Baslinux mailing list

[Lee Forrest <lforrestster AT gmail.com>]

On Mon, Jan 29, 2007 at 05:06:38PM +1030, Greg Mayman wrote:
> On Sun, 28 Jan 2007 19:09:09 +0000, Lee Forrest wrote:
> 
> > DOS is M$. Baby Windows.
> 
> Oh dear God, NO!
> 
> DOS is more like commandline Linux or Unix.

I meant that windows was built on MSDOS, which it was.

Just because an OS can be run from the commandline doesn't mean
it is like linux/unix. Until recently, you could run windows from
the DOS commandline. 

In as much as the creators of windows let anyone but them control
what their OS does. So it didn't need the functionality of a linux
kernel, basesystem, shell.

And some linux distros are run almost completely from a 'graphical
desktop environment' like windows.

> And as for being Micro$not... that only applies to one type,
> MS-DOS which they bought the rights for from someone else.
> 
> At the time there was also IBM-DOS, and nowadays there are DrDOS,
> FreeDOS and several others that have no association at all with
> Micro$not!

If they use basically the same OS as MSDOS, then they have a
_lot_ to do with it.

Just like linux has a lot to do with unix.

> > You picked the hard way to learn linux. But it'll pay off in the long run.
> > I highly recommend the first 10 or so chapters of this course:
> > wget -O - http://rute.2038bug.com/rute.html.tar.bz2 | tar -xvvjf -
> > That'll bring it down and bunzip and untar it and put the files in a 
> > directory
> > called rute.html.
> 
> I made some wild guesses and after doing a ppp-on I opened xterm
> and typed in the command as you gave it to me, and it all worked!

Frankly, I thought that Ron was right, and I was going to offer to
send you an abridged version.

Good work.

wget -O - http://rute.2038bug.com/rute.html.tar.bz2 | tar -xvvjf -

wget is a non-interactive web tool. 

wget http://....

does the same thing as 'going' to the website and going through all the
rigamarole to save the file you are viewing to disk.

-0 tells it to save the file to a file named "-", which means
to stdout. The screen. But it never gets there because it is
piped "|" to the input of tar, which decompresses it -j (be -z
instead of j for gzipped tarball) and extracts the contents of
the archive -x, and sends them to -f file, which is "-", or
stdout, again.

Note that stdout really is a file. /dev/stdout. Try this:

echo FOO > /dev/stdout

then

cat /dev/stdout

-vv just means "very verbose".

(If you left out '-0 file' it would be saved in your current directory
under its name.)

> But I can't read it until I can install the lynx package that I
> have also just downloaded.

That should take you about 5 minutes. I put them in /, though you don't
really have to, and run:

tar tvzf package.tgz | less

Never hurts to take a look at what's there and make sure it isn't going to 
walk
on something important. 

Then tar tvzf pakckage.tgz >> ~/.packages 

Pays to keep track of what you have installed and where.

That ">>" appends the output to the file. If you use one ">" it will
overwrite the file. Both will create the file if it doesn't exist.

I make a note at the beginning and ending of each package entry, identifying
it. Probably should put the date there too. 

Then gzip ~/.packages.

Then run:

pkg package.tar.gz

--------

Often, I only want one or two executables/libraries/docs from a
tarball, and extract them like this:

tar xvzf package.tgz   dir/dir/dir/dir/file 

That last part, full-path-to-file has to be _exactly_ what is 
printed with 'tar tvzf package.tgz'. Good idea to cut&paste those
paths.

You can put as many full-paths-to-files as you want on the
commandline, seperated by spaces.

When you do this, they are installed where it says, not just
copied to the directory you are in...


Lee

-- 
BasicLinux: Small is Beautiful
http://www.basiclinux.com.ru