anyone familiar with the elf format?

to some degree

I am working on a pet-kernel. In my system, there are two kinds of pages (data and capability pages). They are handled mostly similiar. The programmer can specify that certain variables should be stored in capability rather than data pages. I am wondering how to express that fact in the binary.

different elf sections probably?

yes, I am consiering that. It sounds a bit inflexible to me, and I wondered if there was a better solution.

in what sense?
after all there is already e.g. .data and .rodata

lindi-: well, that would imply a convention on section names, which is not what they were made for

adding .capability shouldn't be a huge problem?
hmm

lindi-: but normal loading needn't consider section names at all, they're for humans
anyway, I'll just do it

don't trust my advice

Hello. http://cpp.sourceforge.net/?show=39566 how to do something like this: when I invoke make release I'd like to get CPPFLAGS = $(RELEASEFLAGS), and when I use make debug, I'd like to have CPPFLAGS=$(DEBUGFLAGS) For now I have to manually change the makefile itself whenever I switch build types

in my linkerscript I have .capability ALIGN (4096): {*(.capability)}, but readelf reports the address of the section as 0×408020?

hello everyone!
can I ask something about the gpl v3?
anyone here?

sure
just ask, someone might respond

X releases a program under GPL3. Y gets a copy. X changes the license of his program and stop distributing the sources. how can Y propagate the program now?

He can still propagate as per GPLv3

You mean if Y doesn't have a copy of the source?

he has one

X, by giving Y a copy under the GPLv3 is bound by the terms

Then what's stopping him?

Each time you convey a covered work, the recipient automatically receives a license from the original licensors, to run, modify and propagate that work, subject to this License. You are not responsible for enforcing compliance by third parties with this License.

X isn't bound by the GPLv3 if he wrote it

thus, he has given Y the right to propagate

I dont' see the problem. If Y has a copy, released under the GPL, then Y can use all his GPL-granted rights

License changes aren't "recursive"

I see no license change here

the excerpt i wrote previously says that the recipient automatically recieveis a license from the original licensors
but the original is X.

Yes. So?

which now doesn't give any more copies of his program under the GPL

When X distributed the code to Y, Y got the license
True. So the new stuff isn't GPLed, so Y can't distribute that. He can distribute, modify the old stuff, that X released under the GPL

Y got the license but who says that X want's to continue giving it to downstream recipients?

X did, when he gave Y the copy under the GPL
If by 'giving it' you mean giving everyone a license to use his original software

The problem with Eben Moglen's interpretation comes from the unique construction of the GPL. The GPL purports to grant a new license from the original grantor, rather than each succeeding author of a program granting a sub-license to the next. By the very terms of the statute, however, licenses from the original licensor are only protected if the license was granted before executing the transfer. (New licenses from the original licensor are not allowed,
because the original licensor no longer holds the copyright to the work.) Therefore, further distribution of a GPL program is significantly limited after GPL is revoked. To be precise, the current holders of the program are authorized to modify and distribute the software, if they hold a "written instrument signed by the [former] owner," but the recipients of the software may not further redistribute it.
sorry for the big post
that's from wikipedia article on public domain

how do i use cat ?

man cat

cat file

that's bs

GPL revoked? That sounds like shit

bs?

i want to pipe the output of several things to 1 file, can i use cat for that ?

The GPL expressly states that one cannot revoke it, if I remember.

zap0, i think that cp does this.

huh?

Jafet, i don't think so.

I don't think the GPL states that, but it looks like a clear case of 'promissory estoppel'

all-in-one?

Jafet, the copyright holders don't have to agree with the terms of gpl

can anyone tell me why file(1) reports linux 2.6.6 when it's not 2.6.6 ?

what is it?
it's bullshit
you read it on wikipedia

The copyright holders are free to do whatever they please. Their actions are not covered under the scope of copyright law.

Aim_Here, what is promissory estoppel?

2.6.20

i don't exactly understand what does "automatically" mean in "…the recipient automatically receives a license from the original licensors…"

dictionary

can the original author deny this automatic licensing?
lc2, i know the meaning of the word.

i don't exactly understand what does "automatically" mean in "…the recipient automatically receives a license from the original licensors…"

you had me fooled

Promissory Estoppel is where if someone says 'I won't sue you if you do X', and you do X, then he won't successfully sue you.

lc2, i don't understand it's effect in this context.

the original author cannot deny that automatic licensing
by the very fact that they've released it under the GPL, they're not denying it
"can the original author sue everyone he gives the program to for copyright infringement?"
it's just as silly a question
that's what a license does
you can't simultaneously give permission to do something and not give permission to do something

so wiki is really bs.

welcome to the real world.

go fix it you wikipedian

we don't usually free minds once they reach a certain age

well it really needs fixing.

the mind has problems letting go of such notions as "wikipedia is a reliable source"
delete it all, it's original research

oops
my mistake

wikipedia indeed should not be trusted blindly

i like wikipedia

x makes uses gpl for a program. y gets a copy. then x sells his copyright to Z. Z is not the original author so he doesn't have to automatically license the program under GPL. and X who is the original author CANNOT license the program anymore to anyone!

i like it too

so what happens now?
(if y want to propagate the program further)

x can get a copy from y

I imagine that the estoppel and license arguments still work

"Each time you convey a covered work, the recipient automatically receives a license from the original licensors"

and make changes

In the SCO case, for instance, IBM is using both arguments, alleging they got such licenses from SCO's predecessors

by releasing their program under the GPL, the original licensors (i.e. the copyright holder) *do* automatically license their program to all third parties

They point to things AT&T said, which was two changes of ownership ago

and btw

Three changes

i'm sure you already know that in common law countries, it is a legal tradition to interpret contracts (and licenses) according to their meaning and the values/intent of the communities that spawned

I deleted all that stuff in the wiki host article for a laugh

so really, picking on fine points of wording is a very willy thing to do

I bet a bot will revert it and someone will complain at me

i hope you left a good edit summary
if you do that, then you don't get reverted

Delete a huge pile of mostly unsourced and mostly irrelevant and partly POV original research

good one
silly* btw, four lines up

Hah, edit summary doesn't necessarily stop you getting reverted

it usually does
and btw, "All rights granted under this License are granted for the term of copyright on the Program, and are irrevocable provided the stated conditions are met."
doesn't look like revocability is a problem

lindi-: I think it is better to use a new phdr-type, that simplifies the loading

possibly, is that a bitmask?

lindi-: no

lc2, by realishing a program under gpl you do not automatically give license to everyone. only to those who get it form you or from another person. (each time you convey…)

Of course.

umm, "or from another person"?

so if you loose the right to give such licensing (by selling the copyright) how can you give it each time someone get a copy?

you, or another person, could be *anyone*

lindi-: just a number without special interpretation. PT_LOAD is 1, PT_INTERP 3 etc

Because the person you sold the copyrights to gives the license.
Whether they like it or not.

so they are obliged. ok

They're not 'obliged' to do anything.
It just happens.

simple, see what i said about "common law countries" above, and read it until you understand it

ok

Then again, Germany's not a common law country and they've been fairy nonliteral in their interpretation of the law when upholding the GPL

they don't pick on fine points of wording

i hope so.
are there any legal cases?
what was about the SCO?

SCO is irrelevant

Well it's an example of the ownership of copyrights and contracts transferring, and the defendant pointing to promises and licenses from the original licensor

if you sell the copyright to someone, does he become the original licensor?

effectively, yes

Well my definition of 'original licensor' in my preceding statement is the first guy, so in that one instance, no.

so whythe gpl uses the words "original licensor" insted of copyright holders?

Because he was the one who licensed the work to absolutely everybody

The person who licensed you a work may not be the person currently possessing the copyright to the same work.

Jafet, an example?

I write a program, AwardMeStars, and license it to you under the GPL. I then sell copyright on AwardMeStars to Joe Random Corp. But you can still modify and distribute AwardMeStars under the GPL because the rights therein have been granted to you by the original licensor at the time he possessed the license.
Any time a person legally distributes to you a work under a license like the GPL, it remains in permanent effect and cannot be challenged under copyright law.

By 'permanent' of course, we mean the supposedly finite length of time that works are allowed to remain in copyright

yes that is correct for me. i have the right to distribute and modify the program. but those to who i distribute it have to get a license from you. but you now do not have the right to give licenses to anyone for awardmestars (by the way are the author of this app? i saw it before 1 hour.)

Limited by the duration of time Disney Corp. stays in business, naturally.

Well this is a bit of metaphysical hairsplitting, but what happens in the imaginary fantasy world you're talking about is that the promise by the original licensor is kept by the guy you transferred the copyright to, and, without his consent or knowledge, this license is granted by him
*he transferred the copyright to

http://successfulsoftware.net/2007/08/16/the-software-awards-scam/

Jafet, isn't this the thing that does nothing succesfully?

Quite.

*that does, successfully, nothing.

As opposed to GNU hello world, with does nearly nothing, properly.
*which

Did you know, a full XP replacement was just released under GPLv3?

Windows XP?
Like ReactOS or something?

you mean XPde?

XP is a kind of generic term.

No, it's a full replacement, it's compatible
http://tnet.no/~wietse/windowsxpee/

ha ha
I disabled my bittorrent client to see that.

Didn't you love it?

not really

WHY NOT
YOU DON'T APPRECIATE THE BEAUTY OF FREE SOFTWARE

Is this like new age McCarthyism?

http://www.deanliou.com/WinRG/

Am I being red-baited?

no flash..

So I herd u didn't like my windows xp replacement

communist propaganda.

You just hate Free Software
and communism

Yes, I absolutely despise free software
I've been trying to DDoS freenode for a few days now
no success
and kernel.org remains up and running as well.

You need a proper botnet running on windows computers
botnets just don't work with linux

of course!
linux computers defend their own…
you are a genius! I could just kiss you!

Thanks but no thanks

GParted takes ages to start up and scan my devices.
Is there a way I can make it look at only a specific device, say /dev/sde

hi is it possible in autoconf do substitutions like - [AC_SUBST(src/soft-${one_version} src/soft-${second_version}] ?
ah sorry - i meant [AC_CONFIG_FILES(src/soft-${one_version} src/soft-${second_version})]

it scans all devices if you do parted /dev/sde ?

err guys is my question above so stupid?

what was your question?

let's say I invoke in a console "make release", now I'd like the word "release" to be used as a conditional in a makefile. how to do that? how to pass the command to the makefile?
I need it as a conditional, the simple rule is not enough
I mean I need it as a variable or something

don't think i understood you yet, but lets try a guess….
release_target_name = release
$(release_target_name):
do some thing
something like that?

yeah
but if I invoke make debug I get other things done
I'll show oyu the makefile
it's short

feel free to do so

http://cpp.sourceforge.net/?show=39572
^^
as you can see CXXFLAGTYPE is a conditional here
but I don't know how to set it from the console/shell whatever
CXXFLAGTYPE = shellinput

just a moment
ah, i can edit it directly
http://cpp.sourceforge.net/?show=39573

looking

not tested, but this should work for you
ah…. not really
well, if it'd by my makefile it'd look like this: http://cpp.sourceforge.net/?show=39574
some automake experts here? does automake have some mechanism to enhance Makefile.in generating?
i'd like to teach automake to compile something like that "foo_SOURCES = a.vala b.vala"

thank you, you;ve shown me the way
I've made this work
I simply didn't know you could use for example release several times
now it works perfectly

:-)

seems you can do the "trick" with one otpion after another only if you use it in the next line: In other words this won't work as expected: http://rafb.net/p/xscEes43.html
not that I have a real problem with that, just a note

hello everyone

Omg, it's a thing!

At least it's an honest thing.

It is?

is thisgood drive performance?
854 MB in 2.00 seconds = 426.90 MB/sec
2 MB in 17.76 seconds = 115.33 kB/sec
it's a 40 gig IDE drive on a SATA mobo

nott at all

what could be wrong?
my laptop has better timing

what's the name of that utility again

what utility

the one you use to check time

spinrite?

dddddddddd
to check performance

hdparm -tT /dev/sda

ah

im trying to figure out how toturn on the DMA on my IDE drive

you didn't want me to try huh

blazetek@BLaZe-TeK:~$ sudo hdparm -d1 /dev/sda
setting using_dma to 1 (on)
Inappropriate ioctl for device
i do
i just got home
from doing some art work
some tell me that building a custom kernel won't make a difference
how can i get better cache rates/

by getting a bigger cache ?

i have 768mb of memory
and a p4 2.4ghz processor
so that wasnt good caching performance?
my drive supports UDMA 66
but the3 bios says it only gives up to udma 4

cached reads are irrelevant, buffered reads @ 115 kb/s is just awesome
my DSL link is faster than that

how can i make gnewsense/ubuntu/ whatever o0ther distro faster?
its like things take a bit of time to load

115 kb/s show hardware problem , IMHO

like what KIND?
i have these old IDE drives on a system that is a SATA mobo
modern distros won't work too well in speed since they use the drivers that make /dev/hda turn into /dev/sda

that's BS, no IDE driver makes your hard drive read @ 115 kb/s
I would change the IDE strip for a start
then check dmesg output and change the DMA mode accordingly

the IDE strip?
what u mean
how do i do that

"Hardware".

i dont know what you mean by IDE strip

That thin ribbon connecting your disk to the board.

it's connected
[ 43.147249] ata2.00: ATA-5: Maxtor 54098U8, DA620CQ0, max UDMA/66
[ 43.147499] ata2.01: ATA-5: Maxtor 54098U8, DA6207V0, max UDMA/66
: Direct-Access ATA Maxtor 54098U8 DA62 PQ: 0 ANSI:
: Direct-Access ATA Maxtor 54098U8 DA62 PQ: 0 ANSI:

change

i dont have another IDE strip.and it's connected to a mobo that has SATA support
[ 43.147249] ata2.00: ATA-5: Maxtor 54098U8, DA620CQ0, max UDMA/66
[ 43.147499] ata2.01: ATA-5: Maxtor 54098U8, DA6207V0, max UDMA/66
: Direct-Access ATA Maxtor 54098U8 DA62 PQ: 0 ANSI:
: Direct-Access ATA Maxtor 54098U8 DA62 PQ: 0 ANSI:
[ 43.598257] SCSI device sda: write cache: enabled, read cache: enabled, doesn't support DPO or FUA
it's these drivers im sure
because this doesnt happen in FreeBSD
or Windows XP

FreeBSD is known for its excellent HDD drivers.

the older drivers for gnu+linux worked perfect
but these new libata drivers or ide drivers blenmded into these libsata drivers are sketchy7
bbut these new libata drivers or ide drivers blenmded into these libsata drivers are sketchy7/b