BNETD and FSGS provide(d) ways to create an own battle.net(tm)-compatible server, for example on LANs where IPX gaming is impossible (yes they exist). I have got my latest bnetd source copy from my favorite BSD port mirror;)
Oh, yes, they also told people how to change the registry settings of Blizzard's programmes to connect to their servers. Maybe this was their greatest fault?!
Later than 0.9.5 it doesn't even run natively on OpenBSD, one has to use the Linux/x86-32 version. No, thanks, no Mozilla on Unix for me (Win32 ok). I'm using Opera 5.05 (Linux too) - it's stable anyway, and WTF do you need a graphical browser? I do 98% of my web surfing with Lynx.
Ok, I ack you on this one - for developers, yes. I hate programmes generating core files by myself, and so this should not be the default - unless you prove me wrong. You might take my comment about lazy coders who don't test for a start, but I don't know whether dumping core is always suitable.
Even if your argumentation is correct, to prevent lazy programmers from mistakes, the SECURE implementation is to be preferred by default. You can set FAST by environment variables. Content?
For windows, as for any other OS, it depends on free() which is usually a libc function. Except for mingw32 and very old MS C versions, the compiler is the library vendor, and so it depends on the compiler you are using whether the bug is active or not. (It is still in, though.)
Darn 2 Minute posting limit! Why don't people simply read the comments?
How would you catch the error in standard C, where free() is void, i.e. has no return code?
Believe me or not, the only viable ways are printing an error (warning for the pedantic) message or silently succeeding/ignoring. The former being the default.
Huh? The libz-bug does only affect Linux, on OpenBSD it does print an error. Ok, I didn't test windoze, but there it depends on the compiler (=library vendor). This is a big LOSS for Linux, to be exact (ok now I know I'll be modded down, but I prefer BSD for exactly this reason: security. Give me arguments for Linux/MacOS/whatever and I'll listen.).
Yup, thanks for the pointer, OpenBSD-current is definitively not affected. On the other hand, it _is_ a problem in libz, yeah, really, a bug, which can only let the program crash under glibc (I wonder if libc5 is affected). It does affect execution of programmes on OpenBSD though: an error message is written to fd 2. This can screw up UI programmes.
Well, Borland C++ 5.5 is free as in beer. Also there is mingw32 and cygwin32, but the latter is bad in sense of that programmes compiled with it must be GPL due to the linkage to cygwin1.dll (libc) which is GPL, whereas mingw32 produces native Win32 executables linked against msvcrt.dll which can be found on nearly any Win32 system (except for NT 3.1 and the very first version of Win95).
Pascal might be also free, you can download Turbo Pascal from museum.borland.com, Kylix is IIRC free for Linux, maybe also for Win32, and there is FreePascal.
Also you can install php4, either native or as a plugin for IIS or (better) apache. PHP4 runs quite stable, not as stable as e.g. Pascal, because it originally wasn't developed to support Win32, but it does work. If you aren't doing web stuff you should choose the native version over the webserver plugin.
It has, because this product actually consists of Linux and several packages of the GNU project. An opersting system kernel alone (Linux) is no operating system. (Neither is userland without a kernel, as seen on GNU;)
Errmm... I compiled my own distribution fully with libc5, and if I still was using Linux, I'd do so again, even with kernel 2.4.x - for IPv6 support there is libinet6(?) and who needs LOCALE support anyway? OpenBSD doesn't have it, either.
No wonder they don't, because Wintel is paying off, and their secret contract involves not supporting AMD (at least the new way);)
But I might be wrong, at least isn't Intel said to include x86-64 compatibility stuff into the next Pentium IV releases?
This would be a really, really cool way to get rid of M$ in a large market share, because _if_ IA-64 doesn't pay off, but x86-64 does (and it will, because of its ease to convert from and to x86-32), Intel will activate this, both chips sell (AMDs more I should guess), but M$ OS run only in 32-bit mode;)
If I am not fully confused, Microsoft hasn't
ever criticized open source software, but
Free (as in FSF) Software, specifically
that with the GPL license (I never saw a
microsoftie even know about the LGPL).
Open Source - http://opensource.org
is a completely different thing, and as you
might know, the NT 5 (aka Win2k) TCP/IP-Stack
(IPv6, by the way, might be not) is derived
from an early FreeBSD one. Legally ok.
And, while we are at TCP/IP, the first Winsock
implementation used an API in Berkeley style.
(However I do not know about actual code.)
Tell the OpenBSD team to import Lynx-2.8.4 Ctrl-V e for editor isn't working, but I do like to read my own posts before posting. Ctrl-X e (in 2-8-4) does work.
Slashdot Mirror
(No this wasn't OT but plain wrong):
;)
No one is abusing the battle.net network.
BNETD and FSGS provide(d) ways to create an own
battle.net(tm)-compatible server, for example on
LANs where IPX gaming is impossible (yes they
exist).
I have got my latest bnetd source copy from my
favorite BSD port mirror
Oh, yes, they also told people how to change
the registry settings of Blizzard's programmes
to connect to their servers. Maybe this was their
greatest fault?!
Later than 0.9.5 it doesn't even run natively
on OpenBSD, one has to use the Linux/x86-32 version.
No, thanks, no Mozilla on Unix for me (Win32 ok).
I'm using Opera 5.05 (Linux too) - it's stable
anyway, and WTF do you need a graphical browser?
I do 98% of my web surfing with Lynx.
libglib? Surely you jest...
Ok, I ack you on this one - for developers, yes.
I hate programmes generating core files by myself,
and so this should not be the default - unless you
prove me wrong.
You might take my comment about lazy coders who
don't test for a start, but I don't know whether
dumping core is always suitable.
Even if your argumentation is correct, to prevent
lazy programmers from mistakes, the SECURE
implementation is to be preferred by default.
You can set FAST by environment variables.
Content?
Lynx.
Or fix GNU malloc in your libc. OpenSSH links
/usr/lib/libkrb5.so.2.0 (0x4004f000) /usr/lib/libasn1.so.1.0 (0x40083000) /usr/lib/libkrb.so.9.0 (0x400a0000) /usr/lib/libkafs.so.8.0 (0x400b8000) /usr/lib/libcrypto.so.5.1 (0x400bb000) /usr/lib/libz.so.1.3 (0x40167000) /usr/lib/libc.so.28.2 (0x40173000)
dynamically against free(), doesn't it?
On my system:
$ ldd `which ssh`
/usr/bin/ssh:
-lkrb5.2 =>
-lasn1.1 =>
-lkrb.9 =>
-lkafs.8 =>
-lcrypto.5 =>
-lz.1 =>
-lc.28 =>
Damn Lameness Filter!
It depends on the Linux kernel's internal malloc
implementation.
Damn, how many times do I have to post that?
(As for the topic, I can't check, I'm a happy
BSD user)
It depends on the target platform's malloc implementation
(free, to be exact). OTOH you are correct.
Why went the GNU/Linux community to the Windows
equivalent of the DLL hell and distributing binaries?
OT, but nevertheless interesting.
For windows, as for any other OS, it depends on
free() which is usually a libc function.
Except for mingw32 and very old MS C versions,
the compiler is the library vendor, and so it
depends on the compiler you are using whether
the bug is active or not. (It is still in, though.)
Darn 2 Minute posting limit! Why don't people simply
read the comments?
How would you catch the error in standard C,
where free() is void, i.e. has no return code?
Believe me or not, the only viable ways are printing
an error (warning for the pedantic) message or
silently succeeding/ignoring. The former being the
default.
Huh? The libz-bug does only affect Linux, on OpenBSD
it does print an error.
Ok, I didn't test windoze, but there it depends on
the compiler (=library vendor).
This is a big LOSS for Linux, to be exact (ok now
I know I'll be modded down, but I prefer BSD for
exactly this reason: security. Give me arguments
for Linux/MacOS/whatever and I'll listen.).
Yup, thanks for the pointer, OpenBSD-current is
definitively not affected.
On the other hand, it _is_ a problem in libz,
yeah, really, a bug, which can only let the
program crash under glibc (I wonder if libc5 is
affected).
It does affect execution of programmes on OpenBSD
though: an error message is written to fd 2.
This can screw up UI programmes.
Well, Borland C++ 5.5 is free as in beer.
Also there is mingw32 and cygwin32, but the
latter is bad in sense of that programmes
compiled with it must be GPL due to the linkage
to cygwin1.dll (libc) which is GPL, whereas
mingw32 produces native Win32 executables linked
against msvcrt.dll which can be found on nearly
any Win32 system (except for NT 3.1 and the very
first version of Win95).
Pascal might be also free, you can download Turbo
Pascal from museum.borland.com, Kylix is IIRC free
for Linux, maybe also for Win32, and there is
FreePascal.
Also you can install php4, either native or as
a plugin for IIS or (better) apache.
PHP4 runs quite stable, not as stable as e.g.
Pascal, because it originally wasn't developed
to support Win32, but it does work.
If you aren't doing web stuff you should choose
the native version over the webserver plugin.
whether they also establish a index node,
such as public.bearshare.net or
connect*.gnutellanet.com(sp?)
Or do they just use the existing? Sh.t!
It has, because this product actually consists ;)
of Linux and several packages of the GNU project.
An opersting system kernel alone (Linux) is no
operating system. (Neither is userland without
a kernel, as seen on GNU
Errmm... I compiled my own distribution fully
with libc5, and if I still was using Linux, I'd
do so again, even with kernel 2.4.x - for IPv6
support there is libinet6(?) and who needs
LOCALE support anyway?
OpenBSD doesn't have it, either.
No wonder they don't, because Wintel is paying ;)
;)
off, and their secret contract involves not
supporting AMD (at least the new way)
But I might be wrong, at least isn't Intel said
to include x86-64 compatibility stuff into the
next Pentium IV releases?
This would be a really, really cool way to get
rid of M$ in a large market share, because _if_
IA-64 doesn't pay off, but x86-64 does (and it
will, because of its ease to convert from and
to x86-32), Intel will activate this, both chips
sell (AMDs more I should guess), but M$ OS run
only in 32-bit mode
Try to persuade CmdrTaco to post news
using Photoshop instead of $browser
And they can't provide bnetd and FSGS methods
/usr/ports/games/bnetd && make fetch
to check the CD keys,
just because - heh - their PROPRIETARY way
of checking wouldn't be safe !!!
So, Security by Obscurity?
Does industry never learn?
# cd
If I am not fully confused, Microsoft hasn't
ever criticized open source software, but
Free (as in FSF) Software, specifically
that with the GPL license (I never saw a
microsoftie even know about the LGPL).
Open Source - http://opensource.org
is a completely different thing, and as you
might know, the NT 5 (aka Win2k) TCP/IP-Stack
(IPv6, by the way, might be not) is derived
from an early FreeBSD one. Legally ok.
And, while we are at TCP/IP, the first Winsock
implementation used an API in Berkeley style.
(However I do not know about actual code.)
Hey, that is a cool quote of
/.ed,
The Hackers Handbook
The JARGON file
In its entirety you can view it at,
for example:
http://www.ccil.org/jargon/
Of course, if the site gets
Google*s cache and/or search functions
are your friends.
YMMV
Tell the OpenBSD team to import Lynx-2.8.4
Ctrl-V e for editor isn't working, but I do like
to read my own posts before posting.
Ctrl-X e (in 2-8-4) does work.
Especially when you think over the sentence:
The OS is more than just a kernel.
And when you see the problems in maintainance
Linus encountered since 2.0.