colinux-devel — Cooperative Linux development mailing list

RE: [coLinux-devel] /dev/hd* vs. /dev/cobd*

[Ronald P. <pij...@ds...>]

On Mon, Apr 19, 2004 at 01:17:51PM -0400, Daniel R. Slater wrote:
> Very interesting,
>   I wasn't aware you could do this - you learn something new evey day :=
).
> Would it be possible to use this method to dynamically change /etc/fsta=
b
> during startup? i.e. you could have something like /etc/fstab.native,
> /etc/fstab.colinux and then early in your startup code have something l=
ike:
>
> rm /etc/fstab
>
> if [ x"$BOOTSYSTEM" =3D=3D x"colinux"]; then
>         ln -s /etc/fstab.colinux /etc/fstab
> else
>         ln -s /etc/fstab.native /etc/fstab
> fi
>
> or does the linux startup process read /etc/fstab before executing any =
of
> it's startup scripts?

In fact, I have written something like that. I use the "init=3D/root/prei=
nit"
kernel parameter, where /root/preinit is a script that `fixes' things.
See attachment. If people find this useful, please add to the wiki.

Comments are welcome.

Ronald.

Re: [coLinux-devel] SMP workaround test results

[Richard G. <ric...@ri...>]

<sheepish grin> Yup, problem resolved ;-) </sheepish grin>
----- Original Message ----- 
From: "Ronald Pijnacker" <pij...@ds...>
To: "Richard Goodwin" <ric...@ri...>
Cc: "Cooperative Linux Development" <col...@li...>
Sent: Monday, April 19, 2004 1:10 PM
Subject: Re: [coLinux-devel] SMP workaround test results


> > So far I've had 100% success after about 30 VM cycles (start/reboot)
without
> > hanging. Seems like the workaround is pretty stable. I'm still having a
> > problem inside Gentoo where the gcc process gets killed every time I try
to
> > emerge something, but I don't know if it's related. Also haven't yet
tried
> > to start an xdm (or any dm) so I don't know if things will still hard
lock.
>
> I had a similar experience. It was related to too little memory. Maybe you
> should add a swap device?
>
> Ronald.
>
>

[coLinux-devel] Patch: XML file

[Richard G. <ric...@ri...>]

Patched the XML file to include some comments, good for first time users?

Richard

RE: [coLinux-devel] /dev/hd* vs. /dev/cobd*

[Daniel R. S. <dan...@ya...>]

Very interesting,
  I wasn't aware you could do this - you learn something new evey day :).
Would it be possible to use this method to dynamically change /etc/fstab
during startup? i.e. you could have something like /etc/fstab.native,
/etc/fstab.colinux and then early in your startup code have something like:

rm /etc/fstab

if [ x"$BOOTSYSTEM" == x"colinux"]; then
	ln -s /etc/fstab.colinux /etc/fstab
else
	ln -s /etc/fstab.native /etc/fstab
fi

or does the linux startup process read /etc/fstab before executing any of
it's startup scripts?

Dan

-----Original Message-----
From: col...@li...
[mailto:col...@li...] On Behalf Of John Nelson
Sent: Monday, April 19, 2004 1:01 PM
To: col...@li...
Subject: Re: [coLinux-devel] /dev/hd* vs. /dev/cobd*

Daniel Slater wrote:

> Another nice feature might be to create an entry either a directory or 
> a file in /proc that startup scripts can check for to determine if 
> you're running under colinux - something like
>
> if [ -e /proc/colinux ] then
>
> ..
>
> else
>
> fi
>
> (Excuse me if my bash syntax isn't quite correct, but I think you get 
> the idea. J )
>
In the mean time, you can use the <bootparams> element in the XML file 
to set an environmental variables that you can test on.... for example:
<bootparams>root=vmlinux BOOTSYSTEM=colinux</bootparams>
The colinux daemon will pass this on to the kernel, and the kernel will 
pass on the BOOTSYSTEM portion (which it doesn't understand) to init, 
and init will setup the appropriate environmental variable.

Once setup, your startup scripts can test for this by a number of means, 
such as:

if [ x"$BOOTSYSTEM" == x"colinux"]; then
blah, blah, blah
fi

I'm doing this right now in my dual-boot environment to automatically 
adjust various configuration files (in /etc) at startup and shutdown so 
that everything continues to work. That is, I have effectively created 
separate hardware profiles, and I use the presence of this environment 
variable to distinguish them.


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
_______________________________________________
coLinux-devel mailing list
coL...@li...
https://lists.sourceforge.net/lists/listinfo/colinux-devel

Re: [coLinux-devel] /dev/hd* vs. /dev/cobd*

[John N. <jo...@mo...>]

Daniel Slater wrote:

> Another nice feature might be to create an entry either a directory or 
> a file in /proc that startup scripts can check for to determine if 
> you’re running under colinux – something like
>
> if [ -e /proc/colinux ] then
>
> ….
>
> else
>
> fi
>
> (Excuse me if my bash syntax isn’t quite correct, but I think you get 
> the idea. J )
>
In the mean time, you can use the <bootparams> element in the XML file 
to set an environmental variables that you can test on.... for example:
<bootparams>root=vmlinux BOOTSYSTEM=colinux</bootparams>
The colinux daemon will pass this on to the kernel, and the kernel will 
pass on the BOOTSYSTEM portion (which it doesn't understand) to init, 
and init will setup the appropriate environmental variable.

Once setup, your startup scripts can test for this by a number of means, 
such as:

if [ x"$BOOTSYSTEM" == x"colinux"]; then
blah, blah, blah
fi

I'm doing this right now in my dual-boot environment to automatically 
adjust various configuration files (in /etc) at startup and shutdown so 
that everything continues to work. That is, I have effectively created 
separate hardware profiles, and I use the presence of this environment 
variable to distinguish them.

[coLinux-devel] README patch

[Richard G. <ric...@ri...>]

Alright, this is my first ever diff/patch of any kind, so please be gentle.
It's basically to get the README up to a more current state, and to give
some additional help on first time installations on Windows. I don't have
the CVS dirs, so the files are specified simply as README in the patch.

Comments/feedback requested!

RE: [coLinux-devel] /dev/hd* vs. /dev/cobd*

[Sergey O. <so...@so...>]

No:-( At least for me:-)

> -----Original Message-----
> From: col...@li...
> [mailto:col...@li...] On Behalf
> Of Daniel Slater
> Sent: Monday, April 19, 2004 11:45 AM
> To: Jaroslaw Kowalski
> Cc: col...@li...
> Subject: RE: [coLinux-devel] /dev/hd* vs. /dev/cobd*
>
>
> I forgot about the LABEL= option. Does this still not work
> even with the latest fixes in 0.6.1 that allow properly
> reading the size of a
> /dev/cobd* partition?
>
> Dan
>
>
> -----Original Message-----
> From: Jaroslaw Kowalski [mailto:ja...@zd...]
> Sent: Monday, April 19, 2004 11:40 AM
> To: Daniel Slater; col...@li...
> Subject: Re: [coLinux-devel] /dev/hd* vs. /dev/cobd*
>
>
> It would be great if colinux supported specifying "LABEL=" as
> device name both in kernel parameters and /etc/fstab.
>
> Currently it doesn't seem to work. With this in place you'd
> be able to write in /etc/fstab:
>
> --------------------------
> LABEL=/ROOT / ext3 defaults 0 0
> LABEL=/USR /usr ext3 defaults 0 0
> --------------------------
>
> Under colinux it would map to /dev/cobdXXX, under "hardware"
> linux it would be /dev/hdaXXX Unfortunately I have no kernel
> knowledge to fix this myself, but it should be pretty easy.
>
> +1 for the "/proc/colinux" idea.
>
> Jarek
>
> BTW. I don't think that mapping /dev/cobd to /dev/hda would
> be good. They support different fctls() so it would probably
> do a lot of harm.
>
> ----- Original Message -----
> From: "Daniel Slater" <dan...@au...>
> To: <col...@li...>
> Sent: Monday, April 19, 2004 4:22 PM
> Subject: [coLinux-devel] /dev/hd* vs. /dev/cobd*
>
>
> Currently, colinux creates new block special devices for it's
> drives, /dev/cobd1, /dev/cobd2, etc. Since we're already
> patching the linux kernel, how difficult would it be to patch
> the code responsible for handling the /dev/hd* and /dev/sd*
> block devices so that they are redirected to the colinux
> devices. If this could be done, it would certainly simplify
> the lives of the people with dual-boot systems. The xml
> format could be enhanced to something look like this:
>
>
>
>
>
> <block_device index="0" path="\Device\Harddisk0\Partition1"
> alias_major="3" alias_minor="1" enabled="true" />
>
>
>
>
>
> This would redirect /dev/hda1 to /dev/cobd0.
>
>
>
> Another nice feature might be to create an entry either a
> directory or a file in /proc that startup scripts can check
> for to determine if you're running under colinux - something like
>
>
>
> if [ -e /proc/colinux ] then
>
> ....
>
> else
>
>
>
> fi
>
>
>
> (Excuse me if my bash syntax isn't quite correct, but I think
> you get the idea. :-) )
>
>
>
>
>
>
>
>
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by: IBM Linux Tutorials
> Free Linux tutorial presented by Daniel Robbins, President
> and CEO of GenToo technologies. Learn everything from
> fundamentals to system
> administration.http://ads.osdn.com/?ad_id70&alloc_id638&op=ick
> _______________________________________________
> coLinux-devel mailing list
> coL...@li...
> https://lists.sourceforge.net/lists/listinfo/colinux-devel
>
>

RE: [coLinux-devel] /dev/hd* vs. /dev/cobd*

[Daniel S. <dan...@au...>]

I forgot about the LABEL=3D option. Does this still not work even with =
the
latest fixes in 0.6.1 that allow properly reading the size of a
/dev/cobd* partition?

Dan


-----Original Message-----
From: Jaroslaw Kowalski [mailto:ja...@zd...]=20
Sent: Monday, April 19, 2004 11:40 AM
To: Daniel Slater; col...@li...
Subject: Re: [coLinux-devel] /dev/hd* vs. /dev/cobd*


It would be great if colinux supported specifying "LABEL=3D" as device
name
both in kernel parameters and /etc/fstab.

Currently it doesn't seem to work. With this in place you'd be able to
write
in /etc/fstab:

--------------------------
LABEL=3D/ROOT / ext3 defaults 0 0
LABEL=3D/USR /usr ext3 defaults 0 0
--------------------------

Under colinux it would map to /dev/cobdXXX, under "hardware" linux it
would
be /dev/hdaXXX
Unfortunately I have no kernel knowledge to fix this myself, but it
should
be pretty easy.

+1 for the "/proc/colinux" idea.

Jarek

BTW. I don't think that mapping /dev/cobd to /dev/hda would be good.
They
support different fctls() so it would probably do a lot of harm.

----- Original Message -----=20
From: "Daniel Slater" <dan...@au...>
To: <col...@li...>
Sent: Monday, April 19, 2004 4:22 PM
Subject: [coLinux-devel] /dev/hd* vs. /dev/cobd*


Currently, colinux creates new block special devices for it's drives,
/dev/cobd1, /dev/cobd2, etc. Since we're already patching the linux
kernel, how difficult would it be to patch the code responsible for
handling the /dev/hd* and /dev/sd* block devices so that they are
redirected to the colinux devices. If this could be done, it would
certainly simplify the lives of the people with dual-boot systems. The
xml format could be enhanced to something look like this:





<block_device index=3D"0" path=3D"\Device\Harddisk0\Partition1"
alias_major=3D"3" alias_minor=3D"1" enabled=3D"true" />





This would redirect /dev/hda1 to /dev/cobd0.



Another nice feature might be to create an entry either a directory or a
file in /proc that startup scripts can check for to determine if you're
running under colinux - something like



if [ -e /proc/colinux ] then

....

else



fi



(Excuse me if my bash syntax isn't quite correct, but I think you get
the idea. :-) )

Re: [coLinux-devel] /dev/hd* vs. /dev/cobd*

[Jaroslaw K. <ja...@zd...>]

It would be great if colinux supported specifying "LABEL=" as device name
both in kernel parameters and /etc/fstab.

Currently it doesn't seem to work. With this in place you'd be able to write
in /etc/fstab:

--------------------------
LABEL=/ROOT / ext3 defaults 0 0
LABEL=/USR /usr ext3 defaults 0 0
--------------------------

Under colinux it would map to /dev/cobdXXX, under "hardware" linux it would
be /dev/hdaXXX
Unfortunately I have no kernel knowledge to fix this myself, but it should
be pretty easy.

+1 for the "/proc/colinux" idea.

Jarek

BTW. I don't think that mapping /dev/cobd to /dev/hda would be good. They
support different fctls() so it would probably do a lot of harm.

----- Original Message ----- 
From: "Daniel Slater" <dan...@au...>
To: <col...@li...>
Sent: Monday, April 19, 2004 4:22 PM
Subject: [coLinux-devel] /dev/hd* vs. /dev/cobd*


Currently, colinux creates new block special devices for it's drives,
/dev/cobd1, /dev/cobd2, etc. Since we're already patching the linux
kernel, how difficult would it be to patch the code responsible for
handling the /dev/hd* and /dev/sd* block devices so that they are
redirected to the colinux devices. If this could be done, it would
certainly simplify the lives of the people with dual-boot systems. The
xml format could be enhanced to something look like this:





<block_device index="0" path="\Device\Harddisk0\Partition1"
alias_major="3" alias_minor="1" enabled="true" />





This would redirect /dev/hda1 to /dev/cobd0.



Another nice feature might be to create an entry either a directory or a
file in /proc that startup scripts can check for to determine if you're
running under colinux - something like



if [ -e /proc/colinux ] then

....

else



fi



(Excuse me if my bash syntax isn't quite correct, but I think you get
the idea. :-) )

[coLinux-devel] bugs in release 0.6

[Terje W. <wie...@sa...>]

Hey all

I have tried release 0.6 of colinux, and while I am very impressed there =
=20
are a few bugs that bother me:

-Loading a partition as a block-device doesn't work. I have entered the =20
following block devices:
	<block_device index=3D"0" =20
path=3D"\DosDevices\D:\colinux\Debian-3.0r0.ext3.1gb.img" enabled=3D"true=
">
     </block_device>
     <block_device index=3D"1" path=3D"\Device\Harddisk0\Partition5" =20
enabled=3D"true">
     </block_device>
     <block_device index=3D"2" path=3D"\Device\HarddiskVolume5" enabled=3D=
"true">
     </block_device>
But only the first of them (the debian image)works. This means I have to =
=20
maintain 2 linux installations, which feels quite unneccesary.

-The Xfree86 dummy server freezes the whole machine. I wanted to use it s=
o =20
I could run kdm and connect to that from a remote machine, but with no =20
luck. It might be code trying to access the keyboard or mouse that =20
actually crashes it, I don't know.

-In the debian image, the loopback device is not configured, and the file=
 =20
/etc/hosts is missing. This is, amongst other things, preventing kde from=
 =20
starting.

-Running vncserver, I can get kde running, when the above is fixed. =20
However, when I try to run opera, the whole opera screen goes black and =20
then it crashes after a minute or so. With konqueror, however, browsing =20
works just fine. Also, when I forward X11 via ssh, opera runs fine.

Best regards
Terje Wiesener

PS: I'm not on the list, so if you want to reply to me do it to my email.

[coLinux-devel] /dev/hd* vs. /dev/cobd*

[Daniel S. <dan...@au...>]

Currently, colinux creates new block special devices for it's drives,
/dev/cobd1, /dev/cobd2, etc. Since we're already patching the linux
kernel, how difficult would it be to patch the code responsible for
handling the /dev/hd* and /dev/sd* block devices so that they are
redirected to the colinux devices. If this could be done, it would
certainly simplify the lives of the people with dual-boot systems. The
xml format could be enhanced to something look like this:

=20

=20

<block_device index=3D"0" path=3D"\Device\Harddisk0\Partition1"
alias_major=3D"3" alias_minor=3D"1" enabled=3D"true" />

=20

=20

This would redirect /dev/hda1 to /dev/cobd0.

=20

Another nice feature might be to create an entry either a directory or a
file in /proc that startup scripts can check for to determine if you're
running under colinux - something like=20

=20

if [ -e /proc/colinux ] then

....

else

=20

fi

=20

(Excuse me if my bash syntax isn't quite correct, but I think you get
the idea. :-) )

=20

=20

RE: [coLinux-devel] build environment on Windows

[Daniel R. S. <dan...@ya...>]

Actually, I did successfully build colinux under windows once. Instead of
building a gcc cross-compiler, I built a native gcc compiler, then renamed
the tools ( like gcc, ar, ld, etc ) to their cross-compiler versions ( i.e.
i686-cygwin-gcc, etc). However, since you need a linux kernel source tree
and a built vmlinux, it seemed a little pointless - you will still need a
working linux (colinux works fine) to build the kernel and then you would
need to copy all the linux stuff over to windows.

The only real reason I can see for wanting to build under windows is if you
wanted to try and use Visual Studio and the Microsoft DDK to build. However,
this would involve a considerable number of changes as the compiler switches
are different, the source tree layout for device drivers is different, and
finally, the inline assembly syntax is different (Intel syntax vs. AT&T).

On the other hand, if linux.sys were built with the Microsoft DDK, you could
source debug it with the Microsoft tools. Currently  I don't believe there
is any way to source debug linux.sys since WinDbg doesn't know how to read
gcc symbols and there are no GNU tools for kernel debugging Windows.



-----Original Message-----
From: col...@li...
[mailto:col...@li...] On Behalf Of Matt Behrens
Sent: Monday, April 19, 2004 6:48 AM
To: Robert Citek
Cc: col...@li...
Subject: Re: [coLinux-devel] build environment on Windows

On Mon, Apr 19, 2004 at 02:37:25AM -0500, Robert Citek wrote:

> How does one build coLinux from source on Windows?
> 
> I've been reading through the wiki page:
>   http://colinux.org/wiki/index.php/CygwinToolchain
> It seems as though these instructions are for building coLinux on Linux.

Unless something has changed that I've missed, you can't yet build
on Windows.

RE: [coLinux-devel] build environment on Windows

[peter g. <plu...@p1...>]

you use colinux of course ;)

-----Original Message-----
From: col...@li...
[mailto:col...@li...]On Behalf Of Robert
Citek
Sent: 19 April 2004 08:37
To: col...@li...
Subject: [coLinux-devel] build environment on Windows



Hello all,

How does one build coLinux from source on Windows?

I've been reading through the wiki page:
   http://colinux.org/wiki/index.php/CygwinToolchain
It seems as though these instructions are for building coLinux on Linux.

Regards,
- Robert



-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
_______________________________________________
coLinux-devel mailing list
coL...@li...
https://lists.sourceforge.net/lists/listinfo/colinux-devel
---
Incoming mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.659 / Virus Database: 423 - Release Date: 15/04/2004

---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.659 / Virus Database: 423 - Release Date: 15/04/2004

RE: [coLinux-devel] Using VNC/X ->Win32 Link Off-Network

[Mathias W. <mat...@gm...>]

Hi Michael,

maybe you should disable media sense. There is an article at MSKB:
http://support.microsoft.com/default.aspx?scid=kb;EN-US;239924

--
Mathias

> -----Original Message-----
> From: col...@li... 
> [mailto:col...@li...] On Behalf 
> Of Michael Sisolak
> Sent: Saturday, April 17, 2004 5:47 PM
> To: col...@li...
> Subject: [coLinux-devel] Using VNC/X ->Win32 Link Off-Network
> 
> 
> I have coLinux running on my Windows 2000 laptop with a native network
> configuration (started with the Debian image and added X, VNCserver,
> and KDE 3.2.1).  As long as my laptop is plugged into the 
> network I can
> use a Win32 VNC client to hook into the X system in the coLinux
> instance.  As soon as I unhook from the network, however, the coLinux
> instance and the Win32 instance can no longer communicate with each
> other.
> 
> It appears that in a native networking configuration coLinux can only
> talk to Win32 programs by first passing the packets out to my router
> and then having the router pass them back to the laptop.  
> Thus with the
> laptop off the network the two instances have no way to communicate.
> 
> Is there some way to get off-network usage in a native network
> configuration to work?  Thanks for all your work on CoLinux - it's
> quite an impressive trick.
> 
> Michael Sisolak
> msi...@ya...
> 
> 
> 	
> 		
> __________________________________
> Do you Yahoo!?
> Yahoo! Tax Center - File online by April 15th
> http://taxes.yahoo.com/filing.html
> 
> 
> -------------------------------------------------------
> This SF.Net email is sponsored by: IBM Linux Tutorials
> Free Linux tutorial presented by Daniel Robbins, President and CEO of
> GenToo technologies. Learn everything from fundamentals to system
> administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
> _______________________________________________
> coLinux-devel mailing list
> coL...@li...
> https://lists.sourceforge.net/lists/listinfo/colinux-devel
> 

Re: [coLinux-devel] Using VNC/X ->Win32 Link Off-Network

[<ch...@to...>]

are you using TAP or Bridged networking?
the TAP and setting up connection sharing should work fine when
disconnected. I think there is a disclamer of this sort on the networking
howto for bridged.

chris


> I have coLinux running on my Windows 2000 laptop with a native network
> configuration (started with the Debian image and added X, VNCserver,
> and KDE 3.2.1).  As long as my laptop is plugged into the network I can
> use a Win32 VNC client to hook into the X system in the coLinux
> instance.  As soon as I unhook from the network, however, the coLinux
> instance and the Win32 instance can no longer communicate with each
> other.
>
> It appears that in a native networking configuration coLinux can only
> talk to Win32 programs by first passing the packets out to my router
> and then having the router pass them back to the laptop.  Thus with the
> laptop off the network the two instances have no way to communicate.
>
> Is there some way to get off-network usage in a native network
> configuration to work?  Thanks for all your work on CoLinux - it's
> quite an impressive trick.
>
> Michael Sisolak
> msi...@ya...
>
>
>
>
> __________________________________
> Do you Yahoo!?
> Yahoo! Tax Center - File online by April 15th
> http://taxes.yahoo.com/filing.html
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by: IBM Linux Tutorials
> Free Linux tutorial presented by Daniel Robbins, President and CEO of
> GenToo technologies. Learn everything from fundamentals to system
> administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
> _______________________________________________
> coLinux-devel mailing list
> coL...@li...
> https://lists.sourceforge.net/lists/listinfo/colinux-devel
>

Re: [coLinux-devel] AltGr Keys

[<ch...@to...>]

If you can get networking up maby conecting with putty you could use the
windows keyboard map. Otherwise it would be the same as setting up a
keyboard map in linux. You need to configure your root image to do the
mapping.

chris

> Hi all,
>
> how can I setup colinux consoles to accept the AltGr-Keys (i.e. @ on
> german keyboards)?
> I've setup the correct keyboard map and all keys are working, except the
> AltGr-keys.
> It seems to be a problem in the console.
> If I start xterm inside the console, it works well.
>
> --
> Regards,
> Mathias
>
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by: IBM Linux Tutorials
> Free Linux tutorial presented by Daniel Robbins, President and CEO of
> GenToo technologies. Learn everything from fundamentals to system
> administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
> _______________________________________________
> coLinux-devel mailing list
> coL...@li...
> https://lists.sourceforge.net/lists/listinfo/colinux-devel
>

Re: [coLinux-devel] [ANN] Slackware 9.1.0 image for coLinux (ZipSlack)

[<ch...@to...>]

check out topologilinux.com it is like this but you have access to all of
slackware installer and it will work on an ntfs partition too.
downside is large download a full iso plus another more than half size iso
if you want gnome kde openoffice or nvidea drivers

chris

> hi people,
>
>
> As you can read on Slackware web site, http://www.slackware.org/zipslack/
>
>
> "ZipSlack is a special edition of Slackware Linux that can be installed
> onto
> any FAT (or FAT32) filesystem with about 100 MB of free space."
>
> (...some umsdos related stuff...)
>
> "This distribution is ideal for people who don't have a lot of hard disk
> space, do not have a fast Internet connection to download the entire
> distribution, or who want a Linux distribution they can carry around on a
> Zip disk."
>
>
>
> I created a colinux image similar to the existing Debian, Gentoo and
> Fedora
> images. and it boots without problem, uname -a displays : Linux slackware
> 2.4.25-co-0.6.0
>
>
> the compressed image is now about 37 Mo. But I plan to remove some tools
> and
> applications irrelevant to coLinux like audio and cd burning applications,
> so this number will drastically decrease.
>
>
>
>
> For the impatients, I must point out that this mini-distribution was
> originally targeted as people wanting to try Linux without partitionning
> and
> formatting their hard drive with Windows on it.
>
>
> so right now, Slackware 9.1.0 including :
>
> *most core Unix utilities (many 'a' packages)
> *most necessary network tools to connect to internet (many 'n' packages),
> including ssh
> *some very small games ('y' bsd-games)
> *no "heavy" developpment tools like gcc, perl, python (nothing. gawk and
> bash are part of 'a')
> *no X Window, no KDE, no Mozilla...
>
>
> there are the usual tools pkgtool, installpkg, rpm2targz to install any
> software you would need, but the installation of software is not as
> streamlined as Debian (apt-get) or Gentoo (emerge), so this image would be
> most usefull for Slackware diehards.
>
>
>
> I still have to iron out a few things, but I plan to make available two or
> three images, including a really small one.
>
>
> if there is some interest I could load one image with developpement tools
> necessary to compile most applications, and X Window base applications
> (xterm), all straight from Slackware 'd' and 'xap' official packages.
>
>
>
>
> methodology :
>
>
> I just unzipped zipslack.zip to a FAT32 partition (my D: drive), rebooted
> Windows to give coLinux exclusive access to this drive, created a block
> device similar to :
>
> <block_device index="3" path="\Device\Harddisk1\Partition1"
> enabled="true"></block_device>
>
> and mounted it inside coLinux under umsdos, read only. I then copied all
> files to a directory, permissions were saved (the whole point).
>
> I had to edit /etc/fstab, (changing references like /dev/hda0 to
> /dev/cobd0
> ...), apply some mknod to "declare" /dev/cobd1 and /dev/cobd2 and that was
> all.
>
>
> I could have used a 'real' Slackware installation CD but I wanted to
> quicky
> create a minimal distribution.
>
>
>
> O. Souiry
>
>
>
>
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by: IBM Linux Tutorials
> Free Linux tutorial presented by Daniel Robbins, President and CEO of
> GenToo technologies. Learn everything from fundamentals to system
> administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
> _______________________________________________
> coLinux-devel mailing list
> coL...@li...
> https://lists.sourceforge.net/lists/listinfo/colinux-devel
>

Re: [coLinux-devel] build environment on Windows

[Matt B. <ma...@zi...>]

On Mon, Apr 19, 2004 at 02:37:25AM -0500, Robert Citek wrote:

> How does one build coLinux from source on Windows?
>=20
> I've been reading through the wiki page:
>   http://colinux.org/wiki/index.php/CygwinToolchain
> It seems as though these instructions are for building coLinux on Linux.

Unless something has changed that I've missed, you can't yet build
on Windows.

Re: [coLinux-devel] Using VNC/X ->Win32 Link Off-Network

[Matt B. <ma...@zi...>]

On Mon, Apr 19, 2004 at 09:03:20AM +0200, Marko Bozikovic wrote:

> I had the same problem while I was working on Linux under VMWare on my=20
> laptop.
>  This can be solved by making a small 'loopback' UTP connector: one UTP
> connector and two wires. Plug it in, and Windows detect the card as=20
> connected.
>=20
> I'll have to ask our network guy for the instructions :)

Connect pin 1 to pin 3 and pin 2 to pin 6.

It would be interesting to externally sniff a setup like this to see if
it's "leaking" cross-system packets outside the system.  It probably
is...

[coLinux-devel] build environment on Windows

[Robert C. <rw...@al...>]

Hello all,

How does one build coLinux from source on Windows?

I've been reading through the wiki page:
   http://colinux.org/wiki/index.php/CygwinToolchain
It seems as though these instructions are for building coLinux on Linux.

Regards,
- Robert

Re: [coLinux-devel] Using VNC/X ->Win32 Link Off-Network

[Marko B. <bo...@ks...>]

Michael Sisolak wrote:
> I have coLinux running on my Windows 2000 laptop with a native network
> configuration (started with the Debian image and added X, VNCserver,
> and KDE 3.2.1).  As long as my laptop is plugged into the network I can
> use a Win32 VNC client to hook into the X system in the coLinux
> instance.  As soon as I unhook from the network, however, the coLinux
> instance and the Win32 instance can no longer communicate with each
> other.
> -snip-

I had the same problem while I was working on Linux under VMWare on my laptop.
  This can be solved by making a small 'loopback' UTP connector: one UTP
connector and two wires. Plug it in, and Windows detect the card as connected.

I'll have to ask our network guy for the instructions :)
-- 
Marko
ICQ: 5990814

Him: "Your skin is so soft.  Are you a model?"
Her: "No,"  [blush]  "I'm a cosmetologist."
Him: "Really? That's incredible... It must be very tough to handle
      weightlessness."
    -- "The Jerk"

Re: [coLinux-devel] Minimal Fedora Root

[Jaroslaw K. <ja...@zd...>]

> There are still some missing modules:
> char-major-21, char-major-10-135, block-major-2

char-major-21 is Generic SCSI devices - I'm not sure if it can be included
char-major-10-156 is a real-time-clock. It probably should be included (if
it works - Dan?)
block-major-2 is a floppy disk -  probably cannot be included yet
block-major-3 is a IDE/ATA disk - probably cannot be included yet

> Also it would be usefull, if you include the packages kbd and
> redhat-config-keyboard.

Can't you apt-get them?

> I could not get dhcp to work. But maybe this could not work with bridged
> TAP?

Perhaps. Does bridged TAP propagate broadcasts ?
Can you try with a real bridged networking?

> If I use redhat-config-network to change network settings, they made in
> the wrong file (/etc/sysconfig/networking/ifcfg-eth0). It must be
> changed in /etc/sysconfig/network-scripts/ifcfg-eth0).

Ugh. I never noticed that, maybe that's because I run my system on default
IP... ;-)

Jarek

[coLinux-devel] Seeking Beta Testers for coLinux =?windows-1252?Q?Windows=92_S?= =?windows-1252?Q?ervice_and_Manager_Application?=

[Tony B. <bie...@bi...>]

I’ve started a project to create a coLinux service and manager 
application. The working Beta code launches coLinux as a Windows service 
(localsystem account), and allows users to manage/connect to the coLinux 
process through a system-tray based interface.

I’m looking for a few testers to participate in a controlled Beta 
testing phase. Please email me at bie...@bi... if you’re 
interested.

For more info, please see the screenshots located at this link:

http://www.biermana.org/index.php?p=54

Thanks,
Tony Bierman

RE: [coLinux-devel] running colinux as a service under the NETWORK SERVICES account

[Daniel R. S. <dan...@ya...>]

Thanks for the tip. This is definitely a cool tool. Interestingly, when =
I
bring up the security dialog for a raw partition, I am unable to add the
NETWORK SERVICES account. It doesn't show up when I use the security =
dialog
to search, and it is rejected if I try to type it in.=20

O well - I suspect it is due to the nature of the network services =
account -
looks like I'll just have to create an account to run colinux under and
restrict it's access.

Dan

-----Original Message-----
From: col...@li...
[mailto:col...@li...] On Behalf Of tei
Sent: Sunday, April 18, 2004 10:29 AM
To: col...@li...
Subject: Re: [coLinux-devel] running colinux as a service under the =
NETWORK
SERVICES account

Jaroslaw Kowalski escribi=F3:

> Try this:
>=20
> http://www.sysinternals.com/ntw2k/freeware/winobj.shtml
>=20
> I'm not sure if this will work, but it's pretty powerful.
>=20
> Jarek
>=20

cool tool,

also list Devices has needed to fill our xml config :D
(looks at ArcName folder)


> ----- Original Message -----=20
> From: "Daniel Slater" <dan...@au...>
> To: <col...@li...>
> Sent: Sunday, April 18, 2004 3:52 PM
> Subject: [coLinux-devel] running colinux as a service under the =
NETWORK
> SERVICES account
>=20
>=20
> I have been trying to get colinux to run as a service under the =
NETWORK
> SERVICES account. I have given the account permission to access the
> files and directories in my colinux directory, however I am using a =
raw
> disk partition for my linux filesystem. Does anyone know how I can =
give
> the NETWORK SERVICES account access to my linux partition?
>=20
>=20
>=20
> Dan
>=20
>=20
>=20




-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=3D1470&alloc_id=3D3638&op=3Dcli=
ck
_______________________________________________
coLinux-devel mailing list
coL...@li...
https://lists.sourceforge.net/lists/listinfo/colinux-devel

RE: [coLinux-devel] Minimal Fedora Root

[Daniel R. S. <dan...@ya...>]

Dhcp should work fine with bridged networking, however, the linux kernel
needs to be compiled with Socket Filtering turned on (CONFIG_FILTER=y) to
enable dhcp. It might be a good idea for the colinux kernel config to have
this turned on by default.

I have attached the kernel config file I have been using successfully for
fedora core 1. In addition to having socket filtering, I have turned on
modules for most of the non-experimental filesystems. It also compiles
support for a.out and misc binaries as modules instead of linking them in
statically. 

Dan 


-----Original Message-----


I could not get dhcp to work. But maybe this could not work with bridged
TAP?