberton" <do_not_h...@notemailnotq.cpm>
wrote:
> Alexei A. Frounze" <alexfrun...@gmail.com> wrote in messagenews:04e2aad5-dbd1-42dc...@di5g2000pbc.googlegroups.com...

It's funny that Paul finds MIT/BSD too restrictive (and I still don't
get it) and you find it insufficiently restrictive. You can't please
everyone every time, I guess! :)

OTOH, what prevents PD from being tainted/abused/the-verb-you-find-
most-appropriate by the likes of GPL? Nothing, right?

My preference of BSD/MIT over the default copyright or PD or GPL is
simple:
1. I want to share the code
2. I don't want to impose severe restrictions on what can and can't be
done with it, even if someone drops it into the depths of GPL'd code
(I might revise my view of GPL in the future, but at the moment I
don't care much about it tainting versions of my code produced by
others)
3. I don't want to require others to pay to me or to obtain written
permissions from me or anything alike
4. I want my name to remain there, which I think is a fair non-
monetary price to ask (unless you want to be anal and count every
character of the license and multiply it by the storage and
transmission cost of a single byte times the number of repetitions)

Alex

[kerravon]

On Tuesday, April 16, 2013 7:02:52 PM UTC+10, Rod Pemberton wrote:
> "kerravon" <kerr...@w3.to> wrote in message
>

> > org 0100h
>
> I thought 100h was for DOS .com's ... ;-) Did you want 7C00h
> here?

I'm not sure. I don't understand what effect
that will have. It's currently working.

> > ; jump around buffer
> > jmp bypass
>
> It's my understanding that for compatibility with MS-DOS and Win98
> etc, that the "jmp bypass" needs to be a short jump, i.e., two
> byte encoding, followed by a NOP instruction, for a total of three

Thanks. That was a great help and now my boot
sector is working off hard disk. Not only that,
but I can use FreeDOS now instead of MSDOS. :-)

> > bpb db 51 dup(?)
>
> Ok. You seem to be allocating the correct amount.
>
> Why do you allocate a BPB without setting any values in the BPB?

How else can I do it? The BPB needs to be populated
as part of doing a format. It is variable.

> > ;jmp 0070h:0000h
>
> > mov ax, 0070h
>
> > push ax
>
> > mov ax, 0
>
> > push ax
>
> > retf
>
> Why are you using RETF instead of the absolute JMP that's
> commented out? Is this a MASM limitation? Isn't there a DB form
> for MASM to do that? I might've used RETF earlier to set CS:IP...

wasmr gives this error when I try that:

C:\DEVEL\PDOS\SRC>wasmr -zq -zcm -Dmemodel=tiny pbootsec.asm
pbootsec.asm(58): Error! E085: Only register or label is expected in override

> With that many pushes, I'd look into whether PUSHA was usable here
> and POPA later.

I'm not familiar with those directives.

BFN. Paul.

[Tony]

In article <d9fb750c-0f14-4bac...@googlegroups.com>,
kerr...@w3.to says...

"Gimme, gimme!"? Are you one of those characters that continually seek to get
something for nothing? Feeling entitled?

[Tony]

In article <kkk3hc$c15$1...@speranza.aioe.org>, kl...@unibwm.de says...

>
> On 16.04.2013 14:18, kerravon wrote:
>
>> Because I don't consider a body of work done
>> properly until it has reached the stage of
>> Shakespeare's work. ie belonging to the public,
>> the entire world, with no strings attached
>> whatsoever.
>>
>> Also see this:
>> http://pdos.cvs.sourceforge.net/viewvc/pdos/pdos/pdpgoal.txt?view=markup
>
> Unless something is public domain, it essentially
> hasn't been invented yet, as far as not needing to reinvent it is
> concerned. It's like inventing the cure for cancer and keeping it a
> secret.
>
> Seems you don't understand "free software".

Uh oh, here we go... "they're baa-aack".

> To release software into
> the Public Domain doesn't make it free, the opposite is true, everybody
> has the freedom to enslave it.

Software is inanimate and therefore cannot be enslaved. Save the propaganda for
alt.gpl and alt.jehova.

>
> If you find a medicine for cancer and release it into the Public
> Domain, then a company is free to do a little modification and
> patent the result.

But the public domain recipe is still public domain, yes?

> Then they only have to find a way to modify
> cancer in way that only their patented but not your original
> medicine does work.

How ironic. GPL software seems to be written under that strategy: the spaghetti
code is "free". Uh huh, as free as the "gift" of an elephant is. GPL'd
executables are largely free though, but don't have to be, and there is
commercial software that is free also. Source code is a very different thing
than software (though I have a feeling that "the law" has not yet achieved more
than an infant's knowledge about that yet, perhaps on purpose).

> Your medicine (software) is only free if released
> under GPL,

Plenty of free "medicine" is available. Smoking cleans out the lungs and
prevents cancer, ya know.

> which means that the modified medicine also has to be
> free.

There is only one definition of 'free' that can apply to software, because
software is inanimate, and that definition relates to price.

> FreeBSD license

Did you mean BSD license?

> also doesn't make software "free".

You seem to be projecting fantasy onto "software", akin to Easter Bunny and
Santa Claus fantasy.

[Tony]

In article <kkljpv$vhi$1...@speranza.aioe.org>, do_no...@notemailnotq.cpm
says...

>
> Alexei A. Frounze" <alexf...@gmail.com> wrote in message
> news:04e2aad5-dbd1-42dc...@di5g2000pbc.googlegroups.com...
> > On Apr 16, 5:18 am, kerravon <kerra...@w3.to> wrote:
> > > On Tuesday, April 16, 2013 5:53:17 PM UTC+10, Alexei A.
> Frounze wrote:
>
> > > > Thanks for the offer, but I'm after public
> > > > domain, not FreeBSD.
>
> Sigh, I've just been involved in too many of these PD (US Public
> Domain) versus FOSS copyright arguments arguing the same as you
> Paul ... It's sad to find out that my other Usenet friends
> Alexei, Herbert, and Wolfgang are on the "wrong" side here ...

Please explain ""wrong" side". My take on it is (yes, my .02 is FREE!) that the
GPL crowd want everything for free. Like that current TV commercial where the
Peanuts cartoon characters want everything to be 5 cents. You can talk with them
until you are blue in the face but in the end all you will get from them is that
"we think it should be 5 cents" just the Peanuts characters "demand".

> Otherwise, Paul, I'd be supporting your other statements in this
> thread in regards to this. Instead, I'll just (re)state my
> perspective.
>
> I believe in both PD and copyrighted code. I don't believe in
> FOSS code like GPL or LGPL.

Did you think about that before you wrote it? The words you chose make it seem
like you are stating a religious belief. A license (and I'm not saying that the
ones you mentioned are credible and/or legal ones, mind you) is not "something
to believe in", IMO. "Believing in" stupid crap gets you things like Al Querida.

> Those licenses are too restrictive to
> be called "open source",

They like to create lingo that is weaselly, just like any snake oil salesman's
spiel. If source code is available, why not just say, "source code is
available"? If I was creating a table listing of software, and if I was
targeting the product to developers also (note, ONLY if I was targeting
developers), I would include a column "Source Code" and the entries would
indicate 'Available' or 'Not available'. (Then again, ANYTHING is available for
the right price! You get what you pay for! And sometimes even less!).

> and they have an political agenda:
> anti-company.

Some hippies of the 1960's "never grew up" either. Not that there's anything
wrong with that, though. Peace and love and LSD, yeah!

> They're designed to end code rights of
> corporations.

Not all companies are corporations, so there is more context implied in that
sentence than your previous sentence.

> They're used by people who wish to "destroy" for
> profit software companies.

But what about peace and love and LSD? That is all good, yes? C'mon, drop a
little acid and lose all your hate man. Ya gotta drop those negative vibes.

> But, if you program for a company,

Ooo, ooo.. let ME finish that sentence!! If you program for a company, you're
one of the following:

- stupid
- trying to figure out how to develop software
- naive
- rich (but much less rich than if you knew how to capitalize on your wares)
- content, happy go lucky
- employed (enslaved?)
- an entrepreneur (!)
- a sheeple
- <additions here>

> then that company needs to have ownership rights to make money,

No. Now you're skating on the ice of labor vs. management. The unions would LOVE
to own the work product, but management won't have any of that. It's an "us
against them" behavior, and both sides have their own characteristics.

What is "interesting" is that supposedly software development is more of an
intellectual activity than, say, operating a drill press all day long, yet
programmers give away their intellectual property at the drop of a hat. The
obvious question, then, is, "Are drill press operators smarter than
programmers?".

> and pay your salary as a programmer.

There are a range of possibilities. Companies are getting away with bloody
murder in most current employment-of-programmer scenarios, IMO. Then again,
because most programmers are just "in training"/beginners, maybe I have over-
stated the reality.

> Do you want a job as a
> programmer?

Ewwwww! I'd rather be put in a cage in a biting-rat-infested river with only my
nose out of the water to breath. Oh wait, it's the same thing!

> If you're an "artist" with code, you'd probably like
> to release some your personal code as PD, or maybe MIT or BSD.

And be the proverbial "starving artist", huh. Why do you feel that programmer's
who are "artists" should get less respect than teenage country music singers
singing about their past boyfriends?

> I
> also like MIT or BSD licensed code, but less so. They don't have
> any prohibitions against using their code with GPL or LGPL code.
> That's their weakness. The MIT or BSD licenses can't stop the GPL
> from tainting the MIT or BSD code base. All MIT or BSD code
> becomes mixed with GPL code over time.

I have a feeling that those who use either, use the other. Why a software
company would employ those who have been on open source projects perplexes me
(i.e., your "taint" reason given above). I guess they feel that they can
overcome any obstacles. The small software company may not be able to though.

>
> What's interesting is that a non-Socialist country like the US

Ah ha! I KNEW communism was going to be brought up in this thread!

> allows code without a copyright (not allowed by law, forfeited
> copyright, or copyright term has expired, etc.) which allows for
> code that is effectively owned by all Americans, or anyone else
> for that matter.

"public domain"?

> However, in a Socialist country like Germany,
> all code is copyrighted, upon creation, forever, no matter who
> created it, no matter who paid for it.

Hmm. I thought that applied to source code in the U.S. also. Does one have to
formally "release" the copyright to make it public domain?

> If US tax payers paid for
> the code, then there is no copyright on the code. It's owned by
> all Americans.

Where can I get the source code for that new inter-continental ballistic missile
defense system? I mean, since *I* own it, and all.

> Does the German government retain copyright to
> such code? I've conversed with more than a few Germans who can't
> distinquish between BSD or MIT licensed code and PD code. They
> see them as being equivalent when they're similar but not the
> same. One is copyrighted, but all legal rights that can be
> granted to others have been granted to others. Some rights are
> not transferable. The other is not copyrighted, cannot be
> copyrighted [uncopyrightable product of the US government, or a
> previously copyrighted work from the legal perspective - forfeited
> copyright or expired copyright], and all rights to it are owned by
> everyone. Americans via our taxes have paid for much code within
> many industries. That code is owned by all of us - not by the
> corporations that use it.

So, where is the source code repository? Is it all useless Cobol code?
("Useless" in that unless you are building yet another war machine (YAWM)...).

>
> > > > Why?
> > >
> > > Because I don't consider a body of work done
> > > properly until it has reached the stage of
> > > Shakespeare's work. ie belonging to the public,
> > > the entire world, with no strings attached
> > > whatsoever.
> >
> > How many strings are in a FreeBSD license?
>
> Not enough, IMO. There are no prohibitions against combining
> minimal term MIT or BSD licensed code with more restrictive GPL
> or
> LGPL licensed code.

He, he.. wanna see a GPLer got all fired up? Call his GPL license "more
restrictive"!

> There is no "string" in MIT or BSD licensed
> code requiring that additions, changes, or modifications to the
> code also fall under the original MIT or BSD copyright. I.e.,
> there is no usage exclusion for abusive open source licenses like
> the GPL.

"abusive"? C'mon, say how you really feel.

>
> > Keeping a copy of the license in the source
> > code and/or documentation? Other than that you
> > can do anything with the code.
>
> That's part of the problem. GPL'd and LGPL'd licensed code can be
> combined with other all other Free Open Source Software. The
> GPL/LGPL licenses are seen as being equivalent or compatible open
> source licenses to other licenses like MIT or BSD licenses.
> Unfortunately, that's not true. The GPL and LGPL licenses are
> far, far more restrictive than BSD or MIT licenses.

"Oh man, gotta lose those negative vibes. Turn-on and tune-in man and feel
freeeeeeee (eh-em, ""free" as in"... LSD, yeah, that's the ticket). Have another
hit of acid man, and freely mellow in the freedom of free, man."

> Merging code
> from with different licenses effectively places the resulting
> combination under GPL,

To me, it seems that "GPL license" has not yet been shown to not be an oxymoron.
That is, that it is something legally credible. I mean, anyone(s) who try to
take a word like "free" and pervert it for propaganda reason, can't really be
taken seriously at anything, IMO. It's pretty much a moot point though, because
the code is all junk anyway. THAT may be the real agenda (ref: "smoke and
mirrors"): attempt to make it seem like there is actually something to protect.
Unions trying to create unnecessary jobs? Hey, if it doesn't need to be done,
people should be allowed to stay home and drop acid all day long if that suits
them. I think maybe forcing people to have "jobs" is what creates all the
fantasical nothings. As you can probably tell by now, I hate bullshit.

> either since the GPL portion can't be
> determined in order to be removed, or it enters into legal limbo.

I think "legal limbo" is quite fine for GPLers (i.e., they feel that a useful
tactic in support of their cause (creation of life from the inanimate,
apparently. No, just getting you to "believe in it"!)).

[Tony]

In article <dc67c829-6e8f-49f3...@id10g2000pbc.googlegroups.com>,
alexf...@gmail.com says...

>
> On Apr 17, 12:45 am, "Rod Pemberton" <do_not_h...@notemailnotq.cpm>
> wrote:
> > Alexei A. Frounze" <alexfrun...@gmail.com> wrote in messagenews:04e2aad5-dbd1-42dc...@di5g2000pbc.googlegroups.com...

> My preference of BSD/MIT over the default copyright or PD or GPL is
> simple:
> 1. I want to share the code

What does that mean?

> 2. I don't want to impose severe restrictions on what can and can't be
> done with it,

Not even bomb-making? War machines? Awhile back, I did a tiny bit of grant
research (i.e., I was looking for funding). I found something in the grant
stipulation that I could not accept: that they could use the source code in any
way they wanted to. War machines. I'll be god-damned, if I'm going to help some
fucking government build their fucking war machines. They can go to hell with
their fucking government and their fucking war machines and their fucking money.

> even if someone drops it into the depths of GPL'd code
> (I might revise my view of GPL in the future, but at the moment I
> don't care much about it tainting versions of my code produced by
> others)

> 3. I don't want to require others to pay to me

Why not? You aspire to be a mother Teresa?

> or to obtain written
> permissions from me or anything alike

> 4. I want my name to remain there,

So it is, then, fame that you seek? That does seem to be THE youthful aspiration
these days (i.e., being famous rather than being useful).

> which I think is a fair non-
> monetary price to ask (unless you want to be anal and count every
> character of the license and multiply it by the storage and
> transmission cost of a single byte times the number of repetitions)
>
> Alex

Here's the thing. Where does it end (or begin even)? Does the reward of having
your name in lights in a help dialog box a bit much for just any little thing?
Where is the point where the parallels to graffiti perpetrators fits? Most code
found on the net which is free and requires such seems like an idea for further
research and development by the creator, and not useful for general use. I.e.,
it is nowhere near the level of quality which could demand payment, hence why
should it get recognition at all? The only way really to recognize it would be
to show EXACTLY how and how much of the code was used, so as not to fall prey to
graffiti tagger tactics. Also, I think most commercial developers would find it
easier to just buy something and use it rather than having to introduce a
process into their company that has to deal with a thousand stipulations of a
thousand developers. (Just random thoughts/food for thought in this paragraph.
I'm not saying you don't deserve credit or that your code is shit--I haven't
seen it).

[Tony]

In article <kkk59m$oj$1...@newsreader2.utanet.at>, now...@never.at says...

> Fully agree Herbert,
> If some of us find out something which can improve our code creation
> then we can share it with our friends whithin the few remaining NGs or
> just use it for our very own purpose and keep quite (iaw: dont tell) :)
>
> Patents and Licences are made for the greedy and for the not enough rich...

It's protection, IF (and that's a big "if") you can afford it (remember, the
real cost is the cost of defending your claims in court, and not the filing
fees, which are also too expensive for the small software developer, btw). Once
the cat is out of the bag (embodiment revealed), it all seems easy, but who's
going to pay for all the R&D? The GPL mantra works for those with nothing to
lose (read, they are too early in their abilities, and perhaps will never
achieve any significant level of ability, to create anything significant, but
feel that they are entitled to all source code that IS significant, by thinking
that they have some kind of right to the results of the labors of others,
WITHOUT COMPENSATING THOSE WHO HAVE LABORED, CREATED, INVENTED, SACRIFICED).

I don't like people who want something for nothing (apparently, that's a big
chunk of the world! How depressing.). Why don't they go buy a fucking lotto
ticket instead? Oh, that won't work: it would cost them a buck!

[Rod Pemberton]

"Tony" <a...@some.org> wrote in message
news:MPG.2bda72778...@nntp.aioe.org...

> In article
<d9fb750c-0f14-4bac...@googlegroups.com>,
> kerr...@w3.to says...

> > I don't want any strings whatsoever. If people wish
> > to acknowledge other parties, that's their choice.
> > I don't wish to be a dog in a manger, and I don't
> > wish to force companies or others to do something
> > they don't really want to do.
> >
> > BFN. Paul.
>
> "Gimme, gimme!"? Are you one of those characters
> that continually seek to get something for nothing?
> Feeling entitled?

You've made the wrong assumption about him. He's the giver.
You'd be the givee. If you used his software, you'd be the one
characterized as saying: "Gimme, gimme!".

http://pdos.sourceforge.net/


Rod Pemberton

[Rod Pemberton]

"kerravon" <kerr...@w3.to> wrote in message

news:4a5e12b3-af25-4e7b...@googlegroups.com...

> On Tuesday, April 16, 2013 7:02:52 PM UTC+10, Rod Pemberton
wrote:
> > "kerravon" <kerr...@w3.to> wrote in message

...

> > > bpb db 51 dup(?)
> >
> > Ok. You seem to be allocating the correct amount.
> >
> > Why do you allocate a BPB without setting any values in the
BPB?
>
> How else can I do it? The BPB needs to be populated
> as part of doing a format. It is variable.
>

I'm not sure what you need for yours, but I statically populate
mine. Of course, it's only being used for a floppy bootsector and
I don't have any utils that would fill it in like a format command
...


Rod Pemberton

[Rod Pemberton]

"Tony" <a...@some.org> wrote in message

news:MPG.2bda93ad7...@nntp.aioe.org...
> In article
dc67c829-6e8f-49f3...@id10g2000pbc.googlegroups.com
> alexf...@gmail.com says...

> > 2. I don't want to impose severe restrictions on what
> > can and can't be done with it,
>
> Not even bomb-making? War machines?

I'm sure he means financial.

Most "free" software prior to prior to widespread adoption of the
GPL had anti-government clauses. I'd say that was probably true,
at least for the US, until the early 2000's. So, you can easily
demonize the widespread adoption of the GPL for the widespread
removal the anti-government clause from FOSS licenses. I'm sure
that if you asked supporters of the GPL if they supported use of
their code by the military you'd get a solid: "NO!". OMG, do you
think GPL supporters even realize this? I doubt it. The serious
question is why would the FSF support such a measure in the first
place...? The FSF had to be aware of the standard license terms
of most "free" software of the era, which excluded government use.
Was the FSF hoping that the US government and US military would
use FSF code? Just what end would that serve ... ? Could FSF
code be forced upon US citizens via law? The US government
taxpayers paying for programmers to code FSF code? The US school
system using FSF code? Whatever it is, it's most likely a
malevolent and devious purpose. What type of conspiracy do you
think it is? Do you think the FSF is just a front for the US
government?


Rod Pemberton

[Rod Pemberton]

"Tony" <a...@some.org> wrote in message

news:MPG.2bda984f4...@nntp.aioe.org...

> In article <kkk59m$oj$1...@newsreader2.utanet.at>, now...@never.at

...

> > Fully agree Herbert,
> > If some of us find out something which can improve our code
> > creation then we can share it with our friends whithin the few
> > remaining NGs or just use it for our very own purpose and
> > keep quite (iaw: dont tell) :)
> >
> > Patents and Licences are made for the greedy and for the not
> > enough rich...
>
> It's protection, IF (and that's a big "if") you can afford it
> (remember, the real cost is the cost of defending your
> claims in court, and not the filing fees, which are also
> too expensive for the small software developer, btw).

In the US, you have to prove financial losses upfront. If you
can't, you don't have a court case. That's true even if a patent
or copyright protects your property. This "didn't suffer
financial losses" argument is the basis for "fair use" and other
perfectly legal infringements.

> The GPL mantra works for those with nothing to lose [...]

It's also quite hard to prove financial damages when no one pays
for your software since the financial market for your software is
now immeasurable because no one pays for commercial software to do
that since they're all using GPL for free as in no cost.

> I don't like people who want something for nothing (apparently,
> that's a big chunk of the world! How depressing.).

It's not always a case of people who want something for nothing.
Is it? That implies that they can afford it, but won't, i.e., an
ideologue or a theif. Although many of those people are around, I
doubt those people are the vast majority. I think for most cases,
it's truly people who cannot afford it, e.g., young adults like
college students, inner city poor, or the product priced
appropriately for a rich market but too high for another, less
rich market, etc. I also think that for some cases, it's people
who can't obtain the software via normal retail channels for
various reasons, e.g., credit problems, product no longer sold or
abandonware, product not available in your country, etc.


Rod Pemberton

[Tony]

In article <kkr9n1$mlc$1...@speranza.aioe.org>, do_no...@notemailnotq.cpm
says...

When it comes time that I need code, I want to pay for it. Or trade. I don't
want anything for free. I feel that the "get something for nothing" attitude is
a major problem (it's necessarily taking advantage of the misfortune of others--
consider estate sales just before someone becomes homeless: got quite a bargain
on that big screen TV, huh.) and I want those who are actually creating the
goods to get the compensation--not some MBA pansies who think they are entitled
to ownership of the work product just because they can get away with that
because it's legal. I require a higher level of ethic than that bullshit. And I
want to pay full price. Not half, not double. Now if something is over-priced,
that is a separate issue.

[Tony]

In article <kkranv$pcb$1...@speranza.aioe.org>, do_no...@notemailnotq.cpm
says...

>
> "Tony" <a...@some.org> wrote in message
> news:MPG.2bda93ad7...@nntp.aioe.org...
> > In article
> dc67c829-6e8f-49f3...@id10g2000pbc.googlegroups.com
> > alexf...@gmail.com says...
>
> > > 2. I don't want to impose severe restrictions on what
> > > can and can't be done with it,
> >
> > Not even bomb-making? War machines?
>
> I'm sure he means financial.

I'm sure I meant there are other issues which may be worthy of consideration.

>
> Most "free" software prior to prior to widespread adoption of the
> GPL had anti-government clauses. I'd say that was probably true,
> at least for the US, until the early 2000's.

I do think I may recall reading something like that in source code offerings.

> So, you can easily
> demonize the widespread adoption of the GPL for the widespread
> removal the anti-government clause from FOSS licenses.

So GPL has helped the war industry. Another reason not to like it.

> I'm sure
> that if you asked supporters of the GPL if they supported use of
> their code by the military you'd get a solid: "NO!".

Then why not put in a clause saying that?

> OMG, do you
> think GPL supporters even realize this?

Of course I don't. Like I've said, I think they are kids (intellectually) who
want everything for 5 cents like the Peanuts gang. It takes time to mature and
accomplish but tell that to the teen dropping out of school and joining a gang
instead.

> I doubt it. The serious
> question is why would the FSF support such a measure in the first
> place...? The FSF had to be aware of the standard license terms
> of most "free" software of the era, which excluded government use.
> Was the FSF hoping that the US government and US military would
> use FSF code? Just what end would that serve ... ? Could FSF
> code be forced upon US citizens via law? The US government
> taxpayers paying for programmers to code FSF code? The US school
> system using FSF code? Whatever it is, it's most likely a
> malevolent and devious purpose. What type of conspiracy do you
> think it is? Do you think the FSF is just a front for the US
> government?
>

Very funny (except for the part that is true ;) ).

[Rod Pemberton]

"Tony" <a...@some.org> wrote in message

news:MPG.2bda8a48e...@nntp.aioe.org...
> In article <kkljpv$vhi$1...@speranza.aioe.org>, Rod Pemberton
do_no...@notemailnotq.cpm

> > > > On Apr 16, 5:18 am, kerravon <kerra...@w3.to> wrote:

...

> > > > > Thanks for the offer, but I'm after public
> > > > > domain, not FreeBSD.
> >
> > Sigh, I've just been involved in too many of these PD (US
> > Public Domain) versus FOSS copyright arguments arguing
> > the same as you Paul ... It's sad to find out that my other
> > Usenet friends Alexei, Herbert, and Wolfgang are on the
> > "wrong" side here ...
>
> Please explain ""wrong" side".

They're in favor of GPL (and/or BSD/MIT licenses). I'm in favor
of PD or normal copyrights. I.e., those two that I favor are the
far ends of the spectrum of possible licenses if ranked from least
restrictive to most. The reasons were in the paragraph
immediately following which started: "I believe in both PD and
copyrighted code. ...".

> > Otherwise, Paul, I'd be supporting your other statements in
> > this thread in regards to this. Instead, I'll just (re)state
> > my perspective.
> >
> > I believe in both PD and copyrighted code. I don't believe in
> > FOSS code like GPL or LGPL.
>
> Did you think about that before you wrote it? The words you
> chose make it seem like you are stating a religious belief.
> A license (and I'm not saying that the ones you mentioned
> are credible and/or legal ones, mind you) is not "something
> to believe in", IMO. "Believing in" stupid crap gets you things
> like Al Querida.

It summarizes my position and was easier to say than rewriting the
earlier paragraph that explained it. Here, I've restated it just
for you:

I accept both PD and typical copyrighted code - the latter being
"all rights reserved" for the holder of a copyright such as a
business - as valid since they both have merit and worth. I don't
accept FOSS code like GPL or LGPL since they lack merit and worth.
They (GPL, LGPL) have no purpose which isn't already served by the
other two (PD, normal copyright).

Was that better? 8-| @@ 9_9

> > They're designed to end code rights of
> > corporations.
>
> Not all companies are corporations, so there is more
> context implied in that sentence than your previous sentence.

The difference between a company and corporation is far more
pronounced in England than it is in the US, today.

FYI, most entities that could qualify as companies in the US, are
almost always setup as one of the forms incorporation due to
potential legal liabilites, e.g., Chapter C Corporation,
Subchapter S Corporation, 501(c) "not-for-profit Corporation, or
509(a) charity or foundation which are special 501(c)
Corporations.

This is an introduction to the basic types of valid legal business
entities in the US:
http://www.sba.gov/category/navigation-structure/starting-managing-business/starting-business/choose-your-business-stru

Of course, that includes now "undesirable" legal entities, such as
a sole proprietorship or "Doing-Business-As", Partnerships
(General, Limited, Limited Liability), Limited Liability
Companies, and Cooperatives. While it lists Cooperatives, it
doesn't list Mutuals or a few other special setups valid in the
US. Over the last decade, due to advantageous changes in US law,
Subchapter S Corporations have basically replaced sole
proprietorships, partnerships, and LLCs in the US.

> Hmm. I thought that applied to source code in the U.S. also.
> Does one have to formally "release" the copyright to make
> it public domain?

No.

Well, you weren't paying attention:

1) A work of the US government (requires US government employee to
do the work - I should've clarified that before) cannot be
copyrighted since it was created with taxpayer money.
2) US copyrights terminate after a certain time period. This
material becomes public domain.
3) People willfully forfit any automatic copyright under State,
Federal, or international treaties. The legality of this has not


been fully tested in the US in court, but is generally accepted by
citizens.

> > If US tax payers paid for
> > the code, then there is no copyright on the code. It's owned
> > by all Americans.
>
> Where can I get the source code for that new inter-continental
> ballistic missile defense system? I mean, since *I* own it, and
> all.

That's either:
1) commercial copyrighted product by whomever received the
contract, i.e., no US government employees involved
or:
2) classified and restricted since it was coded by a US government
employees

I.e., if such code is owned by US citizens, the government doesn't
have to comply with the laws simply because they're the government
and when confronted they can claim national security reasons.

> > > > > Why?
> > > >
> > > > Because I don't consider a body of work done
> > > > properly until it has reached the stage of
> > > > Shakespeare's work. ie belonging to the public,
> > > > the entire world, with no strings attached
> > > > whatsoever.
> > >
> > > How many strings are in a FreeBSD license?
> >
> > Not enough, IMO. There are no prohibitions against combining
> > minimal term MIT or BSD licensed code with more restrictive
> > GPL or LGPL licensed code.
>
> He, he.. wanna see a GPLer got all fired up? Call his GPL
> license "more restrictive"!

I just did, and did so truthfully too...

Are you going to submit my name to their attack list?

You know the FSF has a secret counter-intelligence, infiltration,
tax avoidance, and suppression squad to suppress suppressive
persons other than members of their own suppression squad, just
like that Californian sci-fi nanu-nanu "religion" that was almost
outlawed in various EU countries, yes? (Just ... joking.)


Rod Pemberton

[Tony]

In article <kkras7$ptp$1...@speranza.aioe.org>, do_no...@notemailnotq.cpm
says...

The "something for nothing" I was referring to is the result of the R&D. If one
developer (and that can mean a company) does years of R&D which results in a
product or technique that is novel and then "the entitled set" just goes and
reverse-engineers it and gives it away for free, how does the one with all the
time and money and ability get it's money back, make a profit and go on and do
bigger and better things? The same issue occurs in "first to file" patents. It
makes inventing and innovating a risky endeavor. It's not just the end users of
software, but those who pander to them. Maybe these people are getting paid to
program and think it is fair to reverse-engineer and get the goods without any
effort just so THEY can have a job at someone else's expense because they
couldn't invent anything if their life depended on it. Why don't they just stay
at home and smoke dope all day instead? I think it's an issue of ethics and
character and greed and such. I like nice people and maybe I'm "idealistic", but
I don't like bullies and other not-nice people, but some people don't care to
make anything better just as long as they are themselves content. If that's the
way it is, then people should just shut up already when something bad happens
because they are the ones causing it. Someone may DO evil, but those who cause
that ARE evil. Energy cannot be created or destroyed, only transformed. Steal a
little bit from many and think that is OK if you want to, but it isn't. It's not
any better than stealing a lot from one person. All the bullshit about the
individual being the bad one anytime something bad happens and the group trying
to distance themselves from it is bullshit. Accountability. Time to get some,
people. Reap what you sow. Going to church on Sunday does not a good person
make. Everyday there are many choices made by people, and many of those choices
are nothing but selfishness. Capitalizing on the work of others or on the
misfortune of others is evil, and if you do that, then you are evil. Not that I
think that will bother you, but that I hope the nukes you begged for will fall
only on you and not on those you exploited for your own gain--you deserve it and
it would be good riddance (well one can dream nice things can't I?). :P

[Tony]

In article <kkrcjm$uu2$1...@speranza.aioe.org>, do_no...@notemailnotq.cpm
says...

>
> "Tony" <a...@some.org> wrote in message
> news:MPG.2bda8a48e...@nntp.aioe.org...
> > In article <kkljpv$vhi$1...@speranza.aioe.org>, Rod Pemberton
> do_no...@notemailnotq.cpm
> > > > > On Apr 16, 5:18 am, kerravon <kerra...@w3.to> wrote:
> ...
>
> > > > > > Thanks for the offer, but I'm after public
> > > > > > domain, not FreeBSD.
> > >
> > > Sigh, I've just been involved in too many of these PD (US
> > > Public Domain) versus FOSS copyright arguments arguing
> > > the same as you Paul ... It's sad to find out that my other
> > > Usenet friends Alexei, Herbert, and Wolfgang are on the
> > > "wrong" side here ...
> >
> > Please explain ""wrong" side".
>
> They're in favor of GPL (and/or BSD/MIT licenses). I'm in favor
> of PD or normal copyrights. I.e., those two that I favor are the
> far ends of the spectrum of possible licenses if ranked from least
> restrictive to most. The reasons were in the paragraph
> immediately following which started: "I believe in both PD and
> copyrighted code. ...".

And why is it such a big issue? Because people won't do the right thing--pay
people for their wares, efforts, sacrifices. It's a get-something-for-nothing
culture and it sucks. This behavior is seen everyday in all over. Then again,
I'm not "well traveled" or anything so I can just relate to what my environments
have been and what I see on TV and read in the papers and the like.

>
> > > Otherwise, Paul, I'd be supporting your other statements in
> > > this thread in regards to this. Instead, I'll just (re)state
> > > my perspective.
> > >
> > > I believe in both PD and copyrighted code. I don't believe in
> > > FOSS code like GPL or LGPL.
> >
> > Did you think about that before you wrote it? The words you
> > chose make it seem like you are stating a religious belief.
> > A license (and I'm not saying that the ones you mentioned
> > are credible and/or legal ones, mind you) is not "something
> > to believe in", IMO. "Believing in" stupid crap gets you things
> > like Al Querida.
>
> It summarizes my position and was easier to say than rewriting the
> earlier paragraph that explained it. Here, I've restated it just
> for you:

I wasn't saying that I don't understand what you wrote, I was saying that HOW
you wrote it implies more than the just the words themselves mean.

>
> I accept both PD and typical copyrighted code - the latter being
> "all rights reserved" for the holder of a copyright such as a
> business - as valid since they both have merit and worth. I don't
> accept FOSS code like GPL or LGPL since they lack merit and worth.
> They (GPL, LGPL) have no purpose which isn't already served by the
> other two (PD, normal copyright).
>
> Was that better? 8-| @@ 9_9
>

There's nothing wrong with "being a believer" (though it does make one
susceptible to exploitation, of course), I was saying that a legal document is
not "something to believe in". It's such a petty thing, after all.

> > > They're designed to end code rights of
> > > corporations.
> >
> > Not all companies are corporations, so there is more
> > context implied in that sentence than your previous sentence.
>
> The difference between a company and corporation is far more
> pronounced in England than it is in the US, today.
>
> FYI, most entities that could qualify as companies in the US, are
> almost always setup as one of the forms incorporation due to
> potential legal liabilites, e.g., Chapter C Corporation,
> Subchapter S Corporation, 501(c) "not-for-profit Corporation, or
> 509(a) charity or foundation which are special 501(c)
> Corporations.
>
> This is an introduction to the basic types of valid legal business
> entities in the US:
> http://www.sba.gov/category/navigation-structure/starting-managing-business/starting-business/choose-your-business-stru
>
> Of course, that includes now "undesirable" legal entities, such as
> a sole proprietorship or "Doing-Business-As", Partnerships
> (General, Limited, Limited Liability), Limited Liability
> Companies, and Cooperatives. While it lists Cooperatives, it
> doesn't list Mutuals or a few other special setups valid in the
> US. Over the last decade, due to advantageous changes in US law,
> Subchapter S Corporations have basically replaced sole
> proprietorships, partnerships, and LLCs in the US.

Too much response for me just noting that you used 'company' in one sentence and
then 'corporation' as seemingly synonymous in another which followed. You
probably picked-up on some of what I was thinking about corporations, but not to
get overly political or anything, I'm not going to get into that right now here.
(Can't address all the ills of the world in one USENET thread!).

>
> > Hmm. I thought that applied to source code in the U.S. also.
> > Does one have to formally "release" the copyright to make
> > it public domain?
>
> No.
>
> Well, you weren't paying attention:

See my email address? A.D.D. ;)

>
> 1) A work of the US government (requires US government employee to
> do the work - I should've clarified that before) cannot be
> copyrighted since it was created with taxpayer money.
> 2) US copyrights terminate after a certain time period. This
> material becomes public domain.
> 3) People willfully forfit any automatic copyright under State,
> Federal, or international treaties. The legality of this has not
>
>
> been fully tested in the US in court, but is generally accepted by
> citizens.
>
> > > If US tax payers paid for
> > > the code, then there is no copyright on the code. It's owned
> > > by all Americans.
> >
> > Where can I get the source code for that new inter-continental
> > ballistic missile defense system? I mean, since *I* own it, and
> > all.
>
> That's either:
> 1) commercial copyrighted product by whomever received the
> contract, i.e., no US government employees involved
> or:
> 2) classified and restricted since it was coded by a US government
> employees
>
> I.e., if such code is owned by US citizens, the government doesn't
> have to comply with the laws simply because they're the government
> and when confronted they can claim national security reasons.

I was being facetious. Are you always so literal?

>
> > > > > > Why?
> > > > >
> > > > > Because I don't consider a body of work done
> > > > > properly until it has reached the stage of
> > > > > Shakespeare's work. ie belonging to the public,
> > > > > the entire world, with no strings attached
> > > > > whatsoever.
> > > >
> > > > How many strings are in a FreeBSD license?
> > >
> > > Not enough, IMO. There are no prohibitions against combining
> > > minimal term MIT or BSD licensed code with more restrictive
> > > GPL or LGPL licensed code.
> >
> > He, he.. wanna see a GPLer got all fired up? Call his GPL
> > license "more restrictive"!
>
> I just did, and did so truthfully too...
>
> Are you going to submit my name to their attack list?

Did you consider that my comment there was perhaps rhetorical? It was. Seems
rather obvious to me. Are you cognitively challenged that way? Just asking.
There's nothing wrong with that. And not that I intend to be less rhetorical
because of it.

>
> You know the FSF has a secret counter-intelligence, infiltration,
> tax avoidance, and suppression squad to suppress suppressive
> persons other than members of their own suppression squad, just
> like that Californian sci-fi nanu-nanu "religion" that was almost
> outlawed in various EU countries, yes? (Just ... joking.)
>

Too late to try to make more joke of that (and I'm not good at jokes anyway.. go
figure?).

[Herbert Kleebauer]

On 19.04.2013 05:17, Tony wrote:


> There is only one definition of 'free' that can apply to software, because
> software is inanimate, and that definition relates to price.

The "free" in free software has nothing to do with money.
You can sell GPL software at any price you like as long as
the software remains "free", which means, if you distribute a
binary, you also have to provide the source code so anybody
can inspect, modify and redistribute the software.

Suppose a company develops a product for which it needs
software with about 100k lines of code. They can either
develop this code themselves, then the software is
their property and no competitor can profit from their
work. Or they use already existing GPL software and only
have to pay for additional maybe 10k lines of code. They
save much money but as a compensation, they also have to
make this 10k lines of code available for all other
competitors. But if somebody uses this code, then maybe
he also will add some features which then you can use
without much programming costs.

Each company is free to decide which way to go, but I
really don't understand why there are people who consider
GPL as evil. And once GPL software has reached a critical
mass, it can't be stopped anymore.

[Ivan Shmakov]

>>>>> Tony <a...@some.org> writes:
>>>>> alexf...@gmail.com says...

[Cross-posting to news:misc.legal.computing, just in case.]

[...]

>> 3. I don't want to require others to pay to me

> Why not? You aspire to be a mother Teresa?

Somehow, I'd like to ask you the same?

You've posted quite a bit of creative work to this thread
lately, which, no doubt, you now hold your copyright all over.
Somewhat surprising, however, is that you /never/ ever asked for
payment for the reproduction of your creative work -- which was
copied over to perhaps thousands of "newsservers," the Google
Groups archive, at least a handful of private news archives
(including mine, BTW), and at least a dozen of newsreaders.

Now, I wonder, how did you end up effectively abandoning your
exclusive rights like that?

[...]

--
FSF associate member #7257 http://hfday.org/

[Ivan Shmakov]

>>>>> Rod Pemberton <do_no...@notemailnotq.cpm> writes:
>>>>> Alexei A. Frounze" <alexf...@gmail.com> wrote in message...

[Cross-posting to news:misc.legal.computing.]

[...]

> I believe in both PD and copyrighted code. I don't believe in FOSS
> code like GPL or LGPL.

JFTR: public domain, BSD, and the usual flavors of copyleft are
/all/ valid FOSS licenses.

There's a known problem with PD, however, due to the fact that
the very such notion is notoriously absent from certain (as in:
the most of) jurisdictions worldwide. Thus, for those wishing
to disclaim /all/ their rights to the code, I'd recommend to use
the CC0 "license" instead.

> Those licenses are too restrictive to be called "open source", and
> they have an political agenda: anti-company.

I'd argue that they have no such agenda. In fact, such copyleft
licenses prohibit the further encumbrance of the code by /both/
companies /and/ individuals.

(The anti-company /sentiment/ may be due to the fact that it's
indeed the companies that may, given the opportunity to take
over the code, boost its development so that the now-encumbered
result is simply out of competition with the unencumbered one.)

Instead, I'd say that the GPL agenda is /pro-community/, for
soon after the code is published under such a license, even its
original author may find that he or she cannot "take it back"
under his or her exclusive control.

(Assuming that some such community arisen, anyway.)

[...]

> What's interesting is that a non-Socialist country like the US allows
> code without a copyright (not allowed by law, forfeited copyright, or
> copyright term has expired, etc.) which allows for code that is
> effectively owned by all Americans, or anyone else for that matter.

Only until such code ends up in a derivative work. And as soon
as it does -- poof! -- the code is now effectively "owned" by
the author of the derivative work.

[...]

>> How many strings are in a FreeBSD license?

> Not enough, IMO. There are no prohibitions against combining minimal
> term MIT or BSD licensed code with more restrictive GPL or LGPL
> licensed code. There is no "string" in MIT or BSD licensed code
> requiring that additions, changes, or modifications to the code also

> fall under the original MIT or BSD copyright. I. e., there is no

> usage exclusion for abusive open source licenses like the GPL.

That's the very point of copyleft: a copyleft license is a free
software license which explicitly disallows further restriction.

Which makes me wonder, should the BSD license /have/ such a
string, how would you distinguish it from the GPL proper?

[Rod Pemberton]

"Ivan Shmakov" <onei...@gmail.com> wrote in message
news:87ehe6ed...@violet.siamics.net...
> Rod Pemberton <do_no...@notemailnotq.cpm> writes:
...

[FYI, original copyright discussion is mixed within
"public domain bootsector" on alt.os.development.]

> > I believe in both PD and copyrighted code. I don't
> > believe in FOSS code like GPL or LGPL.
>
> JFTR: public domain, BSD, and the usual flavors of
> copyleft are /all/ valid FOSS licenses.

JFTR: You're wrong. Public Domain code is not licensed. Code
must be copyrighted to be licensed. Unfortunately, PD code is not
copyrighted, therefore it can't be licensed. What you meant to
say, but IMO didn't do so correctly is: the FSF and OSI consider
those forms of source code to all be legally compatible. That's
different from saying uncopyrightable code is licensed under
copyright - which is flat out wrong. Unfortunately, that's a
side-effect of memorizing and parroting the standard copyleft
rhetoric to me. You didn't actually think about what you said.

> > Those licenses are too restrictive to be called "open
> > source", and they have an political agenda: anti-company.
>
> I'd argue that they have no such agenda. In fact, such copyleft
> licenses prohibit the further encumbrance of the code by /both/
> companies /and/ individuals.
>

Unfortunately, there is now some loss of context here. "Those"
referred explicitly to the GPL and LGPL. It did not refer to all
copyleft licenses as indicated by your respnse.

> Instead, I'd say that the GPL agenda is /pro-community/, for
> soon after the code is published under such a license, even its
> original author may find that he or she cannot "take it back"
> under his or her exclusive control.

Unfortunately, that's simply untrue.

Under US copyright law, the copyright holder has exclusive
copyrights to the original work. That means the copyright holder
can license his copyrights under any other licenses. That also
means the copyright holder can rescind any license such as the GPL
at any time whether there is a license term prohibiting it or not.
Copyright's are rights. The license isn't. There is no contract
enforcing the license terms. The copyright holder is also granted
copyrights to *ALL* derivative works which do not constitute a
new and original work in their own right, i.e., extremely major
changes. This condition is to prevent theft of copyright via
trivial modifications. That means that the vast majority of
contributions, changes, modifications by others to a GPL'd program
become the copyrighted property of the original author under US
law. So, basically, if you contribute to GPL licensed program,
you're working for free for the original copyright holder since
you've forfitted your rights to him/her due to copyright law.

As for the original author not "taking it back", first they must
prove financial losses to have a valid court case in the US.
That's unlikely for the given situation of it being free software.
But, if they could manage that, and they proved in a court of law
that others were infringing on their copyright, those who are
infringing are guilty of a crime and can be punished by the law.
However, as pointed out by an alt (anarchist, lunatic, and
terrorist) in the original a.o.d. thread, most GPL code is not
high value or high quality, or at least not perceived to be, so
it's unlikely many court cases are going to arise to test any of
these issues.

> > What's interesting is that a non-Socialist country like the
> > US allows code without a copyright (not allowed by law,
> > forfeited copyright, or copyright term has expired, etc.)
> > which allows for code that is effectively owned by all
> > Americans, or anyone else for that matter.
>
> Only until such code ends up in a derivative work. And as soon
> as it does -- poof! -- the code is now effectively "owned" by
> the author of the derivative work.

False. That's not how US copyright law works. See above.

> >> How many strings are in a FreeBSD license?
>
> > Not enough, IMO. There are no prohibitions against combining
> > minimal term MIT or BSD licensed code with more restrictive
> > GPL or LGPL licensed code. There is no "string" in MIT or
> > BSD licensed code requiring that additions, changes, or
> > modifications to the code also fall under the original MIT or
> > BSD copyright. I. e., there is no usage exclusion for
> > abusive open source licenses like the GPL.
>
> That's the very point of copyleft: a copyleft license is a free
> software license which explicitly disallows further restriction.
>

While I understand the backwards rhetoric (it does one thing but
everyone supportive claims it does another), just how do you
rationalize the extreme differences between a simple 2, 3, or 4
term license like MIT or BSD that has near zero further
restrictions while the GPL has literally pages upon pages of
restrictions? I.e., the point, which you seemed to have missed,
is that I believe the GPL and LGPL shouldn't be allowed to be
merged (linking is fine) with substantially "free-er" licensed
code like MIT and BSD. Even though the GPL is the first widely
adopted open-source license, it's _extremely_ restrictive compared
to almost all others that have come after, i.e., it's NOT a free
and open-source software license ANYMORE by comparison.

The problem of merging code with different licenses is license
pollution. When multiple authors work on the same program, a
single piece of code can have a variety of authors claiming
copyrights under a variety of license in a single source code
file. Unfortunately, this isn't legal under US law. US law only
allows a work to be copyrighted under a single copyright. I.e.,
mixing of code under different license within a single file isn't
legal. If the code for each is in different files, it's not an
issue. Normally, this causes simpler licenses like BSD and MIT to
become polluted with a number of non-legitimate GPL copyright
claims that are difficult to eliminate from the code. The problem
is it's indeterminable as to what copyright portions of the code
falls under - only one is allowed by US law. So, in effect, the
more restrictive GPL applies to the entire work until proven
otherwise, even if all but one line of code was non-GPL. So, the
GPL has become the "slap-on" license that you can't "slap-off".
It's used by GPL afficionados in an attempt to make all modifiable
code GPL whether it actually is or not. This also affects a
number of Public Domain projects currently, but would disadvantage
GPL copyright claims. The Public Domain code can't be copyrighted
under US law. It either was copyright once and can't be again, or
it is prohibited by law from being copyrighted. GPL authors have
added small snippets of code for which they've claimed GPL
copyrights. However, the work can only have a single copyright,
the original. In these cases, that "copyright" is Public Domain
and as such can't be copyrighted or have copyright claims like GPL
applied to it. I.e., if such a case ever makes it to court, the
GPL authors will forfit their claims to the Public Domain
"copyright".

> Which makes me wonder, should the BSD license /have/ such a
> string, how would you distinguish it from the GPL proper?

Honestly, I don't see the point you're attempting to make. Are
you equating or conflating the two because BSD license would
have one additional restrictive clause? Absurd.


Rod Pemberton

[Ivan Shmakov]

>>>>> Rod Pemberton <do_no...@notemailnotq.cpm> writes:
>>>>> "Ivan Shmakov" <onei...@gmail.com> wrote...
>>>>> Rod Pemberton <do_no...@notemailnotq.cpm> writes:

[...]

>>> I believe in both PD and copyrighted code. I don't believe in FOSS
>>> code like GPL or LGPL.

>> JFTR: public domain, BSD, and the usual flavors of copyleft are
>> /all/ valid FOSS licenses.

> JFTR: You're wrong. Public Domain code is not licensed. Code must
> be copyrighted to be licensed. Unfortunately, PD code is not
> copyrighted, therefore it can't be licensed. What you meant to say,
> but IMO didn't do so correctly is: the FSF and OSI consider those
> forms of source code to all be legally compatible.

What I've meant to say is that these satisfy the
well-established definitions of both free software and open
source software (of which the acceptance of these "terms" by
FSF, OSI, Debian, etc. is a mere consequence.)

But thanks for the clarification, anyway.

[...]

>>> Those licenses are too restrictive to be called "open source", and
>>> they have an political agenda: anti-company.

>> I'd argue that they have no such agenda. In fact, such copyleft
>> licenses prohibit the further encumbrance of the code by /both/
>> companies /and/ individuals.

> Unfortunately, there is now some loss of context here. "Those"
> referred explicitly to the GPL and LGPL. It did not refer to all

> copyleft licenses as indicated by your response.

Yes. What copyleft licenses do you accept specifically, BTW?

And I'm still curious if you will insist on the "agenda" part?

>> Instead, I'd say that the GPL agenda is /pro-community/, for soon
>> after the code is published under such a license, even its original
>> author may find that he or she cannot "take it back" under his or
>> her exclusive control.

> Unfortunately, that's simply untrue.

> Under US copyright law, the copyright holder has exclusive copyrights
> to the original work. That means the copyright holder can license
> his copyrights under any other licenses. That also means the
> copyright holder can rescind any license such as the GPL at any time
> whether there is a license term prohibiting it or not. Copyright's
> are rights. The license isn't. There is no contract enforcing the
> license terms.

That would seem to suggest that the license is non-binding to
the copyright holder who issued it. IANAL, yet I doubt that.
Could you please provide some evidence proving your point?

> The copyright holder is also granted copyrights to *ALL* derivative
> works which do not constitute a new and original work in their own

> right, i. e., extremely major changes. This condition is to prevent

> theft of copyright via trivial modifications. That means that the
> vast majority of contributions, changes, modifications by others to a
> GPL'd program become the copyrighted property of the original author
> under US law.

Which doesn't seem to align with FSF insisting on the transfer
of copyright to them for any non-trivial (and /not/ "extremely
major") changes to the code they hold their copyright over.

If what you're saying is true, they /already/ hold the copyright
over all such published changes.

(And: they /have/ a lawyer.)

[...]

> As for the original author not "taking it back", first they must
> prove financial losses to have a valid court case in the US.

Indeed. However, this has become somewhat a shaky ground these
days, so I'd assume that even the lawyers employed by the
"content providers," or the state, can (and will) get it wrong.

[...]

> However, as pointed out by an alt (anarchist, lunatic, and terrorist)
> in the original a.o.d. thread, most GPL code is not high value or
> high quality, or at least not perceived to be, so it's unlikely many
> court cases are going to arise to test any of these issues.

Perhaps.

>>> What's interesting is that a non-Socialist country like the US
>>> allows code without a copyright (not allowed by law, forfeited
>>> copyright, or copyright term has expired, etc.) which allows for
>>> code that is effectively owned by all Americans, or anyone else for
>>> that matter.

>> Only until such code ends up in a derivative work. And as soon as
>> it does -- poof! -- the code is now effectively "owned" by the
>> author of the derivative work.

> False. That's not how US copyright law works. See above.

So, unless "extreme major changes" are done to the code in
question, it remains under the public domain?

Still, unlike the case of copyleft, the user of a public domain
program may have trouble altering it, or learning from it.

(My point regarding the absence of the very notion of public
domain in many other jurisdictions still holds.)

[...]

> While I understand the backwards rhetoric (it does one thing but
> everyone supportive claims it does another), just how do you
> rationalize the extreme differences between a simple 2, 3, or 4 term
> license like MIT or BSD that has near zero further restrictions while
> the GPL has literally pages upon pages of restrictions?

That's simple: the latter is more detailed.

(But I wonder if there're any specific restrictions you disagree
with or reject, and why?)

Not to mention that it has "pages upon pages" of permissions
just as well.

> I. e., the point, which you seemed to have missed, is that I believe

> the GPL and LGPL shouldn't be allowed to be merged (linking is fine)
> with substantially "free-er" licensed code like MIT and BSD.

I do not understand. Do you ask for a license which prohibits
merging of the code covered with a "less-free" GPL-licensed
code, yet /allows/ the very same code to be merged with the code
released under "less-free-still" proprietary licenses?

> Even though the GPL is the first widely adopted open-source license,
> it's _extremely_ restrictive compared to almost all others that have

> come after, i. e., it's NOT a free and open-source software license
> ANYMORE by comparison.

That's irrelevant as long as it fits the well-known definitions
thereof. (Anyone's still free to suggest his or her own
definition, yet I doubt that it'd be adopted quickly.)

> The problem of merging code with different licenses is license
> pollution. When multiple authors work on the same program, a single
> piece of code can have a variety of authors claiming copyrights under
> a variety of license in a single source code file. Unfortunately,
> this isn't legal under US law. US law only allows a work to be

> copyrighted under a single copyright. I. e., mixing of code under

> different license within a single file isn't legal. If the code for
> each is in different files, it's not an issue.

That seem to suggest that US law equates "file" and "work,"
which I doubt is the case.

For instance, a statically-linked executable will contain all
the "creative components" right within a single file. Do you
suggest that such executables aren't "legal" under US law?

And what about a song, whose lyrics are copyrighted by one
person, yet the music by one another?

[...]

>> Which makes me wonder, should the BSD license /have/ such a string,
>> how would you distinguish it from the GPL proper?

> Honestly, I don't see the point you're attempting to make. Are you
> equating or conflating the two because BSD license would have one
> additional restrictive clause?

Indeed I do. How do you letter such a clause? Will it be like
"it's explicitly forbidden to merge this code with that under
GNU GPL"? Then, there also is "GNU AGPL," etc., and new ones
may spring at any time. Or will it be "provided that you don't
merge the code with that under a restrictive license"? Fine,
you've made it impossible to incorporate the code into non-free
software just as well, thus effectively making it copyleft,
which the proponents of "unencumbered" free software are often
point out as a "flaw" of copyleft.

[Tony]

In article <kkrla0$p2b$1...@speranza.aioe.org>, kl...@unibwm.de says...

>
> On 19.04.2013 05:17, Tony wrote:
>
>
> > There is only one definition of 'free' that can apply to software, because
> > software is inanimate, and that definition relates to price.
>
> The "free" in free software has nothing to do with money.

I ignore perversions of words which attempt to propagate political agenda and I
ignore those who do that.

[Tony]

In article <87ip3ie...@violet.siamics.net>, onei...@gmail.com says...

Is that a trick question?

[kerravon]

On Apr 19, 1:14 pm, Tony <a...@some.org> wrote:
> > I don't want any strings whatsoever. If people wish
> > to acknowledge other parties, that's their choice.
> > I don't wish to be a dog in a manger, and I don't
> > wish to force companies or others to do something
> > they don't really want to do.
>
> > BFN.  Paul.
>
> "Gimme, gimme!"? Are you one of those characters that continually seek to get
> something for nothing? Feeling entitled?

If anything could describe me, it would be "give you,
give you" not "gimme, gimme". I've spent decades
producing public domain software.

And people are free to want something for nothing
under my philosophy. My code is there, waiting and
willing.

BFN. Paul.

[kerravon]

On Apr 16, 1:14 pm, "Alexei A. Frounze" <alexfrun...@gmail.com> wrote:
> On Apr 15, 7:08 am, kerravon <kerra...@w3.to> wrote:
>

>
> Do you want me to release my FAT12, FAT16 and FAT32 bootsectors under
> FreeBSD license?
>
> The FAT12 and FAT16 versions are available inhttp://alexfru.narod.ru/os/fat/fat-2006-12-03.zip
> (FAT12.ASM and FAT16.ASM under _boot). The code is for NASM.
>
> Alex

Hi Alex.

I am interested in seeing the mathematical manipulations you
do on the bpb. But I got an error with your file. I tried downloading
twice, same result.

C:\scratch\ddd>unzip \download\fat-2006-12-03.zip
Archive: /download/fat-2006-12-03.zip
End-of-central-directory signature not found. Either this file is
not
a zipfile, or it constitutes one disk of a multi-part archive. In
the
latter case the central directory and zipfile comment will be found
on
the last disk(s) of this archive.
unzip: cannot find zipfile directory in /download/fat-2006-12-03.zip,
and cannot find /download/fat-2006-12-03.zip.zip, period.

Any ideas?

Thanks. Paul.

[kerravon]

On Apr 18, 12:28 pm, "Alexei A. Frounze" <alexfrun...@gmail.com>
wrote:

> On Apr 17, 12:45 am, "Rod Pemberton" <do_not_h...@notemailnotq.cpm>
> wrote:

> My preference of BSD/MIT over the default copyright or PD or GPL is
> simple:
> 1. I want to share the code

PD is the best for that.

> 2. I don't want to impose severe restrictions on what can and can't be
> done with it, even if someone drops it into the depths of GPL'd code
> (I might revise my view of GPL in the future, but at the moment I
> don't care much about it tainting versions of my code produced by
> others)

PD is best for that.

> 3. I don't want to require others to pay to me or to obtain written
> permissions from me or anything alike

PD is just as good at that.

> 4. I want my name to remain there, which I think is a fair non-
> monetary price to ask (unless you want to be anal and count every
> character of the license and multiply it by the storage and
> transmission cost of a single byte times the number of repetitions)

My preference is that you would receive a monetary price
for your work, and that in turn you relinquish all copyright.
Someone like the CSIRO (Australian research) should be
writing this code, not me and you. They should be writing
it and releasing it under the public domain, as their job.
As Australia's contribution to the world of software.

In the absence of a government (North Korea, anyone)
doing the right thing, I'm trying to compensate for the
malfeasance. How much would you charge to make your
code public domain? How much would you charge for a
universal boot sector that works on all of FAT 12/16/32,
but just reads 3 consecutive sectors?

BFN. Paul.

[Tony]

In article <82674f67-8738-423b...@l2g2000pbn.googlegroups.com>,
kerr...@w3.to says...

>
> On Apr 19, 1:14 pm, Tony <a...@some.org> wrote:
> > > I don't want any strings whatsoever. If people wish
> > > to acknowledge other parties, that's their choice.
> > > I don't wish to be a dog in a manger, and I don't
> > > wish to force companies or others to do something
> > > they don't really want to do.
> >
> > > BFN.  Paul.
> >
> > "Gimme, gimme!"? Are you one of those characters that continually seek to get
> > something for nothing? Feeling entitled?
>
> If anything could describe me, it would be "give you,
> give you" not "gimme, gimme". I've spent decades
> producing public domain software.

OK. If the software was good, I hope you prospered from it some way.

>
> And people are free to want something for nothing
> under my philosophy. My code is there, waiting and
> willing.
>

People with the ability to pay but the unwillingness to pay, irk me. Especially
those who exploit the unfortunate.

[kerravon]

Actually, what I would like most is for my existing code to
be updated with the required maths. How much would you
charge to get this to work:

http://pdos.cvs.sourceforge.net/viewvc/pdos/pdos/src/pbootsec.asm?view=markup

You can add your name to it, and also say that your mods
are PD. I can tell you that I will be happy to see your name
in the code - one of the contributors to PDOS. But I refuse
to guarantee that neither I nor someone else will remove
your name, since the material is all PD.

BFN. Paul.

[Ivan Shmakov]

>>>>> Tony <a...@some.org> writes:
>>>>> In article <kkk3hc$c15$1...@speranza.aioe.org>, kl...@unibwm.de says...

[Cross-posting to news:misc.legal.computing, just in case.]

[...]

>> To release software into the Public Domain doesn't make it free, the
>> opposite is true, everybody has the freedom to enslave it.

> Software is inanimate and therefore cannot be enslaved.

Indeed, I tend to agree with that.

However, the /users/ of software are typically "animate," and
/can/ be "enslaved." E. g., as in the case of vendor lock-in.
And by definition, free software is the one which grants its
users certain freedoms, such as the freedom to learn from it,
the freedom to improve, and the freedom to share (both improved
and unimproved versions.) Certainly, one may argue that the
choice of the term is unfortunate or misleading, but it's not
unlike arguing about a definition of velocity or momentum.

Alas, I have to admit that I have to use non-free software at
times. For instance, I have this "funny" BIOS which insists on
making a copy of itself on a hard disk -- a feature which cannot
be disabled (without re-building it, anyway; which, however,
requires vendor-specifc software for some 1000 USD or so.)

Not at all unsurprising is the fact that by writing itself to
the end of a hard disk, this BIOS both screws up the GPT's
backup, /and/ sets up a HPA incompatible with the primary copy,
thus instantly making my system unbootable.

[Herbert Kleebauer]

On 20.04.2013 06:22, Tony wrote:

>>> There is only one definition of 'free' that can apply to software, because
>>> software is inanimate, and that definition relates to price.
>>
>> The "free" in free software has nothing to do with money.
>
> I ignore perversions of words which attempt to propagate political agenda and I
> ignore those who do that.

Yes, ignore reality and continue to live in your fantasy
world as long as you can. Maybe you should read:

http://www.gnu.org/philosophy/free-sw.html :

Thus, �free software� is a matter of liberty, not price.
To understand the concept, you should think of �free� as
in �free speech,� not as in �free beer�.

[kerravon]

On Apr 20, 4:30 pm, Tony <a...@some.org> wrote:
> In article <82674f67-8738-423b-a63e-e1532c319...@l2g2000pbn.googlegroups.com>,
> kerra...@w3.to says...

>
> > If anything could describe me, it would be "give you,
> > give you" not "gimme, gimme". I've spent decades
> > producing public domain software.
>
> OK. If the software was good, I hope you prospered from it some way.

Well what I did was a mammoth engineering exercise.
The goal was to make the world a better place, not
personal prosperity. I'm satisfied with what has been
achieved, even though it is not the sort of thing that
is recognized.

> > And people are free to want something for nothing
> > under my philosophy. My code is there, waiting and
> > willing.
>
> People with the ability to pay but the unwillingness to pay, irk me. Especially
> those who exploit the unfortunate.

I don't consider people wanting to save money to be a
bad thing. I give my code away freely even if billionaires
are using it. I'm quite happy that anyone at all uses my
code which was decades in the making.

BFN. Paul.

[kerravon]

On Apr 20, 4:35 pm, kerravon <kerra...@w3.to> wrote:
> On Apr 20, 4:17 pm, kerravon <kerra...@w3.to> wrote:
>
> > On Apr 18, 12:28 pm, "Alexei A. Frounze" <alexfrun...@gmail.com>
> > wrote:
> Actually, what I would like most is for my existing code to
> be updated with the required maths.

Updated in masm syntax (to be compiled with wasmr), not nasm.

BFN. Paul.

[Tony]

In article <370f7aa6-d58b-4275...@di5g2000pbc.googlegroups.com>,
kerr...@w3.to says...

>
> On Apr 20, 4:30 pm, Tony <a...@some.org> wrote:
> > In article <82674f67-8738-423b-a63e-e1532c319...@l2g2000pbn.googlegroups.com>,
> > kerra...@w3.to says...
> >
> > > If anything could describe me, it would be "give you,
> > > give you" not "gimme, gimme". I've spent decades
> > > producing public domain software.
> >
> > OK. If the software was good, I hope you prospered from it some way.
>
> Well what I did was a mammoth engineering exercise.
> The goal was to make the world a better place, not
> personal prosperity. I'm satisfied with what has been
> achieved, even though it is not the sort of thing that
> is recognized.

Well maybe that is a good reason to SELL your wares instead of giving them away.
That way, people have some idea of the market value of them, their usefulness,
etc. Consider that a singer can sing songs all day long, giving away his "gift
to the world", but with a really bad voice, a lot of people may even pay him to
stop!

Then, what you do with the money, is your own business. The only way I know how
to make thinks better is to build things, fix things (not necessarily only
technical things), but to do that, I need resources and build-up some facility
for doing bigger and better things. The only way I know how to do that, is to
make something, and sell it. I have no problems with traditional "hang out a
shingle and offer wares". It's not like I think money is inherently evil, nor is
wanting to get paid for products and services.

Other approaches and motivations are, of course, fine. To each their own (as
long as it doesn't impose upon others).

>
> > > And people are free to want something for nothing
> > > under my philosophy. My code is there, waiting and
> > > willing.
> >
> > People with the ability to pay but the unwillingness to pay, irk me. Especially
> > those who exploit the unfortunate.
>
> I don't consider people wanting to save money to be a
> bad thing. I give my code away freely even if billionaires
> are using it. I'm quite happy that anyone at all uses my
> code which was decades in the making.
>

If I was buying from D. Trump (though I'd never buy from him if I didn't have to
because I would instead seek out someone who had less or was needy and could
provide adequate product or service), I'd want to "save money". If someone is
already below the line, I'm not going to bicker on price with him.

[Tony]

In article <kktd7f$f7o$1...@speranza.aioe.org>, kl...@unibwm.de says...

>
> On 20.04.2013 06:22, Tony wrote:
>
> >>> There is only one definition of 'free' that can apply to software, because
> >>> software is inanimate, and that definition relates to price.
> >>
> >> The "free" in free software has nothing to do with money.
> >
> > I ignore perversions of words which attempt to propagate political agenda and I
> > ignore those who do that.
>
> Yes, ignore reality and continue to live in your fantasy
> world as long as you can. Maybe you should read:
>
> http://www.gnu.org/philosophy/free-sw.html :
>

> Thus, “free software” is a matter of liberty, not price.
> To understand the concept, you should think of “free” as
> in “free speech,” not as in “free beer”.

Y'all go on, and on, and on, and on... with your same spiel all the time, like
those religious groups who come knocking on your door trying to convince you
that the sky is falling and you must prepare for it. I'm never ever going to
join yours or anyone else's religion, so don't waste your breath.

Y'all really need to read up on why hard-sell tactics don't work and how they
can actually reduce sales and damage a brand or company reputation. Y'all are
way beyond that level with me, FYI.

[Tony]

In article <87y5cdd7...@violet.siamics.net>, onei...@gmail.com says...

>
> >>>>> Tony <a...@some.org> writes:
> >>>>> In article <kkk3hc$c15$1...@speranza.aioe.org>, kl...@unibwm.de says...
>
> [Cross-posting to news:misc.legal.computing, just in case.]
>
> [...]
>
> >> To release software into the Public Domain doesn't make it free, the
> >> opposite is true, everybody has the freedom to enslave it.
>
> > Software is inanimate and therefore cannot be enslaved.
>
> Indeed, I tend to agree with that.
>

There is only one definition of 'free software': software for which the monetary
price is zero. Nothing more to say beyond that. That is all it can be. Anything
else is bullshit. If you want to be a bullshitter, that is your perogative.

Software and source code, BTW, are different things.

[Ivan Shmakov]

>>>>> Tony <a...@some.org> writes:

[...]

> There is only one definition of 'free software': software for which
> the monetary price is zero.

[...]

ACK, thanks. Though I'd rather paraphrase it as follows:

There is only one definition of stubbornness, and it's the one where
I am right, and you are wrong.

[kerravon]

On Saturday, April 20, 2013 5:12:58 PM UTC+10, Tony wrote:
> In article <370f7aa6-d58b-4275...@di5g2000pbc.googlegroups.com>,

>

> > Well what I did was a mammoth engineering exercise.
> > The goal was to make the world a better place, not
> > personal prosperity. I'm satisfied with what has been
> > achieved, even though it is not the sort of thing that
> > is recognized.
>
> Well maybe that is a good reason to SELL your wares
> instead of giving them away.

I sell my skills commercially already. At home
I spend my time compensating for what I consider
to be the failure of every government in the
world. Every government in the world should have
deemed the Microsoft monopoly as unacceptable,
and asked the nation's R&D to develop a rival.

BFN. Paul.

[Tony]

In article <0da9f69c-da51-442b...@googlegroups.com>,
kerr...@w3.to says...

So if I read you correctly, for lack of the ability to create a better
mousetrap, you want some kind of government protection? Or want to declare war
on Microsoft? It sounds like you are whining.

[Tony]

In article <87txn1d...@violet.siamics.net>, onei...@gmail.com says...

>
> >>>>> Tony <a...@some.org> writes:
>
> [...]
>
> > There is only one definition of 'free software': software for which
> > the monetary price is zero.
>
> [...]
>
> ACK, thanks. Though I'd rather paraphrase it as follows:
>
> There is only one definition of stubbornness, and it's the one where
> I am right, and you are wrong.

You just made yourself look really stupid.

[Ivan Shmakov]

>>>>> kerravon <kerr...@w3.to> writes:

[...]

> I sell my skills commercially already. At home I spend my time
> compensating for what I consider to be the failure of every
> government in the world.

Could you please summarize your success so far? In particular,
what is the estimated number of installations of the software
already written?

TIA.

> Every government in the world should have deemed the Microsoft
> monopoly as unacceptable, and asked the nation's R&D to develop a
> rival.

There are rivals, though I don't quite understand why it's the
nations' R&Ds that should be in charge for their development?

Somehow, I feel the software development done by international
teams (as it tends to be in the case of free software in
general) to be equally valid, if not better, to that done by the
within-the-borders teams.

[Ivan Shmakov]

>>>>> kerravon <kerr...@w3.to> writes:
>>>>> On Apr 18, 12:28 pm, "Alexei A. Frounze" wrote:

[Cross-posting to news:misc.legal.computing.]

>> My preference of BSD/MIT over the default copyright or PD or GPL is
>> simple: 1. I want to share the code

> PD is the best for that.

I believe it isn't, for there are jurisdictions lacking the very
notion of public domain.

In such jurisdictions, the code which the author declares public
domain effectively cannot be legally used at all, unless the
author also provides a "usual" license (even if one having the
very same effect as a public domain dedication would have.)

For this reason, I'd recommend using a CC0 public domain
dedication instead [1].

[1] http://creativecommons.org/about/cc0

[...]

>> 4. I want my name to remain there, which I think is a fair
>> non-monetary price to ask

[...]

> My preference is that you would receive a monetary price for your
> work, and that in turn you relinquish all copyright.

Irrespective of one's own preferences, certain jurisdictions may
require that the author is attributed for his or her creative
work or contribution. Effectively, it may not be possible for
the author to abandon his or her right to be attributed.

[...]

[kerravon]

On Saturday, April 20, 2013 6:37:32 PM UTC+10, Tony wrote:
> In article <0da9f69c-da51-442b...@googlegroups.com>,

>
> So if I read you correctly, for lack of the ability to create a better
> mousetrap, you want some kind of government protection? Or want to declare war
> on Microsoft? It sounds like you are whining.

Our societies are capitalist, and to a major
extent, that works fine.

However, from time to time, capitalist societies
produce harmful monpolies. When that happens,
government can and should step in to end the
monopoly.

People have ummed and aahed about the Microsoft
(Windows) and IBM (z/OS) monpolies, but the
monopolies remain unscathed.

Practically any government in the world has the
ability to set up a team to break those
monopolies.

If you consider that to be war/government protection/
whining, that is your choice. From my perspective
it is just the most logical reaction to the stimuli.

BFN. Paul.

[Mike Gonta]

"kerravon" wrote:

> I am interested in seeing the mathematical manipulations you
> do on the bpb.

The LBA of the first of three contiguous sectors of the first file
written to a newly formatted media with 512 byte sectors is calculated
as follows.

first = bpb.reserved_sector_count + bpb.hidden_sectors;
if(bpb.sectors_per_fat){ // FAT12/FAT16
first += bpb.number_of_fats * bpb.sectors_per_fat;
first += bpb.directory_entries >> 4; // 16 (32 byte) entries per sector
}else{ // FAT32
first += bpb.number_of_fats * ebpb.sectors_per_fat;
first += bpb.sectors_per_cluster; // FAT32 root directory
}


Mike Gonta
look and see - many look but few see

http://mikegonta.com

[kerravon]

On Apr 20, 7:15 pm, Ivan Shmakov <oneing...@gmail.com> wrote:

> >>>>> kerravon  <kerra...@w3.to> writes:
>
> [...]
>
>  > I sell my skills commercially already.  At home I spend my time
>  > compensating for what I consider to be the failure of every
>  > government in the world.
>
>         Could you please summarize your success so far?

Well thanks to PDPCLIB, I was able to port GCC to MVS,
CMS and VSE (mainframe systems). This (in my opinion)
made C a universal language. ie there is no computer
system in commercial use that doesn't have a standard,
or free, C compiler.

It was a huge engineering effort, but not something that
a large number of people wanted. I only know of 1 guy
(Mike Noel) who is using GCCMVS to develop KICKS
(a rival to CICS).

But at my place of employment, I was able to respond to
a customer request with "here is some C code that will
do what you want, and here is a free C compiler if you
need it". It's all worth it to be able to do that.

>  In particular,
>         what is the estimated number of installations of the software
>         already written?

173 people have downloaded this version of GCCMVS:

https://sourceforge.net/projects/gccmvs/files/GCCMVS/GCC%203.2.3%20MVS%208.5/gccmvs-3_2_3-8_5-xmit.zip/stats/timeline?dates=2010-04-14+to+2013-04-20

And 623 have downloaded the latest version:

https://sourceforge.net/projects/gccmvs/files/GCCMVS/GCC%203.4.6%20MVS%201.0/gccmvs-3_4_6-1_0-xmit.zip/stats/timeline?dates=2010-04-14+to+2013-04-20

(which has a bug in it, which I can't fix, so I don't use it).

The C runtime library has had 892 downloads of the latest version:

https://sourceforge.net/projects/pdos/files/pdpclib/pdpclib%203.10/pdpc310.zip/stats/timeline?dates=2010-04-14+to+2013-04-20

PDOS has 1115 downloads:

https://sourceforge.net/projects/pdos/files/pdos/pdos%200.86/stats/timeline?dates=2001-04-14+to+2013-04-20

But PDOS is very immature so I don't think those 1115 people are doing
anything with it.

My challenge is to make PDOS mature.

>  > Every government in the world should have deemed the Microsoft
>  > monopoly as unacceptable, and asked the nation's R&D to develop a
>  > rival.
>
>         There are rivals

I don't consider there to be a rival to Windows. Windows
has made itself industry standard.

> , though I don't quite understand why it's the
>         nations' R&Ds that should be in charge for their development?

Because history has shown that no company is willing to
fund the building of a rival to Windows. Government
should step in when that happens.

>         Somehow, I feel the software development done by international
>         teams (as it tends to be in the case of free software in
>         general) to be equally valid, if not better, to that done by the
>         within-the-borders teams.

Yes, ideally there should be a United Nations Operating System (UNOS).
We need one of those for the PC platform and one for the mainframe
platform. My currenty hobby is to fill both of those holes in the
software industry. ie I'm doing what I think the UN (or any member
of the UN) should be doing.

BFN. Paul.

[kerravon]

On Apr 20, 7:38 pm, Ivan Shmakov <oneing...@gmail.com> wrote:

> >>>>> kerravon  <kerra...@w3.to> writes:
> >>>>> On Apr 18, 12:28 pm, "Alexei A. Frounze" wrote:
>
>         [Cross-posting to news:misc.legal.computing.]
>
>  >> My preference of BSD/MIT over the default copyright or PD or GPL is
>  >> simple: 1. I want to share the code
>
>  > PD is the best for that.
>
>         I believe it isn't, for there are jurisdictions lacking the very
>         notion of public domain.
>
>         In such jurisdictions, the code which the author declares public
>         domain effectively cannot be legally used at all, unless the
>         author also provides a "usual" license (even if one having the
>         very same effect as a public domain dedication would have.)
>
>         For this reason, I'd recommend using a CC0 public domain
>         dedication instead [1].
>
> [1]http://creativecommons.org/about/cc0

Ok, I've updated my documentation to refer to that:

C:\devel\pdos>cvs diff readme.txt
Index: readme.txt
===================================================================
RCS file: /cvsroot/pdos/pdos/readme.txt,v
retrieving revision 1.20
diff -r1.20 readme.txt
8,9c8,9
< without restriction.
<
---
> without restriction, as discussed here:
> http://creativecommons.org/publicdomain/zero/1.0/


BFN. Paul.

[kerravon]

On Apr 20, 8:44 pm, "Mike Gonta" <mikego...@gmail.com> wrote:
> "kerravon" wrote:
> > I am interested in seeing the mathematical manipulations you
> > do on the bpb.
>
> The LBA of the first of three contiguous sectors of the first file
> written to a newly formatted media with 512 byte sectors is calculated
> as follows.
>
>   first = bpb.reserved_sector_count + bpb.hidden_sectors;
>   if(bpb.sectors_per_fat){ // FAT12/FAT16
>     first += bpb.number_of_fats * bpb.sectors_per_fat;
>     first += bpb.directory_entries >> 4; // 16 (32 byte) entries per sector
>   }else{ // FAT32
>     first += bpb.number_of_fats * ebpb.sectors_per_fat;
>     first += bpb.sectors_per_cluster; // FAT32 root directory
>   }
>

Thanks. There are also manipulations required to convert
the sector number into CHS.

And a major problem I have is that the boot sector is in
assembler, and my assembler has gone rusty. I was
wondering whether I could call a C subroutine to do the
manipulations as shown above. But the boot sector
requires things like an ID at the end of the sector. I think
it will be quite messy to try to use C at that point.

BFN. Paul.

[Alexei A. Frounze]

Try the copy from Ben's page at http://www.fysnet.net/zips/fat-2006-12-03.zip

Alex

[Alexei A. Frounze]

I haven't thought about that. What's a reasonable price in your
opinion?

> How much would you charge for a
> universal boot sector that works on all of FAT 12/16/32,
> but just reads 3 consecutive sectors?

I'm afraid a universal FAT12/16/32 bootsector might be impossible to
do. FAT32 is particularly complex and I remember I had trouble
squeezing it into the 512 bytes of the bootsector. Adding FAT12/16 on
top of that may really be impossible. But I'll take a look to see how
much can be saved by removing the .COM/.EXE-related part.

Alex

[Alexei A. Frounze]

On Apr 18, 10:36 pm, Tony <a...@some.org> wrote:
> In article <dc67c829-6e8f-49f3-b59a-1d5b12149...@id10g2000pbc.googlegroups.com>,
> alexfrun...@gmail.com says...

>
>
>
> > On Apr 17, 12:45 am, "Rod Pemberton" <do_not_h...@notemailnotq.cpm>
> > wrote:

> > > Alexei A. Frounze" <alexfrun...@gmail.com> wrote in messagenews:04e2aad5-dbd1-42dc...@di5g2000pbc.googlegroups.com...

> > My preference of BSD/MIT over the default copyright or PD or GPL is
> > simple:
> > 1. I want to share the code
>

> What does that mean?

>
> > 2. I don't want to impose severe restrictions on what can and can't be
> > done with it,
>

> Not even bomb-making? War machines? Awhile back, I did a tiny bit of grant
> research (i.e., I was looking for funding). I found something in the grant
> stipulation that I could not accept: that they could use the source code in any
> way they wanted to. War machines. I'll be god-damned, if I'm going to help some
> fucking government build their fucking war machines. They can go to hell with
> their fucking government and their fucking war machines and their fucking money.

>
> > even if someone drops it into the depths of GPL'd code
> > (I might revise my view of GPL in the future, but at the moment I
> > don't care much about it tainting versions of my code produced by
> > others)

> > 3. I don't want to require others to pay to me
>

> Why not? You aspire to be a mother Teresa?

>
> > or to obtain written
> > permissions from me or anything alike

> > 4. I want my name to remain there,
>

> So it is, then, fame that you seek? That does seem to be THE youthful aspiration
> these days (i.e., being famous rather than being useful).

>
> > which I think is a fair non-
> > monetary price to ask (unless you want to be anal and count every
> > character of the license and multiply it by the storage and
> > transmission cost of a single byte times the number of repetitions)
>

> > Alex
>
> Here's the thing. Where does it end (or begin even)? Does the reward of having
> your name in lights in a help dialog box a bit much for just any little thing?
> Where is the point where the parallels to graffiti perpetrators fits? Most code
> found on the net which is free and requires such seems like an idea for further
> research and development by the creator, and not useful for general use. I.e.,
> it is nowhere near the level of quality which could demand payment, hence why
> should it get recognition at all? The only way really to recognize it would be
> to show EXACTLY how and how much of the code was used, so as not to fall prey to
> graffiti tagger tactics. Also, I think most commercial developers would find it
> easier to just buy something and use it rather than having to introduce a
> process into their company that has to deal with a thousand stipulations of a
> thousand developers. (Just random thoughts/food for thought in this paragraph.
> I'm not saying you don't deserve credit or that your code is shit--I haven't
> seen it).

Tony, you seem to be enjoying opposing everyone and everything, no
matter who or what it is, or simply trolling. Would this be a correct
interpretation of your replies in this thread?

'cause I don't quite understand your opposition to my well meaning
code sharing intents.

Alex

[Alexei A. Frounze]

Remember also that my bootsector doesn't read sectors from predefined
locations, it reads a file from the root directory. If you don't want
that and can guarantee that the file isn't going to move (if, let's
say, you write it as the first on the newly formatted partition, mark
it as "system" and prohibit moving "system" files in your disk
defragment(at?)ors as DOS did), then I'm not sure what we're talking
about here. It should be a fairly trivial task.

Alex

[kerravon]

On Apr 20, 11:02 pm, "Alexei A. Frounze" <alexfrun...@gmail.com>
wrote:

> On Apr 20, 5:49 am, "Alexei A. Frounze" <alexfrun...@gmail.com> wrote:

Hi Alex. I was able to download your file from the
new location - thanks for that.

The filenames are different from what you said
though. But I found boot16.asm, and I can see
that you have put a lot of work into it.

> > I haven't thought about that. What's a reasonable price in your
> > opinion?

$20?

> locations, it reads a file from the root directory. If you don't want
> that and can guarantee that the file isn't going to move (if, let's
> say, you write it as the first on the newly formatted partition, mark
> it as "system" and prohibit moving "system" files in your disk
> defragment(at?)ors as DOS did), then I'm not sure what we're talking
> about here.

I am happy to live within the restriction that the file isn't
going to move, and needs to appear first on disk, with
consecutive sectors.

> It should be a fairly trivial task.

I hate programming. In particular I hate assembler
coding. My forte is fixing bugs in developed (by
others) C code which has a working printf. Also I
have ideas for code development.

But I am taking things one at a time. At the moment
my boot sector has hardcoded values in it which I
want to avoid.

So what is to you a trivial task is for me a small
nightmare. That's why it has taken me so long to
develop PDOS. Almost 2 decades. But I have now
acknowledged my limitations and seek to outsource
development of PDOS. My hope is to find Indian
etc programmers who are willing to work on PDOS
for $1/hour, which is $1 more than any other free
and open source project pays.

BFN. Paul.

[kerravon]

On Apr 20, 11:39 pm, kerravon <kerra...@w3.to> wrote:
> On Apr 20, 11:02 pm, "Alexei A. Frounze" <alexfrun...@gmail.com>
> wrote:

> > It should be a fairly trivial task.

Just to clarify - I'm not after a super boot sector like you
have. I'm after a simple one like I already have. I just
need mine fixed to get rid of the hardcoding.

mov cl, 26 ; +++ needs to be 010h for floppy
mov ch, 00h ; track 0
mov dh, 03h ; head ; +++ should be 01h for floppy

I'm expecting about 10-20 lines of code needs to be written
in order for this hardcoding to be replaced.

I am not after a complete revamp of the boot sector. I just
want my current design to be completed.

BFN. Paul.

[Mike Gonta]

"kerravon" wrote:

> There are also manipulations required to convert
> the sector number into CHS.

unsigned int sector = lba % bpb.sectors_per_track + 1;
unsigned int head = (lba / bpb.sectors_per_track) % bpb.number_of_heads;
unsigned int cylinder = lba / bpb.sectors_per_track / bpb.number_of_heads;

// al = number of sectors to read (write)
// ah = 2 (read)
// es:bx = buffer
unsigned short cx = ((cylinder & 0xFF) << 8) | (((cylinder & 0x300) >> 2)
| (sector & 0x3F));
// dl = drive id from BIOS on boot sector entry
unsigned char dh = head;

When booting from USB it is recommended to use int 0x13, ah=8 (get drive
parameters)
to obtain geometry. I have observed some versions of some BIOS return the
info that
is read from the bpb, but this is not a standard nor guaranteed.

> And a major problem I have is that the boot sector is in
> assembler, and my assembler has gone rusty. I was
> wondering whether I could call a C subroutine to do the
> manipulations as shown above. But the boot sector
> requires things like an ID at the end of the sector. I think
> it will be quite messy to try to use C at that point.

I am not suggesting that the code be written in C, this is merely the
"mathematics".
I am also not suggesting that the assembler code be written in this "C"
style.

[Tony]

In article <c76b92f3-c8ec-468a...@mf10g2000pbb.googlegroups.com>,
kerr...@w3.to says...

> Well thanks to PDPCLIB, I was able to port GCC to MVS,
> CMS and VSE (mainframe systems). This (in my opinion)
> made C a universal language.

OK. So you want to know if that is so, yes? Or maybe you care to be not burdened
with it more but celebrate it much?


I am not "I", but wouldn't I resent it if I was "C" and you were claiming to be
it?

[Tony]

In article <81b67614-6c47-4276...@pl9g2000pbb.googlegroups.com>,
kerr...@w3.to says...

>
> On Apr 20, 7:38 pm, Ivan Shmakov <oneing...@gmail.com> wrote:
> > >>>>> kerravon  <kerra...@w3.to> writes:
> > >>>>> On Apr 18, 12:28 pm, "Alexei A. Frounze" wrote:
> >
> >         [Cross-posting to news:misc.legal.computing.]
> >
> >  >> My preference of BSD/MIT over the default copyright or PD or GPL is
> >  >> simple: 1. I want to share the code
> >
> >  > PD is the best for that.
> >
> >         I believe it isn't, for there are jurisdictions lacking the very
> >         notion of public domain.
> >
> >         In such jurisdictions, the code which the author declares public
> >         domain effectively cannot be legally used at all, unless the
> >         author also provides a "usual" license (even if one having the
> >         very same effect as a public domain dedication would have.)
> >
> >         For this reason, I'd recommend using a CC0 public domain
> >         dedication instead [1].
> >
> > [1]http://creativecommons.org/about/cc0
>
> Ok, I've updated my documentation to refer to that:
>
>

Starving artist? What the fuck is your problem man, you gotta kill people to get
noticed. Just ask the rob who doesn't even realize he is a chatroom. I can't
help him, even if I wanted to. I didn't mean to intrude. I didn't think I was
"time-traveling".

You are not my problem. And when you die, your sins will die with you. You know
why they will? Because they are apparent.

[Tony]

In article <47ffd6c8-0f77-4657...@googlegroups.com>,
kerr...@w3.to says...

>
> On Saturday, April 20, 2013 6:37:32 PM UTC+10, Tony wrote:
> > In article <0da9f69c-da51-442b...@googlegroups.com>,
> >
> > So if I read you correctly, for lack of the ability to create a better
> > mousetrap, you want some kind of government protection? Or want to declare war
> > on Microsoft? It sounds like you are whining.
>
> Our societies are capitalist, and to a major
> extent, that works fine.
>

Don't you DARE "our" to me!! I'm an actor! (Ok, I'm a director, but I tried to
dance, and I'm still not afraid of it).

[Tony]

In article <47ffd6c8-0f77-4657...@googlegroups.com>,
kerr...@w3.to says...
>

> On Saturday, April 20, 2013 6:37:32 PM UTC+10, Tony wrote:
> > In article <0da9f69c-da51-442b...@googlegroups.com>,
> >
> > So if I read you correctly, for lack of the ability to create a better
> > mousetrap, you want some kind of government protection? Or want to declare war
> > on Microsoft? It sounds like you are whining.
>
> Our societies

Ya think? Do you really think that I am of "your society". Are you fucking
dense, or just fucking stupid? I am not you. I will never ever be you.

> are capitalist, and to a major
> extent, that works fine.
>
> However, from time to time, capitalist societies
> produce harmful monpolies. When that happens,
> government can and should step in to end the
> monopoly.

There is no war. You are the war.

>
> People

As if you knew people.

> have ummed and aahed about the Microsoft
> (Windows) and IBM (z/OS) monpolies, but the
> monopolies remain unscathed.

Don't even give me a reason to shut your mouth, I have known for 20 years that
Windows was paltry, and people aroound me knew that too, but I had loves to
seek, I couldn't land them apparently, but the technology attack is wrong,
wrong, wrong, and I have watched it for 20 years and you got nothing to attack
on the front of attacking Microsoft, because I've known the better mouse trap
for 2o years. From the get go Windows is wrong. Be it as it is, you couldn't
bring forth anything better.

You diss MS because you can't. I donno, I diss me cuz, you won? what else do you
want??!

shut up, stop whining.

[s_dub...@yahoo.com]

On Apr 19, 12:56 pm, "Rod Pemberton" <do_not_h...@notemailnotq.cpm>
wrote:
> "Ivan Shmakov" <oneing...@gmail.com> wrote in message
>
[snipped]
>
> > That's the very point of copyleft: a copyleft license is a free
> > software license which explicitly disallows further restriction.
>
> While I understand the backwards rhetoric (it does one thing but
> everyone supportive claims it does another), just how do  you
> rationalize the extreme differences between a simple 2, 3, or 4
> term license like MIT or BSD that has near zero further
> restrictions while the GPL has literally pages upon pages of
> restrictions?  I.e., the point, which you seemed to have missed,
> is that I believe the GPL and LGPL shouldn't be allowed to be
> merged (linking is fine) with substantially "free-er" licensed
> code like MIT and BSD.  Even though the GPL is the first widely
> adopted open-source license, it's _extremely_ restrictive compared
> to almost all others that have come after, i.e., it's NOT a free
> and open-source software license ANYMORE by comparison.
>
> The problem of merging code with different licenses is license
> pollution.  When multiple authors work on the same program, a
> single piece of code can have a variety of authors claiming
> copyrights under a variety of license in a single source code
> file.  Unfortunately, this isn't legal under US law.  US law only
> allows a work to be copyrighted under a single copyright.  I.e.,
> mixing of code under different license within a single file isn't
> legal.  If the code for each is in different files, it's not an
> issue.  Normally, this causes simpler licenses like BSD and MIT to
> become polluted with a number of non-legitimate GPL copyright
> claims that are difficult to eliminate from the code.  The problem
> is it's indeterminable as to what copyright portions of the code
> falls under - only one is allowed by US law.  So, in effect, the
> more restrictive GPL applies to the entire work until proven
> otherwise, even if all but one line of code was non-GPL.  So, the
> GPL has become the "slap-on" license that you can't "slap-off".
> It's used by GPL afficionados in an attempt to make all modifiable
> code GPL whether it actually is or not.  This also affects a
> number of Public Domain projects currently, but would disadvantage
> GPL copyright claims.  The Public Domain code can't be copyrighted
> under US law.  It either was copyright once and can't be again, or
> it is prohibited by law from being copyrighted.  GPL authors have
> added small snippets of code for which they've claimed GPL
> copyrights.  However, the work can only have a single copyright,
> the original.  In these cases, that "copyright" is Public Domain
> and as such can't be copyrighted or have copyright claims like GPL
> applied to it.  I.e., if such a case ever makes it to court, the
> GPL authors will forfit their claims to the Public Domain
> "copyright".
>
> > Which makes me wonder, should the BSD license /have/ such a
> > string, how would you distinguish it from the GPL proper?
>
> Honestly, I don't see the point you're attempting to make.  Are
> you equating or conflating the two because BSD license would
> have one additional restrictive clause?  Absurd.
>
> Rod Pemberton

I had two questions.

a. Does the author have to claim copyright? I vaguely remember angst
that (c) wasn't legally the same as the copyright symbol (U.S. venue),
thus impeding copyright claim.

b. Is the U.S. signatory to the Berne Convention?

(b.) according to Wikipedia:
The United States initially refused to become a party to the
Convention, since that would have required major changes in its
copyright law, particularly with regard to moral rights, removal of
the general requirement for registration of copyright works and
elimination of mandatory copyright notice. This led to the Universal
Copyright Convention in 1952 to accommodate the wishes of the United
States. But on March 1, 1989, the U.S. Berne Convention Implementation
Act of 1988 was enacted, and the U.S. Senate ratified the treaty,
making the United States a party to the Berne Convention,[11] and
making the Universal Copyright Convention nearly obsolete.[12]

So, by (b.) copyright (a.) falls automatically to the author. -- even
an anonymous author, -- for a term of 50 years.

Someone might argue the legality of public domain designation prior to
the copyright end of term?
Is the authors true recourse a very permissible copyright license??

c. Is there wording in the Berne Convention as to 'Public Domain'?
(the hidden question is whether an author has abdication rights to a
work covered by the Berne Convention.)

Steve

[Ivan Shmakov]

>>>>> s dubrovich@yahoo com <s_dub...@yahoo.com> writes:

[...]

> But on March 1, 1989, the U.S. Berne Convention Implementation Act of
> 1988 was enacted, and the U.S. Senate ratified the treaty, making the

> United States a party to the Berne Convention, and making the

> Universal Copyright Convention nearly obsolete.

> So, by (b.) copyright (a.) falls automatically to the author. -- even
> an anonymous author, -- for a term of 50 years.

Yes. (AFAIK. IANAL.)

> Someone might argue the legality of public domain designation prior
> to the copyright end of term? Is the authors true recourse a very
> permissible copyright license??

The first's debatable, yet the use of such a license will
certainly help a lot, should the would-be "user" of such work be
under a jurisdiction lacking the very notion of public domain.

> c. Is there wording in the Berne Convention as to 'Public Domain'?
> (the hidden question is whether an author has abdication rights to a
> work covered by the Berne Convention.)

As there /are/ jurisdictions which have no notion of public
domain, I'd assume that there's no such wording in the
convention itself.

PS. Check also the TRIPS agreement.

[wolfgang kern]

Ivan replied to Steve:

[...]

>> Someone might argue the legality of public domain designation prior
>> to the copyright end of term? Is the authors true recourse a very
>> permissible copyright license??

> The first's debatable, yet the use of such a license will
> certainly help a lot, should the would-be "user" of such work be
> under a jurisdiction lacking the very notion of public domain.

>> c. Is there wording in the Berne Convention as to 'Public Domain'?
>> (the hidden question is whether an author has abdication rights to a
>> work covered by the Berne Convention.)

> As there /are/ jurisdictions which have no notion of public
> domain, I'd assume that there's no such wording in the
> convention itself.

Sorry for I can't follow on this whole discussion,
whenever I posted a few lines of Asm-code here or in other NGs,
they were meant to help or at least meant to be point to a discussion
about programming/hardware-matters and I never expected nor asked for
payment nor copyright-comments on the lines I posted (this were always
free as in free beer and free to use as in No-Licence applied!)

Methink that publishing ideas and hints and tricks for a certain hardware
cannot imply fees of any kind. Except if you buy s book then pay for it!

Information need to be free, and no school can pass by!

Billy the Greedy may think about this different ...
and therefore he might keep the status number one as the most cursed
persons on this planet.

__
wolfgang

[Stanley Daniel de Liver]

On Sat, 20 Apr 2013 07:39:15 +0100, Herbert Kleebauer <kl...@unibwm.de>
wrote:

> On 20.04.2013 06:22, Tony wrote:
>
>>>> There is only one definition of 'free' that can apply to software,
>>>> because
>>>> software is inanimate, and that definition relates to price.
>>>
>>> The "free" in free software has nothing to do with money.
>>
>> I ignore perversions of words which attempt to propagate political
>> agenda and I
>> ignore those who do that.
>
> Yes, ignore reality and continue to live in your fantasy
> world as long as you can. Maybe you should read:
>
> http://www.gnu.org/philosophy/free-sw.html :
>
> Thus, “free software” is a matter of liberty, not price.
> To understand the concept, you should think of “free” as
> in “free speech,” not as in “free beer”.
>
>
I think he's here for the full half-hour argument.


--
It's a money /life balance.

[Rod Pemberton]

<s_dub...@yahoo.com> wrote in message
news:ebd8572e-0b52-4e0d...@j20g2000yqo.googlegroups.com...

[copyright]

> I had two questions.
>
> a. Does the author have to claim copyright? I vaguely remember
> angst that (c) wasn't legally the same as the copyright symbol
> (U.S. venue), thus impeding copyright claim.

...

> b. Is the U.S. signatory to the Berne Convention?
>
> (b.) according to Wikipedia:
> The United States initially refused to become a party to the
> Convention, since that would have required major changes in its
> copyright law, particularly with regard to moral rights, removal
> of the general requirement for registration of copyright works
> and elimination of mandatory copyright notice. This led to the
> Universal Copyright Convention in 1952 to accommodate the
> wishes of the United States. But on March 1, 1989, the U.S.
> Berne Convention Implementation Act of 1988 was enacted,
> and the U.S. Senate ratified the treaty, making the United
> States a party to the Berne Convention,[11] and
> making the Universal Copyright Convention nearly obsolete.[12]
>
> So, by (b.) copyright (a.) falls automatically to the author. --
> even an anonymous author, -- for a term of 50 years.
>
> Someone might argue the legality of public domain designation
> prior to the copyright end of term?
> Is the authors true recourse a very permissible copyright
> license??
>
> c. Is there wording in the Berne Convention as to 'Public
> Domain'?
> (the hidden question is whether an author has abdication rights
> to a work covered by the Berne Convention.)

People keep arguing that we're subject to B.C., automatically and
unconditionally, usually *vehemently* so by EU citizens. But,
that's not the way they apply the copyright law here. It's as if
the B.C. doesn't exist here. I.e., treaties don't seem to
override State and Federal law. My understanding, (which could be
incorrect) is that a court case has to go to special US court that
handles international laws for B.C. to apply. There is generally
no reason for a copyright case to go to an international court.
Normally, Federal laws apply, since people file their copyright
cases in Federal courts. This is so their copyrights can be
enforced across State boundaries. However, you can also apply in
State courts where State laws would apply. But, your rights are
limited to State law, and the results of your case is limited to
the territory of that State. That causes a problem for your court
case if the copyright infringement occured in another State than
yours. So, most people file in Federal courts. Perhaps, they
file their case in international court if infringement occurs
outside the US... (?)


Rod Pemberton

[Rod Pemberton]

"Ivan Shmakov" <onei...@gmail.com> wrote in message
news:8738umd...@violet.siamics.net...
> >>>>> Rod Pemberton <do_no...@notemailnotq.cpm> writes:

[...]

Sorry for the delay, my video card died, darn BGAs unsolder
themselves when they get hot.

I'm going to snip alot. I really didn't want to get into this
topic in the first place as noted in the a.o.d. thread. I've
discussed it much in the past. I just wanted to restate my
opinions to my a.o.d. "friends" and move on. I'm not sure where
you came from ...

> > I. e., the point, which you seemed to have missed, is that I

> > believe the GPL and LGPL shouldn't be allowed to be
> > merged (linking is fine) with substantially "free-er"
> > licensed code like MIT and BSD.
>

> I do not understand. Do you ask for a license which prohibits
> merging of the code covered with a "less-free" GPL-licensed
> code, yet /allows/ the very same code to be merged with the code
> released under "less-free-still" proprietary licenses?
>

No.

Basically, I want a single license, similar to BSD or MIT, that
would require any changes, modifications, additions, etc by
licensees to comply with the original copyright and license terms,
i.e., prohibit changes, modifications, additions, etc by under
other licenses.

The problem is "slap-on" GPL licenses which cause license
pollution. E.g., say the original code is BSD or MIT, then along
comes "Mr. FSF" who makes some trivial changes. He doesn't want
to release his code additions under the original BSD or MIT
license terms. So, he slaps on GPL license. Now, the code has
two licenses on the same file(s).

As for FSF requiring contribution of copyright, that's probably
for separate files, but if not, then they're probably just
covering their bases.

> > The problem of merging code with different licenses is
> > license pollution. When multiple authors work on the
> > same program, a single piece of code can have a variety
> > of authors claiming copyrights under a variety of license
> > in a single source code file. Unfortunately, this isn't legal
> > under US law. US law only allows a work to be

> > copyrighted under a single copyright. I. e., mixing of code

> > under different license within a single file isn't legal. If
> > the> code for each is in different files, it's not an issue.
>

> That seem to suggest that US law equates "file" and "work,"
> which I doubt is the case.
>
> For instance, a statically-linked executable will contain all
> the "creative components" right within a single file. Do you
> suggest that such executables aren't "legal" under US law?
>

a) No. (see below)
b) You're conflating two things: source code (protected by
copyright) and executable (not ...).

> >> Which makes me wonder, should the BSD license /have/ such a
> >> string, how would you distinguish it from the GPL proper?
>
> > Honestly, I don't see the point you're attempting to make.
> > Are you equating or conflating the two because BSD license
> > would have one additional restrictive clause?
>

> Indeed I do. How do you letter such a clause? Will it be like
> "it's explicitly forbidden to merge this code with that under
> GNU GPL"? Then, there also is "GNU AGPL," etc., and new ones
> may spring at any time. Or will it be "provided that you don't
> merge the code with that under a restrictive license"?
...

> Fine,
> you've made it impossible to incorporate the code into non-free
> software just as well,

Untrue. You're equating "merging" and "linking".

(I'd swear in one of the numerous past conversations I've had on
copyrights, all these exact issues, expressed nearly identically
too, came up previously with someone else ...)

For merging, any changes, additions, or modifications to licensed
code in a single file are in all in the same file and IMO should
have a single license. For linking, each file is separate, each
can have it's own license. E.g., a source file from BSD or MIT
when compiled can be linked with a source file from GPL when
compiled which is linked with a source file from a proprietary
license when compiled. The issue I was complaining about is when
the BSD or MIT, GPL, and proprietary licensed code all end up in
the same source file. Basically, if someone wants to make GPL
changes to BSD or MIT code, they'd have a few choices 1) put the
GPL code in it's own file 2) put the GPL code in with other GPL
code 3) decide to "rebrand" the code as BSD or MIT and put it in
with the BSD or MIT code it modifies. That way, the BSD or MIT
code can be separated from the GPL code and commercial code,
easily. If one file or multiple files has code under multiple
licenses, it can't be separated out by license. I call the
inability to *not* separate code by license because it's has
multiple licenses: license pollution. E.g., say I'd like to build
a project purely using BSD code. I've found a program that I need
code from. One of the source code files has the critical code.
Originally, it was released as BSD, but someone has made GPL
changes. However, the original project can't be located anymore.
How do I re-use the code I need and guarantee the code is purely
BSD? That's the issue. If the BSD code and GPL code aren't in
separate files, then it can't be done. That's why there is a need
for a license that prohibits merging unless it falls under the
original terms. Did that clarify the issue finally?


Rod Pemberton

[kerravon]

On Apr 20, 11:39 pm, kerravon <kerra...@w3.to> wrote:

> On Apr 20, 11:02 pm, "Alexei A. Frounze" <alexfrun...@gmail.com>
> wrote:
>
> > > I haven't thought about that. What's a reasonable price in your
> > > opinion?
>
> $20?

Hi Alexei.

You haven't replied to this.

Would you at least be able to make the bpb etc variables
public domain so that I don't need to derive fresh names?

Thanks. Paul.

[Ivan Shmakov]

>>>>> Rod Pemberton <do_no...@notemailnotq.cpm> writes:
>>>>> "Ivan Shmakov" <onei...@gmail.com> wrote...

> Sorry for the delay,

There's hardly any need to apologize: you weren't under an
obligation to respond to begin with.

Indeed, I seem to understand the problem you're having. What I
don't understand, however, is how GPL (and not, say, particular
GPL proponents or users) makes this issue any worse (than any
other non-BSD license in existence), /and/ how a specific
license (unless a copyleft one) could help one to resolve it.

[...]

> The problem is "slap-on" GPL licenses which cause license pollution.

> E. g., say the original code is BSD or MIT, then along comes

> "Mr. FSF" who makes some trivial changes. He doesn't want to release
> his code additions under the original BSD or MIT license terms. So,
> he slaps on GPL license. Now, the code has two licenses on the same
> file(s).

How is it going to be different from, say, "Mr. Proprietor," who
makes trivial changes and slaps on a "free for non-commercial
use only" (or any other, for that matter) license?

[...]

> b) You're conflating two things: source code (protected by copyright)
> and executable (not ...).

I'm not a copyright law expert, but I find it hard to believe
that an executable (or any other form of binary code) somehow
escapes copyright.

First of all, the binary code results from a "trivial" (as in:
non-creative) transformation of the source, and thus should be
covered by the very same copyright as the source. Also, would
there be no copyright on binaries, how'd it be possible for the
proprietary software vendors to demand a per-copy payment for
binary software distributions (which is quite a commonplace)?

[...]

>> Fine, you've made it impossible to incorporate the code into
>> non-free software just as well,

> Untrue. You're equating "merging" and "linking".

I didn't say that the non-free software developers will
necessarily /link/ against a BSD-licensed file. On the
contrary, I see nothing in the BSD license to prohibit them from
"merging" their own non-free changes into the very same file
instead, and distributing the result under a non-free license as
a whole. (And please note that there /are/ non-free software
packages, which are distributed as source code bundles.)

Two more things to mention:

* "linking" as a term may become inapplicable when it comes to
libraries for /interpreted/ languages, and similar cases;

* occasionally, it may make sense to ship the most part (or the
whole) of the package's source concatenated into a single file
(check, e. g., SQLite.)

[...]

> Basically, if someone wants to make GPL changes to BSD or MIT code,
> they'd have a few choices 1) put the GPL code in it's own file 2) put
> the GPL code in with other GPL code 3) decide to "rebrand" the code
> as BSD or MIT and put it in with the BSD or MIT code it modifies.
> That way, the BSD or MIT code can be separated from the GPL code and
> commercial code, easily. If one file or multiple files has code
> under multiple licenses, it can't be separated out by license. I
> call the inability to *not* separate code by license because it's has

> multiple licenses: license pollution. E. g., say I'd like to build a

> project purely using BSD code. I've found a program that I need code
> from. One of the source code files has the critical code.
> Originally, it was released as BSD, but someone has made GPL changes.
> However, the original project can't be located anymore. How do I
> re-use the code I need and guarantee the code is purely BSD? That's
> the issue. If the BSD code and GPL code aren't in separate files,
> then it can't be done.

I seem to understand the inconvenience. However, to reiterate
it once more, GPL is /no different/ in that respect to any other
license whatsoever, and there's little sense in singling it out.

> That's why there is a need for a license that prohibits merging
> unless it falls under the original terms. Did that clarify the issue
> finally?

Yes.

[Tony]

In article <kl6j16$mt4$1...@newsreader2.utanet.at>, now...@never.at says...

> Information need to be free, and no school can pass by!

That's a lame attempt at trying to justify property taxes to pay for your kids.
Dude, you use your dick, YOU must be responsible for the result. How much do you
owe?

[Ivan Shmakov]

>>>>> Stanley Daniel de Liver <ad...@127.0.0.1> writes:

[...]

>> http://www.gnu.org/philosophy/free-sw.html :

>> Thus, “free software” is a matter of liberty, not price. To
>> understand the concept, you should think of “free” as in “free
>> speech,” not as in “free beer”.

> I think he's here for the full half-hour argument.

I guess he just exercises his right of free speech. Which is,
as we all now understand, the right to offer one's opinions and
ideas at no price.

--
FSF associate member #7257 Alas, free advice is worth what you pay for it.

[Rod Pemberton]

"kerravon" <kerr...@w3.to> wrote in message
news:77a2dfb5-f856-4bef...@k6g2000pbq.googlegroups.com...

> Would you at least be able to make the bpb etc variables
> public domain so that I don't need to derive fresh names?

Can't you just use the names in RBIL?


RP

[kerravon]

On Apr 27, 9:19 am, "Rod Pemberton" <do_not_h...@notemailnotq.cpm>
wrote:
> "kerravon" <kerra...@w3.to> wrote in message

Where can I find that in here?

http://www.ctyme.com/rbrown.htm

(I can't find bpb)

Thanks. Paul.

[James Harris]

Paul, I've not been closely following this thread so the following
answer may not be what you are looking for but if you want the
official names of fields in the various FAT-format boot sectors
including the BPB check out Microsoft document fatgen103.doc. A quick
Google for it brings up a PDF form:

http://staff.washington.edu/dittrich/misc/fatgen103.pdf

As you will see there are variants depending on when the boot sector
was created.

If it helps there is an example of assembly use in the source code at

http://codewiki.wikispaces.com/fatbootsect.nasm

Do a search on that page for the first field, BS_OEMName.

James

[Rod Pemberton]

"kerravon" <kerr...@w3.to> wrote in message

news:f1d9210d-3a08-4710...@ph9g2000pbb.googlegroups.com...

> On Apr 27, 9:19 am, "Rod Pemberton"
<do_not_h...@notemailnotq.cpm>
> wrote:
> > "kerravon" <kerra...@w3.to> wrote in message
> >
news:77a2dfb5-f856-4bef...@k6g2000pbq.googlegroups.com...
...

> > > Would you at least be able to make the bpb etc variables
> > > public domain so that I don't need to derive fresh names?
> >
> > Can't you just use the names in RBIL?
> >
>

> Where can I find that in here?
>
> http://www.ctyme.com/rbrown.htm
>
> (I can't find bpb)
>

Well, they're right here on the CTYME website:
http://www.ctyme.com/intr/rb-2985.htm

(Okay, okay, so I searched the original text files for them... ;-)

If you have the original RBIL zip'd text files, the BPB is Table
01663 and the EBPB is Table 01664. Both are in INTERRUP.G file
within inter61b.zip under Int 21h, AH=53h.

They're also available from here on the DJGPP website:
http://www.delorie.com/djgpp/doc/rbinter/id/97/29.html


FYI, basically, there are three places online to get RBIL. One is
the original file set of 6 main zip's and some other files. The
other two are online html-ized versions provided by CTYME and
DJGPP. The online html versions don't include some of the other
non-interrupt related text files like those for the CMOS or PORTS
or MEMORY locations.

The original zip'd text files are here:
http://www.cs.cmu.edu/afs/cs/user/ralf/pub/WWW/files.html


James mentioned and provided a linked for the MS fatgen103.doc.
There is also an international standard for FAT12/16. It's known
as either ISO/IEC 9292 or ECMA-107. The ECMA version is here:

http://www.ecma-international.org/publications/standards/Ecma-107.htm

Many other standards for media and disks are available as ECMA
standards too.

http://www.ecma-international.org/publications/standards/Stnindex.htm

HTH,


Rod Pemberton

[kerravon]

Thanks James and Rod.

Someone popped out of the ether who was willing to rewrite
the boot sector. Here is the current state:

http://pdos.cvs.sourceforge.net/viewvc/pdos/pdos/src/pbootsec.asm?view=markup

The code doesn't work, and I'm trying to desk-check it to
see if I can spot the problem.

If anyone here can spot the issues, please let me know.

Also note that he changed it to org 07c00h, but that caused
a linker error so I put it back to 0100h.

Thanks. Paul.

[kerravon]

On Apr 28, 6:50 pm, kerravon <kerra...@w3.to> wrote:
>
> http://pdos.cvs.sourceforge.net/viewvc/pdos/pdos/src/pbootsec.asm?vie...

>
> The code doesn't work, and I'm trying to desk-check it to
> see if I can spot the problem.

I've found some problems (and updated the link), but it's
still not working. I think the major problem is that ds and
cs should be set to 7b0h (and ip should be 0100+), even
if it is called with cs:ip of 0000:07c00.

I will see if I can negotiate that code change.

BFN. Paul.

[s_dub...@yahoo.com]

On Apr 28, 3:50 am, kerravon <kerra...@w3.to> wrote:
> Thanks James and Rod.
>
> Someone popped out of the ether who was willing to rewrite
> the boot sector. Here is the current state:
>

> http://pdos.cvs.sourceforge.net/viewvc/pdos/pdos/src/pbootsec.asm?vie...

>
> The code doesn't work, and I'm trying to desk-check it to
> see if I can spot the problem.
>
> If anyone here can spot the issues, please let me know.
>
> Also note that he changed it to org 07c00h, but that caused
> a linker error so I put it back to 0100h.
>
> Thanks.  Paul.

Can you link it as a .SYS file? -AIR that is equivalent to a flat
binary image?

-this is an error:

"
ReadSingleSector proc
push ax
push bx
push cx
push dx
push es
call Lba2Chs ;Grab our CHS
RetryRead:
call ResetDrive ;Get drive ready..
mov dl, [BootDisk] ;Grab our boot disk
mov ax, 0x0201 ;Read function, one sector
int 13h
jc RetryRead
push es
push dx
push cx
push bx
push ax
ret
ReadSingleSector endp
"

The pushes before the RET need to be POP.

The destination is ES:BX. For me, I'd set ES to 0070h (that appears
to be where you want this to go) and BX := 0000h, but on each read
loop in ReadSingleSector, BX needs to be incremented by sector-size,
also the source CHS needs to increment, otherwise, aren't you reading
the same sector over?

hth.

Here's some snippets from one of mine, but this is not fat, but 4k
block based, still it shows: the BPB fields, and the method of using
variables for C,H,S for ReadSectors whose values are set during the
conversion routine; LSN_CHS.

(NASM Syntax, jmp's are long unless specified SHORT JMP)

;;Rombios loads this to 0000:7C00, DL==boot drv.

[SECTION .cseg vstart=0]

BS_jmpBoot:
jmp START

;;-----------------------------------------------------------------------75
;; win95 overwrites OEM_ID, don't depend on it.

BS_OEMName db "HOS boot" ;; OEM_ID
BPB_BytesPerSec dw 0200h
BPB_SecPerClus db 01h
BPB_RsvdSecCnt dw 0012h ;; not normal
BPB_NumFATs db 00h ;; normally 2
BPB_RootEntCnt dw 00E0h ;; directory entries
BPB_TotSec16 dw 0B40h ;;720k=05A0h
BPB_Media db 0F0h
BPB_FatSz16 dw 0009h ;; typical 1.44mb floppy
BPB_SecPerTrk dw 0012h ;; ditto
BPB_NumHeads dw 0002h
BPB_HiddSec dd 00000000h ;preceding vol
BPB_TotSec32 dd 00000000h
BS_DrvNum db 00h
BS_Reserved1 db 00h
BS_BootSig db 29h ;;indicates the next
;; three fields are valid, newer fat spec
BS_VolID dd 0FFFFFFFFh
BS_VolLab db "BOOTS HOS "
BS_FilSysType db " "

;; added local string, not a part of the specification.
BS_Img_Vers db "v.0r0, Single Stage Boot"

;; boot sector + (sectors per blk - 1)
BLK_RSV dw 1 ;; blocks before directory file.
DIR_SZ dw 11 ;; directory file, size in 4k blocks.
VOL_SZ dw 360 ;; 2880 sectors / 8 = 360 blocks.

;;-----------------------------------------------------------------------75
;; code located at 0000:7C00. Chg data segment:offset to 07C0:0000h.
;;

START:
cli
mov ax, 07C0h
mov DS, ax
mov ES, ax
; mov fs,ax
; mov gs,ax

;; create stack at SS:0000h SP:7C00h

mov ax, 0000h
mov SS, ax
mov sp, 7C00h
sti

;; put bootstrap message to console

mov si, msgBoot
call DisplayMessage

;;-----------------------------------------------------------------------75
;; One sector is read into 07C0:0000h, now load remainder (7) of
block.

LOAD_BBLK: ;; read into memory (7C00:0200)

mov ax, 1 ;; LSN to start with..
mov cx, 7 ;; remaining sectors per 4k block.
mov bx, 0200h ;; copy above boot strap sector.
call ReadSectors ;; if success, remainder of block follows.
. . .



;;-----------------------------------------------------------------------75
;; ReadSectors
;; reads CX sectors from disk starting at AX`LSN into memory location
ES:BX
;;-----------------------------------------------------------------------75

ReadSectors:
.MAIN:
mov di, 0005h ;; five retries for error
.SECTORLOOP:
push ax
push bx
push cx
call LSN_CHS
mov ah, 02h ;; BIOS read sector
mov al, 01h ;; read one sector
mov ch, BYTE [absoluteTrack] ;; track
mov cl, BYTE [absoluteSector] ;; sector
mov dh, BYTE [absoluteHead] ;; head
mov dl, BYTE [BS_DrvNum] ;; drive
int 13h ;; invoke BIOS
jnc .SUCCESS ;; test for read error

xor ax, ax ;; BIOS reset disk
int 13h ;; invoke BIOS
dec di ;; decrement error counter
pop cx
pop bx
pop ax
jnz .SECTORLOOP ;; attempt to read again
int 18h

.SUCCESS:
mov si, msgProgress ;; Conout a period for each read.
call DisplayMessage
pop cx
pop bx
pop ax
add bx, WORD [BPB_BytesPerSec] ;; queue next buffer
inc ax ;; queue next sector
loop .MAIN ;; read next sector
RET

;;-----------------------------------------------------------------------75
;; LSN_CHS
;; Convert Logical Sector Number to Cylinder, Head, Sector.
;; convert AX == LSN addressing scheme to CHS addressing scheme.
;; absolute sector =
;; (logical sector / sectors per track) + 1
;; absolute head =
;; (logical sector / sectors per track) MOD numb of heads
;; absolute track =
;; logical sector / (sectors per track * number of heads)
;;-----------------------------------------------------------------------75
;; div DX:AX by word -> AX=Quotient, DX=Remainder

LSN_CHS:
xor dx, dx ;; prepare dx:ax for operation
div WORD [BPB_SecPerTrk] ;; AX=Quotient, DX=Remainder
inc dl ;; adjust for sector 0
mov BYTE [absoluteSector], dl
xor dx, dx ;; prepare dx:ax for operation
div WORD [BPB_NumHeads] ;; AX=Quotient, DX=Remainder
mov BYTE [absoluteHead], dl
mov BYTE [absoluteTrack], al
RET

;;-----------------------------------------------------------------------75
;; D A T A
;;-----------------------------------------------------------------------75

absoluteSector db 00h ;; absolute, final, values to pass to INT 13h.
absoluteHead db 00h
absoluteTrack db 00h

. . .

TIMES 510-($-$$) DB 0
DW 0AA55h

-- this is for floppy boot only. see the use of BPB_BytesPerSec as a
runtime variable gotten from the BPB static field, for example.

-- I'm not familiar with pdos fat compatibility, but fat cluster size
has had various sizes other than a sector size, has modern fat settled
on sector-size == cluster size?

-Steve

[Rod Pemberton]

"kerravon" <kerr...@w3.to> wrote in message

news:97ffcbef-698c-47cc...@g5g2000pbp.googlegroups.com...

> Also note that he changed it to org 07c00h, but that caused
> a linker error so I put it back to 0100h.
>

Linker ... ?

In general, an assembler should be able to produce a flat binary
without any linking... There should be an option to produce a
.com, or as Steve noted, a .sys file. My guess is the linker is
part of the problem, i.e., you're fighting your tools.

Yes, if you're insisting on keeping linking, I think you can work
around it by adjusting cs, ip, etc as you're attempting to do now.


Rod Pemberton

[Rod Pemberton]

"kerravon" <kerr...@w3.to> wrote in message

news:fcc889ac-ce6c-4923...@pl9g2000pbb.googlegroups.com...

> On Apr 28, 6:50 pm, kerravon <kerra...@w3.to> wrote:

> > The code doesn't work, and I'm trying to desk-check it to
> > see if I can spot the problem.
>
> I've found some problems (and updated the link), but it's
> still not working. I think the major problem is that ds and
> cs should be set to 7b0h (and ip should be 0100+), even
> if it is called with cs:ip of 0000:07c00.

Did you want CS=0, IP=7C00h, and DS=07B0h ?

AIUI, CS and IP are used for code only, unless you're using a CS
segment override somewhere. So, they just need to point to
address 7C00h. CS<<4+IP. I.e., I don't think they need to be
adjusted.

However, AIUI, DS and the offset encoded in the instruction are
used for the address of data.

AIUI, offsets in the encoded instructions which reference data
with an ORG of 100h will have 100h added to them, ORG of 7C00h
will have 7C00h added to them, ORG of 0h will have nothing added.

You can check the offsets encoded into various instructions by
assembling with different ORG's then disassembling - look at the
encoded bytes not the address.

E.g., compare ORG 0h with ORG 7C00h. The encoded offsets should
have 7C00h added to them, and stored little-endian. FYI,
branching instructions use relative offsets, so they'll look a
little different from data the instructions.

So, I think offsets for data references in the code are:

DS + instruction encoded offset

Where 'instruction encoded offset' is:

ORG + byte offset from start of code

Resulting with:

DS + ORG + byte offset from start of code

So, that gives us DS<<4+ORG=7C00h, with an ORG of 100h, then
DS=07B0h.

Sorry, I'm definately not 100% sure that's correct either ...
I mostly do 32-bit.


Rod Pemberton

[Rod Pemberton]

"Rod Pemberton" <do_no...@notemailnotq.cpm> wrote in message
news:klmokm$lka$1...@speranza.aioe.org...

Clarification, that's when the code is loaded to address 7C00h.
If you're loading the code as a .com too, then the load address
will be something else. I think CS must equal DS for that.

He doesn't have CS or DS, but here is Ben Lunt's list of DOS entry
values:
http://www.fysnet.net/yourhelp.htm

[Mike Gonta]

"Rod Pemberton" wrote:

> In general, an assembler should be able to produce a flat binary
> without any linking...

That depends on the assembler. In NASM and FASM the "org" directive
allows the assembler to "fixup" a flat binary. This is half the job
of the linker. In MASM and GAS the "org" directive skips ahead,
filling in the gap with zero's.

> Yes, if you're insisting on keeping linking, I think you can work
> around it by adjusting cs, ip, etc as you're attempting to do now.

Typically, runtime adjustments have to be made with MASM code that
isn't linked. Code that is linked has an implied "org" of zero.

With NASM and FASM there is no need to use a linker to create a
boot sector. It's when you want to combine assembly (not inline
assembly) with C in the same executable that a linker is required.

It's easy to write a real mode boot sector in GAS (with Intel syntax).
On Windows you need a GCC cross compiler with ELF output (which can
be downloaded - you don't have to make one).
Assemble with AS -- as boot.s -o boot.o

.intel_syntax noprefix
.text
.code16
// your code and data here.
.org 0x1FE
.word 0xAA55

Link with LD -- ld -T boot.ld -o boot.bin
The ld script (boot.ld):

OUTPUT_FORMAT ("binary")
SECTIONS
{
. = 0x7C00;
.text :
{
boot.o (.text);

[kerravon]

On Apr 30, 1:36 am, "s_dubrov...@yahoo.com" <s_dubrov...@yahoo.com>
wrote:

> On Apr 28, 3:50 am, kerravon <kerra...@w3.to> wrote:
> The pushes before the RET need to be POP.

Thankyou.

> The destination is ES:BX.

And thanks a lot for pointing at the ES not being set
correctly.

Good news is that the bootloader works now, although
it still has 1 hack in it which needs to be removed.

BFN. Paul.

[Rod Pemberton]

"Mike Gonta" <mike...@gmail.com> wrote in message
news:klmrr9$uds$1...@speranza.aioe.org...

> "Rod Pemberton" wrote:

> > In general, an assembler should be able to produce a flat
> > binary without any linking...
>
> That depends on the assembler. In NASM and FASM the "org"
> directive allows the assembler to "fixup" a flat binary. This is
> half the job of the linker. In MASM and GAS the "org" directive
> skips ahead, filling in the gap with zero's.
>

...

> > Yes, if you're insisting on keeping linking, I think you can
> > work around it by adjusting cs, ip, etc as you're attempting
> > to do now.
>
> Typically, runtime adjustments have to be made with MASM code
> that isn't linked. Code that is linked has an implied "org" of
> zero.
>
> With NASM and FASM there is no need to use a linker to create a
> boot sector. It's when you want to combine assembly (not inline
> assembly) with C in the same executable that a linker is
> required.
>
> It's easy to write a real mode boot sector in GAS (with Intel
> syntax).

And, "It [should be just as] easy ... in GAS"
(without Intex syntax) too. ;-)

GAS syntax is fine. It's just different.

> On Windows you need a GCC cross compiler with ELF output (which
> can be downloaded - you don't have to make one).

You don't need a "GCC cross compiler".
You don't need "ELF output" either.

DJGPP uses GCC and GAS but not GLIBC. DJGPP can be used under
Windows in a DOS console through XP or so (no cross compiler).
DJGPP can produce flat binaries. DJGPP produces COFF object files
(not ELF).

> Assemble with AS -- as boot.s -o boot.o
>

...

> .intel_syntax noprefix
> .text
> .code16
> // your code and data here.
> .org 0x1FE
> .word 0xAA55
>

Well, I've not seen .org used like that, so far. I'd have to
check the GAS manual to find out if that's valid usage.
Maybe, tomorrow...

I generally see the GAS .fill directive used there instead, e.g.,
something like:

.fill 0x1fe - (. - start) , 1, 0

> Link with LD -- ld -T boot.ld -o boot.bin

Perhaps, you want this:

ld --oformat binary --Ttext 0x7C00 -e start -o boot.bin boot.o

You should be able to produce a flat form binary without a linker
script using --oformat binary. You should be able to specify the
entry lable with -e. You should also be able to specify the
starting address using the -Ttext option.

I think the -Ttext option is the actual "org" address here. I'm
not sure if -Ttext address is or isn't or can do the same thing as
the GAS .org assembler directive. I'd have to review the GAS
manual on .org...

> The ld script (boot.ld):
>
> OUTPUT_FORMAT ("binary")
> SECTIONS
> {
> . = 0x7C00;
> .text :
> {
> boot.o (.text);
> }
> }
>

I think this is unneeded.

HTH,


Rod Pemberton

[Mike Gonta]

"Rod Pemberton" wrote:

> "Mike Gonta" wrote:
>> On Windows you need a GCC cross compiler with ELF output (which
>> can be downloaded - you don't have to make one).
>
> You don't need a "GCC cross compiler".
> You don't need "ELF output" either.

"ELF output" is short for ELF object file format, LD generates the
binary from the object file, there is no ELF executable.

>
> DJGPP uses GCC and GAS but not GLIBC. DJGPP can be used under
> Windows in a DOS console through XP or so (no cross compiler).
> DJGPP can produce flat binaries. DJGPP produces COFF object files
> (not ELF).

DJGPP produces a DOS executable output. If it could be run on
Windows without a virtual machine/emulator like MinGW and Cygwin
it would be by definition a "GCC cross compiler" since the output
is not a native Windows PE.

LD cannot produce a binary from a native Windows GCC (with PE coff
object file format).

>> Link with LD -- ld -T boot.ld -o boot.bin
>
> Perhaps, you want this:
>
> ld --oformat binary --Ttext 0x7C00 -e start -o boot.bin boot.o

Yes, for a simple boot sector the command line version is much simpler.
The -e (program entry) is redundant for a binary format and isn't
required. LD will warm that there is no default "_start" but still
output correctly. Putting this at the beginning of the code will keep
it quiet.

.global _start
_start:

> I think the -Ttext option is the actual "org" address here.

"org" only in the sense of FASM or NASM. ".org" in GAS is the same as
"org" in MASM" - the code is expanded from the current location by the
value specified. Putting this at the start,

.org 0x7C00

will result in an undesired code size of at least 0x7C00 bytes (plus the
actual code).

[Alexei A. Frounze]

On May 1, 4:55 pm, "Mike Gonta" <mikego...@gmail.com> wrote:
> "Rod Pemberton" wrote:
> > "Mike Gonta" wrote:
> >> On Windows you need a GCC cross compiler with ELF output (which
> >> can be downloaded - you don't have to make one).
>
> > You don't need a "GCC cross compiler".
> > You don't need "ELF output" either.
>
> "ELF output" is short for ELF object file format, LD generates the
> binary from the object file, there is no ELF executable.
>
>
>
> > DJGPP uses GCC and GAS but not GLIBC.  DJGPP can be used under
> > Windows in a DOS console through XP or so (no cross compiler).
> > DJGPP can produce flat binaries.  DJGPP produces COFF object files
> > (not ELF).
>
> DJGPP produces a DOS executable output. If it could be run on
> Windows without a virtual machine/emulator like MinGW and Cygwin
> it would be by definition a "GCC cross compiler" since the output
> is not a native Windows PE.

By default, DJGPP produces a 16-bit DOS EXE with a COFF binary glued
to it.
That 16-bit part checks if there's a DPMI host, which any 32-bit
Windows to date is. If there isn't, it either loads one (e.g. CWSDPMI)
or has one inside and "installs" it. Then (perhaps, after an
additional relocation step), the DPMI host can run the 32-bit code
contained in the COFF part. In 32-bit Windows there's a virtual
machine for DOS and DPMI (=Dos Protected Mode Interface, AFAIR) apps.
I don't know if you want to count this virtual machine as a virtual
machine since it's built-in into Windows, but it's there.

> LD cannot produce a binary from a native Windows GCC (with PE coff
> object file format).

I'm not sure I understand the sentence.

Off-hand I do not remember if DJGGP's LD can produce flat binaries.
Perhaps it doesn't. That would explain why I wrote a linker/relocator
and a loader for DGJPP's COFF binaries. I'd need to check it to be
sure.

Alex

[Mike Gonta]

"Alexei A. Frounze" wrote:

> "Mike Gonta" wrote:

>> LD cannot produce a binary from a native Windows GCC (with PE coff
>> object file format).

> I'm not sure I understand the sentence.

This is the clasic "ld: cannot perform pe operations on non pe output file"
You cannot use the MinGW or Cygwin versions with PE target to produce
flat binaries with LD. You need to use a "cross compiler" with ELF target.
This error is probably a result of other things as well, but there seems
to be a lot directly related to creating a flat binary. There are objdump
methods to extract the .text and even .text and .data. However, C will
put the data at the beginning and objdump will not rearrange the sections
and thus the flat binary will not work.

[Mike Gonta]

"Rod Pemberton" wrote:
> "Mike Gonta" wrote:

>> .intel_syntax noprefix
>> .text
>> .code16
>> // your code and data here.
>> .org 0x1FE
>> .word 0xAA55
>>
>
> Well, I've not seen .org used like that, so far. I'd have to
> check the GAS manual to find out if that's valid usage.
> Maybe, tomorrow...
>
> I generally see the GAS .fill directive used there instead, e.g.,
> something like:
>
> .fill 0x1fe - (. - start) , 1, 0

.org, .fill and .set are variations of .space and all do the same
thing - advance the location counter and fill the intervening
bytes.