[Rule-list] New slinky images, copyright question
Devon
devon at tuxfan.homeip.net
Thu Mar 28 03:46:28 EET 2002
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On Wednesday 27 March 2002 07:57 pm, Martin Stricker wrote:
> Devon wrote:
> > I've reverted to full file names in the package lists, so the install
> > needs the packages from the official 7.2 release. Including updates
> > will break it. I do however also have lists generated (and a script
> > to automate the list creation process) for a fully updated set of
> > packages.
>
> Why this step back?
It wasn't because I wanted to, I can tell you that. :)
I'm struggling to come up with a nice easy way to take the package
basename, and say "OK, find the rpm package with the same base name, and
install it." This has problems, because all of the different kernel
packages have the same basename, as do the glibc{i386,i686} packages.
For example, saying install kernel*i386.rpm looks promising, except that
it will install the kernel, the kernel-headers, and kernel-pcmcia
packages. The same holds true for similar names. "file" comes to mind,
which matches file, file-utils, filesystem, rootfiles, audiofile...
When the rpm files are available locally, this is ugly, but fairly easy
to do. Here's the nasty looking little section of script:
RPM_PATH="/mnt/cdrom/RedHat/RPMS/"
CUT="cut -d "/" -f 6"
RPM_COMMAND="rpm -ivh --nodeps ${RPM_PATH}"
RPM_TEST_COMMAND="rpm -qp --qf "%{NAME}""
for package in `cat /root/scripts/base_packages`
do
candidates1=`ls $RPM_PATH$package* | $CUT | grep i386`
candidates2=`ls $RPM_PATH$package* | $CUT | grep noarch`
candidates=`echo $candidates1 && echo $candidates2`
if [ -n "$candidates" ]; then
for match in `echo $candidates`
do
test_match=`"$RPM_TEST_COMMAND" "$match"`
if [ "$test_match" = "$package" ]; then
"$RPM_COMMAND" "$match"
if [ $? -ne 0 ] ; then
echo "Error reported while installing $match."
echo Error reported while installing\
$match. >> /root/scripts/install_log
fi
fi
done;
else
echo "OOPS! Failed to locate $package."
echo Failed to locate $package. >> /root/scripts/install_log
fi
done;
Adding ftp and http support to the script required some additions, which
still worked fine, but for each package, the installer had to connect to
the server to run the querry, then connect again to download the matching
package. It seemed like a lot of added bandwidth, so for the moment, I
removed it. I do intend to add it back, and if anyone has better ideas on
how to accomplish this, I'd love to hear them.
Also, to accomplish this, the installer needs a directory listing from
the server, and the stripped down 'wget' from busybox doesn't create a
directory listing when pointed at a directory like the full wget does.
Additionally, you can't say: "rpm -ivh http://{server/path/kernel*"
You have to supply the full filename.
This means the user would have to obtain a directory listing and copy it
to the floppy, transfer it, etc. It was getting too complicated.
> Link to the GPL here, I think it's http://www.gnu.org/copyleft/ , but
> I'm not sure.
I'll add that, thank you.
I did include a file called COPYING, with the text of the GPL in the
directory with the scripts.
- -D
- --
pgp key: http://www.tuxfan.homeip.net:8080/pgpkey.txt
Red Hat Linux 7.2 in 8M of RAM: http://www.freesoftware.fsf.org/rule/
- --
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: For info see http://www.gnupg.org
iD8DBQE8onX0eMAUbzJhSVcRAu/6AJ9Yb1RDMXIGLdRWaOM/9XECJiskJACbBPiZ
xgk6W8I8ycNYGJPRwseeoqE=
=R4PR
-----END PGP SIGNATURE-----
_______________________________________________
Rule Project HOME PAGE: http://www.freesoftware.fsf.org/rule/
Rule Development Site: http://savannah.gnu.org/projects/rule/ Rule-list at mail.freesoftware.fsf.org
http://mail.freesoftware.fsf.org/mailman/listinfo/rule-list
This full static mirror of the Run Up to Date Linux Everywhere Project mailing list, originally hosted at http://lists.hellug.gr/mailman/listinfo/rule-list, is kept online by Free Software popularizer, researcher and trainer Marco Fioretti. To know how you can support this archive, and Marco's work in general, please click here