Re: Introducing live-xmaker
Hi On Fri, Aug 22, 2008 at 07:57:59PM +0200, Simon Schiele wrote: And of course because code says more than thousand words there is the function xml.sh, the helper lh_xml (for im- and export) and an example config.xml attached to this mail. a few details (but open for every kind of change/rewrite): - export will only write non-default values - only depends on hexdump (busybox version works too) - fast (enough) - nothing implimented for binary yet (but there is uuencode in busybox, too) I missed xml.ah before. So you're not only reinventing your own format, but also your own XML parser? This will work on the XML you emit. So you're not really working with XML. You're working with the specific format that your exporter has generated, which happens to be a subset of XML. This will break when you try to use other tools to process your XML data. (I also use this option as a last resort elsewhere. But there the XML was forced on me and I did my best not to use it, because it only made life more complicated. Using it from start with this known limitation is again not the best design in the world.) -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: Introducing live-xmaker
On Sat, Aug 23, 2008 at 10:54:57AM +0100, Brendan Sleight wrote: Hi Tzafrir, 2008/8/23 Tzafrir Cohen wrote: This will break when you try to use other tools to process your XML data. Works for me with other tools, see below. [EMAIL PROTECTED]:~/debian-live$ python Python 2.5.2 (r252:60911, May 28 2008, 08:35:32) [GCC 4.2.4 (Debian 4.2.4-1)] on linux2 Type help, copyright, credits or license for more information. from amara import binderytools template = binderytools.bind_file('config.xml') print template.live_helper.common.lh_cache_packages False print template.live_helper.chroot.lh_language de exit() I do not think anyone is forcing people to use XML, but it may be useful as an option to some people who like XML. Kindly answer to what I actually wrote. He wrote his own XML parser in xml.sh. The XML config you generate in your python script might work with it. Or not. Likewise for any other XML library and generator out there. Maybe using XML is a good idea. But this implementation fails to address a number of substantial issues I have raised. -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: Introducing live-xmaker
lh_bootstrap lh_chroot lh_binary fi ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: Introducing live-xmaker
On Wed, Aug 20, 2008 at 11:40:03PM +0100, Brendan Sleight wrote: Hi All, I put together a little python script, live-xmaker. [1][2][3]. Live-xmaker is a command line front-end to live-helper, inspired by morphix-mmaker[4][5] (written by Alex de Landgraaf). This allows all the configurations for a liveCD build to be specified in one XML file. For example:- live-xmaker --xml /usr/share/live-xmaker/templates/lightgui_all.xml --output ./lightgui.iso Why the duplicate configration? lh_config generates a bunch of config files. By creating your own configuration file you're merely creating an extra layer to get information through. What does thing extra layer give you? Can you try to parse the information from the config directory? Maybe thr right way to do that is through the shell. See the following perl code that uses 'env -i sh -ec' to get a shell process to parse shell variables. http://svn.digium.com/svn/zaptel/branches/1.4/kernel/xpp/utils/zconf/Zaptel/Config/Defaults.pm A sample program that uses it: http://svn.digium.com/svn/zaptel/branches/1.4/kernel/xpp/utils/zapconf -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: problem for building a live cd
On Fri, Jul 04, 2008 at 03:11:44PM +0200, Marco Bardelli wrote: Hi at all, i have a strange problem to build a live cd. I'm behind a proxy which i setted in apt.conf the Acquire::http_proxy statement, and i have any problem to install pkgs via apt-get. When i run lh_build the output is: P: Begin caching bootstrap stage... P: Begin bootstrapping system... P: Retrieving Release E: Couldn't download Release! debootstrap failed to download even the Relase fie of some apt source. P: Begin caching bootstrap stage... P: Begin caching chroot stage... P: Begin mounting /proc... E: lh_chroot_proc missing That's because debootstrap did not install anythng later on. Is a problem solvable via a particular configuration in the config directory or is a more hard problem ?? Bad mirror (or some strange networking error)? Debootstrap has failed. -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: Database applications
On Wed, May 28, 2008 at 11:21:38AM +0100, Colin Turner wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hi Tzafir, Tzafrir Cohen wrote: (yet anyway)), live-helper works away solidly, but then crashes in postinst when the database would ordinarily be created since mysql is installed, but not running in the chroot. I don't think that debian-live is the only scenario where a package is installed in a chroot. No, I'm sure it isn't, but maybe it's the non-running mysql that's the problem. Is there any accepted way around this? I have thought of special init scripts and special packages, but I wanted to ask if there was a known workaround for this problem first. What if mysql is already running on the main system? It is / was. I think the issue was that the socket path was pointing to the /var/run/ directory in the chroot where there was no running server. This is somewhat good: if someone tried to connecto to port 3306 and assumed this this allowed setting up data in the for the package the result would be a silent failure, which is even worse. Anyway, any chance that application can work with an sqlite database? Hmm, I don't know, I've never played with sqlite, but looking at it briefly, I think that would cause more work since I'd have to do a lot of testing on a new backend to ensure the application still worked correctly. One workaround I wondered about was amending the postinst script to check if it was a live-helper install, and if so to not call the database configuration sections. I could then execute an init script that performed either a dpkg-reconfigure on the packages or directly called a script to import the SQL. Calling dpkg-reconfigure from an init script? What if dpkg would call that init script through some other call flow? What about purging the data? You can no longer rely on the files of the package to be present for that. Does that sound credible, and if so is there a reliable way of detecting that I'm in the live-helper build environment? What other legitimate reasons are there for mysql-server not to be running? * Package is being test-installed in a pbuilder environment :-) -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: Database applications
On Mon, May 19, 2008 at 10:39:16AM +0100, Colin Turner wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hi Everyone, I am trying to make a live-cd to demo some web apps we are creating at our university. When I add my two extra .deb files (not in the archive (yet anyway)), live-helper works away solidly, but then crashes in postinst when the database would ordinarily be created since mysql is installed, but not running in the chroot. I don't think that debian-live is the only scenario where a package is installed in a chroot. Is there any accepted way around this? I have thought of special init scripts and special packages, but I wanted to ask if there was a known workaround for this problem first. What if mysql is already running on the main system? Anyway, any chance that application can work with an sqlite database? -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: The Debian Live manual
On Sun, May 18, 2008 at 07:48:50PM +0100, Kai Hendry wrote: I could imagine contributing more if it wasn't XML. http://natalian.org/archives/2008/05/15/10-years-of-xml/ Why not markdown and a front end like ikiwiki? Personally I prefer asciidoc. But I didn't start that document so I don't get to dictate that. I don't contribute much to http://wiki.debian.org/DebianLive/ because I hate editing anything without vim git. That docuument is available for you to edit with vim and git. You, I know that the syntax is sub-optimal. -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: [RFC] The future of LH_BINARY_IMAGES
On Thu, May 15, 2008 at 01:10:11PM -0300, Otavio Salvador wrote: Chris Lamb [EMAIL PROTECTED] writes: If no-one has any major objections or queries I'll go ahead and cleanup the code to assume that LH_BINARY_IMAGES contains a single value. I do believe that we shouldn't try to support all kinds of use in a tool and instead support the simple one. The more complicated use scenarios can be done using a layer above the tool (as the code you've shown that simulates the multiple image support). From my POV it ought to be cleaned up and it'll reduce/simplify a bit of code. I posted a while ago a trivial wrapper script to create multiple binary outputs. The downside is that you have to run lh_clean --binary in between. How could I avoid generating the squashfs multiple times? I believe that the squashfs should be the same with multipl formats. Its generation probably takes most of the time of the binary stage. -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: Add live CD user to fuse group?
On Sat, May 10, 2008 at 08:02:55PM -0500, maybeway36 wrote: Is there a way to specify what extra groups the live CD user can be part of using lh_config? You can always add this in a hook script: cat EOF config/chroot_local-hooks/group_fuse #! /bin/sh # Add live CD user to group 'fuse' adduser user fuse EOF chmod +x config/chroot_local-hooks/group_fuse -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: running livecd from a hd without installing it
Hi It's kind of frustrating when you keep ignoring suggestions that are not exactly the way a specific live distribution behave. On Sun, May 11, 2008 at 12:01:24AM +0200, Ricardo Gabriel Herdt wrote: I searched for some info about using grub as the liveCD boot-loader Have you looked into setting 'LH_BOOTLOADER=grub' in config/common or config/binary ? and heard it's not so efficient in some systems. What do you mean by efficient? Well, it is better than nothing. Frankly I don't think that breaking the initrd of live-initramfs to smaller pieces is that posssible. And breaking to modules after that is basically impossibe without actually chanign Debian. (And you can still use a USB storage device) -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: running livecd from a hd without installing it
On Sat, May 10, 2008 at 12:14:39AM +0200, Ricardo Gabriel Herdt wrote: Hi all, Dyne:bolic has a feature called docking where the user is able to simply copy the system directory from the CD to a partition (what can be done from inside a windows system too) and boot using it instead the cd content. I find that awesome, since I can run my system everywhere and with high speed, not having to work on a system running from a cd. I also don't have to change the installed system, after using linux, I just erase the dyne directory and leave. The other cool thing is the so called nesting, where the system automatically recognizes a existent *.nst file that stores the users data and config files (/etc content for example) and loads it. Is there a way to do those things with a DebianLiveCD? If not, I really think that is something to be thought about, than we all know how liveCDs are much more useful than simple show-linux-to-your-friends systems. Have you tried simply copying the contents of the CD somewheer and using an alternative boot loader? -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: Default user decisions
On Fri, May 02, 2008 at 10:19:11AM +1000, Trent W. Buck wrote: On Thu, May 01, 2008 at 05:39:07PM +0300, Tzafrir Cohen wrote: On Thu, May 01, 2008 at 09:48:45AM -0400, John Reese wrote: Marco Amadori wrote: ssh wise, Ubuntu's choice is more secure, because it disallows ssh logins if the local console user did not provide a new password. I think that using a NULL password like ubuntu do and providing both an interactive way to change it and a boot parameter could be the way I would like to have the user password managed. That way we could have a more secure default image approach, a secure personal use approach and the ability to set a password easely at build time. I have to agree with this. I really like the Ubuntu approach to securing the root/default users, and I'd like to put my support behind making this behavior the preferred method. A user has to install ssh explicitly, anyway. But what happens when that secure user installs a service that doesn't care about empty passwords? What kind of user? - An end user running the default Debian Live system; changing the /cow which will be lost on boot. - A similar end user running a customized Live system created by an (intermediary) lh user; or - An (intermediary) lh user, creating a Live system with on-by-default attackable services like ssh and httpd? That user has installed them and should know how to secure them. Almost all httpd provide only read-only information by default. It seems to me that these cases are fundamentally different, and should be considered separately. Fine. And just to remind you that openssh-server is not installed by default. Now that we mentioned httpd: here are some useful bits from my configuration: My apache hook: Usually I have a separate hook for each task. But this one gre a bit larger than usual: Begin config/chroot_local-hooks/apache #!/bin/sh # generate a certificate for apache. # FIXME: This one should really be done at runtime. Anyone can know the # secret key of this server. And I have no hope of using the real host # name before boot. # # But until I figure how to run things on runtime, let's generate a # static certificate: #export RANDFILE=/dev/random; #mkdir /etc/apache2/ssl # #openssl req $@ -new -x509 -days 365 -nodes -out /etc/apache2/ssl/apache.pem \ # -keyout /etc/apache2/ssl/apache.pem # [The snippet above has been replaced by a single certificate I have # generated once. Needless to say you should not consider this secure. # Thus you cannot really protect from a man-in-the-middle attack. But # what would you expect from a live CD? :-) ] # the default apache site has a NameVirtualHost for all ports. # I want to add a different one for port 443. And I don't need it # anyway: a2dissite default a2ensite default80 # Enable ssl: a2enmod ssl # Enable mod_proxy for ajaxterm: a2enmod proxy_http # Make sure apache is started by default: sed -i 's/^NO_START=.*/NO_START=0/' /etc/default/apache2 End config/chroot_local-hooks/apache As you can see, apache does not run by default. Apache is also takes some effort to create an SSL site. Here is something the reduces security a bit as it allows remote users to get an information about installed software), but then again, what do I have to hide? ;-) On the other hand it exposes all the great documentation (and has encourged me to contribute more documentation and make it available through debian-doc). Begin config/chroot_local-hooks/dwww_permit #!/bin/sh # allow anyone to browse the documentation on the CD: sed -i '/allow/s/from .*/from all/' /etc/apache2/conf.d/dpkg-www End config/chroot_local-hooks/dwww_permit I also have some useful documentation on the CD. Thus I want to expose it as well: # Begin config/chroot_local-includes/etc/apache2/conf.d/live_media.conf # Make the content of the CD available for apache under # http://server/media Alias /media /live_media Directory /live_media Options Indexes MultiViews /Directory # End config/chroot_local-includes/etc/apache2/conf.d/live_media.conf I also permit access through ajaxterm. This is why I had to enable SSL. Though ajaxterm seems to be quite dead and maybe it is time to package anyterm. The configuration I use is basically the one included in the current package in Lenny. Allow web-based access to the IRC support channel. Sadly nobody really uses this. Uses the package cgiirc: ## Begin config/chroot_local-includes/etc/cgiirc/cgiirc.config # CGI:IRC configuration file. # # Check /usr/share/doc/cgiirc/examples/cgiirc.config.full.gz # for more details. # Take care about applying debian-specific settings like # `image_path' if you intend to just copy it! default_server = irc.freenode.net default_port = 6667 default_channel = #my-channel default_name
Re: Default user decisions
On Thu, May 01, 2008 at 09:48:45AM -0400, John Reese wrote: Marco Amadori wrote: Alle Sunday 27 April 2008, Daniel Baumann ha scritto: The only other live system I know remotely, uses username 'ubuntu' and no password. That made me curious what others think. So, what do *you* guys think? Should it be left as is? Or do you have other preferences? How do other live systems do it? ssh wise, Ubuntu's choice is more secure, because it disallows ssh logins if the local console user did not provide a new password. I think that using a NULL password like ubuntu do and providing both an interactive way to change it and a boot parameter could be the way I would like to have the user password managed. That way we could have a more secure default image approach, a secure personal use approach and the ability to set a password easely at build time. I have to agree with this. I really like the Ubuntu approach to securing the root/default users, and I'd like to put my support behind making this behavior the preferred method. A user has to install ssh explicitly, anyway. But what happens when that secure user installs a service that doesn't care about empty passwords? -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: Default user decisions
On Mon, Apr 28, 2008 at 09:15:33AM +1000, Trent W. Buck wrote: On Sun, Apr 27, 2008 at 03:00:34PM +0300, Tzafrir Cohen wrote: Instead of having a static, predictable, easy-to-crack password, I would suggest taking these steps: Here you assume that someone will actually bother to take action with a live CD. Users expect it to Just Work[tm]. I'm advocating a negligible amount of extra work for the live-helper user, not the *end* user. (Last time I checked, this is merely a matter of whether 13home and a couple of other scripts are present in live-initramfs.) - possibly, prompt for confirmation at build time if BOTH 1) the guest user is enabled; AND 2) any blacklisted packages (e.g. openssh-server) are installed. Something like openssh-server is to be installed, but the insecure guest user is enabled, with a predictable username and password. Do you accept this gaping security hole? There is a pretty good chance that the user will not be at the console more than necessary[*]. Such a propmt will needlessly stall the boot (recall you have to do it before sshd starts) As before, I am talking about live-helper users (you), not end users (your customers). So please re-read my mail and tell me where do you think extra work can help. As I have already mentioned, none of the proposals for extra work for me actually helps the end user. I want to just be able to boot the CD and access the system remotely. In such a case there is simply nothing that sets apart a legitimate user from one that isn't. I also expect a typical system to be up for a pretty short time, and hence the impact of a malicious take-over is significantly reduced. As a rule, asking the user is something I hope to avoid with the live CD. Normally such solutions are just not applicable, and the default have to work. I'm not sure if the third point is worthwhile, since various network layouts make different packages worthy of blacklisting. That is, the blacklist is bound to have a bunch of false positives and negatives. [*] As for more than necessary - what does it take to boot to the CD automatically after a timeout of, say, 60 seconds? -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: Rescue flavor improvements
On Sun, Apr 27, 2008 at 08:27:33PM -0300, Tiago Bortoletto Vaz wrote: On Sun, Apr 27, 2008 at 11:12:05PM +0100, Alex Owen wrote: Can we AND and OR these conditions. I guess OR is like this: #if ARCHITECTURE amd64 #include stuff #endif #if MODE ubuntu #include stuff #endif This works. Does nesting the if's work too? to AND the conditions like this: #if ARCHITECTURE amd64 #if MODE ubuntu some-ubuntu-amd46-package #endif #endif This doesn't work (yet). Maybe Chris is working on this, not sure. What exactly is the proposed syntax? cpp's syntax does not allow string values to variables. And if we don't want to use cpp's syntax, we might as well using '#' for anything other than a comment. Also: should it be possible to use any environment variable, or just one from a pre-selected list? -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: Default user decisions
Hi On Mon, Apr 28, 2008 at 09:51:02AM +0200, Maarten ter Huurne wrote: On Monday 28 April 2008, Tzafrir Cohen wrote: I also expect a typical system to be up for a pretty short time, and hence the impact of a malicious take-over is significantly reduced. If the system has permanent storage like a harddisk or flash stick, a take-over could last longer than the uptime. If sensitive data is captured, the damage is done regardless of how long the system stays up after that. Also, what is a short time for a human attacking the system is ages for an automated system. If the system is a USB stick, changing the default on a per-copy basis is relatively simple. If the system is an ISO image, it is far too dificult for anyone to bother. It's nice to say that defualt passwords are dangerous. But this is how it's done with practically any router box and such, where local access does not really exist. ssh is not installed by default. If you install it, you need to secure it. Someone wrote here before that using a default password gives a false sense of security, as you might think it is better than an empty password (for some scenarios it does help, but it is indeed meerly obfuscation with no real extra layer of security to support it). However, now that you feel safe with the empty password of that user, that is disabled, you go and install an imapd on your box. Suddenly anybody in the network *can* use your disabled for remote operation user account. -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: LH_BINARY_IMAGES was: Rescue flavor improvements
On Mon, Apr 28, 2008 at 11:09:37AM +0100, Alex Owen wrote: The obvious example is LH_BINARY_IMAGES,which can contain permutations of iso, usb-hdd and net. So if I find a place in the code where LH_BINARY_IMAGES is expected to be single valued (ie not a list of image types) then that is a bug? This is just a theoretical question, right? -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: Default user decisions
On Sun, Apr 27, 2008 at 03:00:34PM +0300, Tzafrir Cohen wrote: [*] As for more than necessary - what does it take to boot to the CD automatically after a timeout of, say, 60 seconds? config/binary: LH_SYSLINUX_TIMEOUT=20 Works well. -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: LH_BINARY_IMAGES was: Rescue flavor improvements
On Mon, Apr 28, 2008 at 12:55:36PM +0200, Daniel Baumann wrote: Tzafrir Cohen wrote: This is just a theoretical question, right? in theory, every variable whose name is grammatically a plural should accept multiple values (in form of a spaces seperated list). in practise, for most variables above, this is true. unfortunately, LH_BINARY_IMAGES is not one of them. it would need a bit of work to re-arrange some stuff in order to be able to build different images types in one shot, but ensuring that the chroot is not tainted by the different auxillary packages that are required for the different binary targets (e.g. dosfstutils, syslinux, genisoimage etc.). In the mean time: LOG_FILE=$build_root/log # Clean most things. Do keep cache. lh_clean --all 21 | tee $LOG_FILE mkdir -p tmp/ # Let's get to work: lh_build 21 | tee -a $LOG_FILE mv binary.iso tmp/ # build hdd-usb: lh_clean --binary 21 | tee -a $LOG_FILE LH_BINARY_IMAGES=usb-hdd lh_build 21 | tee -a $LOG_FILE #mv binary.img tmp/ # build netboot tarball: lh_clean --binary 21 | tee -a $LOG_FILE LH_BINARY_IMAGES=net lh_build 21 | tee -a $LOG_FILE #mv binary-net.tar.gz tmp/ mv tmp/* . -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: Default user decisions
On Sun, Apr 27, 2008 at 11:09:06AM +0200, Daniel Baumann wrote: Hi, some days ago, somone on irc asked about thoughs of having, by default, no password for the live user. Currently, we have the username 'user' with password 'live', which I *personally* consider to be a reasonable choice. The only other live system I know remotely, uses username 'ubuntu' and no password. That made me curious what others think. So, what do *you* guys think? Should it be left as is? Or do you have other preferences? How do other live systems do it? Password should not be required for that user in {g|k}dm (e.g: in case there is a need to logout and re-login). But disabling password elsewhere is probably not a good idea. The console login uses the login program which may be used by some other programs for remote login. There are telnet ans rsh/rlogin that nobody really uses (right?) Various other optional packages (e.g: openssh-server) will allow remote login. I suppose it is a bad idea to make it password-less. -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: installing debian-live for etch?
On Wed, Apr 23, 2008 at 11:36:32AM -0500, Josh Lawrence wrote: On Wed, Apr 23, 2008 at 11:28 AM, Daniel Baumann [EMAIL PROTECTED] wrote: Debian Live is nothing you can install, it's not a distribution. It is a 100% Debian system with a tweak to run from a read-only medium (e.g. cdrom). If you want to install Debian, use a regular Debian installation medium. Daniel, You're exactly right, sorry I wasn't more clear before...what I'm looking for is a way to use the live-helper scripts on an Etch box. I know that live-helper exists in the Lenny repos, so I did some Googling on the proper way to keep a mixed system; I added the Lenny repos to the box, specified Etch as my preferred repo in /etc/apt/apt.conf, and I was able to get the live-helper scripts. Does this sound like a reasonable approach to take? You don't need Lenny repos on your build server. You just need live-helper and preferably also live-initramfs . You can get backports of those packages, or build them yourself. (I'm using live-helper from git mostly directly and casper instead of live-initramfs and my target is an Etch system. I have a simple wrapper script to add the helpers directory to the PATH) -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: running a command with sudo
On Tue, Apr 22, 2008 at 12:55:17AM +0100, Chris Lamb wrote: Alex Owen wrote: The above should work... but read visudo man page... This is all totally pointless. Just append your line to /etc/sudoers in a live-initramfs hook: this is what it does itself, so it's difficult to see why you could have any qualms about doing so too. To quote a previous message of mine: But that's a really technical issue. My main issue is with the fact that my package needs to edit /etc/sudors . Works? sure. Elegant? no. Will work for the live CD. Will fail later in real life. That is: any better Debian-blessed way for a package to edit /etc/sudoers ? I prefer to put my changes in packages and not in one-time scripts, as this makes them also usable outside of the specific environment of debian-live. And the whole point of Debian-Live is to use unmodified Debian. -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Bug#476204: toboot in isolinux prompt and other problems
On Mon, Apr 14, 2008 at 11:21:36PM -0400, Joey Hess wrote: f2.txt has a newline after the word to, so boot: is written on the next line The immediate offender is: diff --git a/helpers/lh_binary_syslinux b/helpers/lh_binary_syslinux index 8b8fc3f..da3952b 100755 --- a/helpers/lh_binary_syslinux +++ b/helpers/lh_binary_syslinux @@ -270,7 +270,6 @@ Configure_syslinux_templates () -e s|{\$LH_MEDIA}|${MEDIA}| \ -e s|{\$LH_VERSION}|${VERSION}| \ -e s|{\$LH_SYSLINUX_TIMEOUT}|${LH_SYSLINUX_TIMEOUT}0| \ - -e 's|\ $||g' \ $template_file done f2.txt lacks the text F2 in the corner, all the other screens have their number there f2.txt could move the URL to the website down to the next line since the full url doesn't fit on the line it's on Reminder: in vim, make sure you use noeol. I with those files: :set noeol bin|w instead of simply saving, just to be sure. f4.txt says F3 in the corner And F8 is intentionally left blank? -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: Cross packaging issues with module-assistant
On Thu, Apr 10, 2008 at 07:22:47PM +1000, Trent W. Buck wrote: On Thu, Apr 10, 2008 at 11:15:36AM +0200, Giorgio Pioda wrote: I'm trying to compile extra modules inside the lh chroot. Using module-assistant leads systematically to wrong linking with wrong headers. For example I'm working in a 486 chroot inside a lenny-amd64 host and I get downloaded the amd64 headers instead of the desired 486. Are you calling m-a with the -r flag to select the appropriate kernel? By default it uses the RUNNING kernel, which is the one outside the chroot (i.e. wrong). If you have sh code to find what argument to pass to m-a within the chroot, I'm interested; currently I'm just hard-coding it in my call to m-a. I have a dummy package that build-requires a bunch of linux-header packages and the relevant -source package, and then runs: m-a -u . -i -t -f \ -k `echo /usr/src/linux-headers-[1-9]*.*.*-*-* | grep -v '*'` \ build PACKAGENAME (and goes on to package the products in the dummy deb). I just feed it to pbuilder and get the modules for all the kernels I need. -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: Simple GUI
On Mon, Apr 07, 2008 at 11:11:05AM +0200, Erwan Le Gall wrote: Me again, One of our client has ordered a GUI to build Live-CD and Live UsbKey. To do that we have worked on a PyQT interface (cause our client use KDE and live-magic wasn't doing the job). Debian Live is the real engine, infact we have only make a graphical interface on it. At this moment, we have almost finished and the client will deliver the product in about 1 month (certainly GPLv3 or Cecill or both)... But during this time, I still have the sources and I can work on it to prepare the community deliver. The aim of this GUI is to be simple and usable by a standard adminsys without any doc read or complex things to know. The principals features are : -templates collection manager : each image is build with different configuration. Each configuration can be saved on a simple xml file. This file is a kind of Live specification to rebuild on another computer. Why an XML file? We already have a configuration format (a bunch of variables in pre-set locations). A tar / zip archive of those would work just as well. It would also allow saving additional debs and such. In fact, I simply dump the config directory on my CD to make it possible to rebuild it :-) -Packages manager : A simple distribution package manager (synaptic like), very simple local packages manager. -Options manager : *Lang/keyboard (At this moment only french and Eng are well managed... But it's very easy to change that). *Hostname/username *Live type : USB or iso *Bootsplash Right. And show that image with the right number of colors... *Usplash *Timeout *Encryption *Interactive mode (allow you to choose somes options like lang at the Live start) *Partitions automount (Off, automount, forensic, Interactive) *DHCP (Off/on/interactive) *Optional Option : capabilities to add option to lh_config *Hooks scripts manager -Build screen (whit a summary and a follow of lh_build output) -Image manager : *Test the image with qemu directly into the application *Copy the image on a CD/DVD or usbkey directly by the same application *Test the CD/DVD or USBKey with qemu directly with the application Is a USB image the same as a CD image? If not, what do you actually mean here? -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: liveCD for audio work
On Wed, Mar 12, 2008 at 11:58:20PM +0100, Daniel Baumann wrote: chymian wrote: to build squashfs aufs modules you can use module-assistant if you want to do it properly (aka 'the debian way'), you rebuild a debian kernel tree (completely), and build matching linux-modules-extra-2.6 (with squashfs/aufs/unionfs only) for it. Building packages with m-a is just as good, and quite simple. You can easily feed m-a multiple for kernels (using -l). Just be sure to use -t -i -f . -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: lh_build devpts-live already mounted
On Thu, Feb 14, 2008 at 09:46:10AM -0500, Thomas H. George wrote: lh_build ends with a message mount: devpts-live is already mounted or chroot/dev/pts is busy I found this problem discussed in the January archive as bug 459697. I still have the problem. Is there a solution? Hmmm... what version were you useing, again? -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: predefine interfaces by MAC with udev?
On Thu, Feb 07, 2008 at 12:00:57PM +0100, Chris Fanning wrote: Hi, We're trying to boot diskless servers that have two network interfaces. We boot from one interface (dhcpd/MAC), but when initramfs kicks in, udev (isn't it?) sets the interface names and then (often) the boots process fails becuase it's looking for the filesystem on the wrong interface. We can't use the MAC as an option for 'ifconfig' or 'ip add add', so we need to change something with udev, right? KERNEL==eth*, SYSFS{address}==00:12:34:fe:dc:ba, NAME=eth0 KERNEL==eth*, SYSFS{address}==00:56:78:98:76:54, NAME=eth1 I've looked around the scripts but nothing as jumped out at me. /etc/udev/rules.d/z25_persistent-net.rules is generated by /etc/udev/rules.d/z45_persistent-net-generator.rules It might provide you some inspiration. -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: setting up a couple of user accounts
On Wed, Dec 19, 2007 at 06:41:29AM -0700, John Schmidt wrote: Hi, I have some software that I would like installed on a LiveCD for demonstration purposes. The installation requires the addition of a couple of user accounts with the software in those accounts. Is there a way to add additional users? Using the old make-live, I would create a separate chroot and do my installation there, and then use the following comand: make-live --include-chroot /home/jas/debianliveSCIRun/ --hook=chown 999:99 9 /home/casper -R I looked through the exmple scripts for debian-live but didn't see anything, any points to FAQs or something I perhaps overlooked would be appreciated. Add hook scripts in config/chroot_local-hooks . Those scripts are run in the chroot itself after everything was installed. You can use something of the sort of: #!/bin/sh # add some users: adduser whatever --nopassword echo whatever:thepassword | chpasswd Alternatively, put a list of: 'username password' lines in config/chroot_local-includes/users_to_add and use a hook script of the sort of: #!/bin/sh # add some users: cat /users_to_add | while read user passowrd; do adduser user --nopassword echo user:password | chpasswd done rm /users_to_add ## The above are completely untested. -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
etch problems: auto LH_INITRAMFS, \${LH_UNION_FILESYSTEM}, lh_clean
Hi I started working with debian-live on Etch again, and as usual wasted a number of days just to get it built properly. Bugs encontered: 1. lh_build can run wild and (if left alone) force you to reboot the build system due to over-swapping. Fix: handle lh_build --usage. 2. If LH_INITRAMFS is not set in the configuration files, the build will fail (somewhere at lh_binary, due to DESTDIR being undefined). 3. Unionfs modules would not install and thus the unionfs mount fails. --- a/functions/defaults.sh +++ b/functions/defaults.sh @@ -407,7 +407,7 @@ Set_defaults () # Set linux packages if [ -z ${LH_LINUX_PACKAGES} ] then - LH_LINUX_PACKAGES=linux-image-2.6 \${LH_UNION_FILESYSTEM}-modul es-2.6 + LH_LINUX_PACKAGES=linux-image-2.6 ${LH_UNION_FILESYSTEM}-module s-2.6 if [ ${LH_CHROOT_FILESYSTEM} = squashfs ] then Note that defaults.sh guarantees somewhere above that part that LH_UNION_FILESYSTEM is set. Is the above a hook to override it later? 4. (minor issue) All the text screens of syslinux are now terminated with a newline. This seems to break the original intention of the text there, that uses the word Boot: of syslinux as part of the sentence. The one of f3 has an extra line. -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: lh_build fails in: invoke-rc.d: initscript mysql, action start failed.
On Wed, Dec 05, 2007 at 08:25:42AM +0100, René Mérou wrote: Is it a problem of my config? Any chance you have mysql server running on your system and the mysql server on the chroot fails to bind to port 3306? -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: lh_clean all launch a big number of processes
On Sat, Nov 24, 2007 at 06:00:51PM +0100, René Mérou wrote: I remember 5730 process and i was not able to refresh my descktop windows. Usuall, now, i have 204 proc at gkrem. I started 2 times lh_clean all and the 2 times i was forced to turn off my computer or restart X windows. Ugly quick-fix: index 771b852..c7ffb59 100755 --- a/helpers/lh_clean +++ b/helpers/lh_clean @@ -118,7 +118,9 @@ do ;; *) - ${0} --usage + #${0} --usage + echo TODO: here should be usage message of $0: + echo == $USAGE === ;; esac done ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel -- Tzafrir Cohen icq#16849755 jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: d-l and quieting $APT
On Wed, Sep 19, 2007 at 10:35:39AM -0400, Justin Pryzby wrote: Hi Everyone, Is there any way to make apt/aptitude quiet? d-l does lots of [de]installations and apt/dpkg accounts for almost all of the output remaining after setting LH_QUIET=enabled. An alternative is to minimize the number of calls to $apt, which is also good for performance. I mean that a single apt call should install all the packages in the chroot and probably never remove them. Any reason to not do this? Multiple calls give a different resolution of dependencies. Generally will get more packages. -- Tzafrir Cohen icq#16849755jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
[patch] syslinux screen fixes
Hi I attach fixes to the syslinux startup screens. I only fixed the English live ones, as those were the ones I could test. Changes: * Removing end-of-line in the end, to keep the word boot: in the same line * And thus adding an extra line to all of them (except boot.txt itself). * s/F3/F4/ in f4.txt . Problems: I still can't place F2 in the proper place on the screen. No idea why. Adding an extra space makes it slide of the screen. -- Tzafrir Cohen icq#16849755jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir diff --git a/templates/syslinux/en/boot.txt b/templates/syslinux/en/boot.txt index 56dffca..2fe3ead 100644 --- a/templates/syslinux/en/boot.txt +++ b/templates/syslinux/en/boot.txt @@ -1,3 +1,3 @@ splash.rle -Press F1control and F then 1 for help, or ENTER to +Press F1control and F then 1 for help, or ENTER to \ No newline at end of file diff --git a/templates/syslinux/en/f1.txt.live b/templates/syslinux/en/f1.txt.live index f6f33ef..ff132ae 100644 --- a/templates/syslinux/en/f1.txt.live +++ b/templates/syslinux/en/f1.txt.live @@ -22,4 +22,5 @@ For F1-F9 type control and F then the digit 1-9 For F10 type control and F then the digit 0 -Press F2 through F10 for details, or ENTER to + +Press F2 through F10 for details, or ENTER to \ No newline at end of file diff --git a/templates/syslinux/en/f10.txt b/templates/syslinux/en/f10.txt index 3cf379e..8652bcf 100644 --- a/templates/syslinux/en/f10.txt +++ b/templates/syslinux/en/f10.txt @@ -20,4 +20,5 @@ More information about the Debian Live project can be found at -Press F1control and F then 1 for the help index, or ENTER to + +Press F1control and F then 1 for the help index, or ENTER to \ No newline at end of file diff --git a/templates/syslinux/en/f2.txt.live b/templates/syslinux/en/f2.txt.live index f79062a..8532c24 100644 --- a/templates/syslinux/en/f2.txt.live +++ b/templates/syslinux/en/f2.txt.live @@ -1,9 +1,10 @@ -0fPREREQUISITES FOR RUNNING DEBIAN07 09F207 +0fPREREQUISITES FOR RUNNING DEBIAN07 09F207 You must have at least 64 megabytes of RAM to run this Debian Live system. See the Live Manual or the FAQ for more information; both -documents are available at the Debian Live web site, 0fhttp://debian-live.alioth.debian.org/07 +documents are available at the Debian Live web site, +0fhttp://debian-live.alioth.debian.org/07 Thank you for choosing Debian! @@ -20,4 +21,5 @@ Thank you for choosing Debian! -Press F1control and F then 1 for the help index, or ENTER to + +Press F1control and F then 1 for the help index, or ENTER to \ No newline at end of file diff --git a/templates/syslinux/en/f3.txt.live b/templates/syslinux/en/f3.txt.live index b48ab8d..1b90fe6 100644 --- a/templates/syslinux/en/f3.txt.live +++ b/templates/syslinux/en/f3.txt.live @@ -13,6 +13,7 @@ + To use one of these boot methods, type it at the prompt, optionally followed by any boot parameters. For example: boot: live acpi=off @@ -20,4 +21,4 @@ followed by any boot parameters. For example: If unsure, you should use the default boot method, with no special parameters, by simply pressing enter at the boot prompt. -Press F1control and F then 1 for the help index, or ENTER to +Press F1control and F then 1 for the help index, or ENTER to \ No newline at end of file diff --git a/templates/syslinux/en/f4.txt.live b/templates/syslinux/en/f4.txt.live index 6a76aca..be53488 100644 --- a/templates/syslinux/en/f4.txt.live +++ b/templates/syslinux/en/f4.txt.live @@ -1,4 +1,4 @@ -0fADDITIONAL BOOT METHODS07 09F307 +0fADDITIONAL BOOT METHODS07 09F407 0fAvailable boot methods:07 @@ -20,4 +20,5 @@ -Press F1control and F then 1 for the help index, or ENTER to + +Press F1control and F then 1 for the help index, or ENTER to \ No newline at end of file diff --git a/templates/syslinux/en/f5.txt b/templates/syslinux/en/f5.txt index 58bbb94..7d01b24 100644 --- a/templates/syslinux/en/f5.txt +++ b/templates/syslinux/en/f5.txt @@ -20,4 +20,5 @@ module.param=value, for example: libata.atapi_enabled=1 -Press F1control and F then 1 for the help index, or ENTER to + +Press F1control and F then 1 for the help index, or ENTER to \ No newline at end of file diff --git a/templates/syslinux/en/f6.txt b/templates/syslinux/en/f6.txt index 772fd50..0552aba 100644 --- a/templates/syslinux/en/f6.txt +++ b/templates/syslinux/en/f6.txt @@ -17,7 +17,8 @@ disable buggy APIC interrupt routing 0fnoapic nolapic07 (partly) disable ACPI 0facpi=noirq07 or 0facpi=off07 disable USB0fnousb07 + For example
Re: RAM usage
On Sat, Sep 15, 2007 at 03:32:12PM +0200, Daniel Baumann wrote: Ben Armstrong wrote: Aha! Yet another reason etch is a bad default that has nothing to do with etch, but with the fact, that the 486 kernel flavour doesn't support highmen. this is also the case in lenny. and yes, the 486 kernel should be the default flavour; not just because its the default d-i flavour, but because it runs on hardware that the 686 doesn't support; otoh, the 486 supports all hardware that the 686 does, but only some of them not optimally (such as machines with more than 1gb ram). It also has SMP disabled. So you support up to oine CPU and 1GB of memory. But you do support a bunch of extra CPUs. Yay! The 686 works on Intel ppro Pentuim II and better, AMD K7 and better, some of the Vias (not sure exactly which). On my CD I have: $ cat config/chroot LH_LINUX_FLAVOURS=686 -- Tzafrir Cohen icq#16849755jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
[patch] incorrect default value for LH_INITRAMFS
When a value for LH_INITRAMFS is not set, defaults.sh will set it to 'auto' rather than setting it to something sensible (as is when it is set explicitly to 'auto'. -- Tzafrir Cohen icq#16849755jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir diff --git a/functions/defaults.sh b/functions/defaults.sh index 05b1e0d..ae9c8a4 100755 --- a/functions/defaults.sh +++ b/functions/defaults.sh @@ -168,7 +168,7 @@ Set_defaults () if [ -z ${LH_INITRAMFS} ] then LH_INITRAMFS=auto - else + fi if [ ${LH_INITRAMFS} = auto ] then case ${LH_MODE} in @@ -190,7 +190,6 @@ Set_defaults () ;; esac fi - fi # Setting losetup if [ -z ${LH_LOSETUP} ] || [ ! -x ${LH_LOSETUP} ] ___ debian-live-devel mailing list debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: BOOT PENDRIVE .IMG SOLVED!
Hi Federico, A petty procedural post: On Sun, Jul 01, 2007 at 10:26:00AM -0300, Frederico Brandt wrote: Hi again, guys. I could mount the .img file and looking around syslinux.cfg file, just editing it removing the lines `/binary/` inside it. And so I could boot the image without problems. That`s all. Welcome to the list. We appreciate your contribution. I would appreciate it if you: 1. Please do not use capital letters for your subject. It is normally considered as SHOUTING. 2. When you follow-up on a message, please reply to it, so your follow-up message gets listed in the same thread (even if you change the subject). BTW: thanks for posting a summary message after your problem has been solved. Cheers, -- Tzafrir Cohen icq#16849755jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ Debian-live-devel mailing list Debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: custom kernel
On Wed, Jun 13, 2007 at 12:27:22PM +0400, Vladimir Stavrinov wrote: On Tue, Jun 12, 2007 at 11:03:44PM +0200, Daniel Baumann wrote: i do not care about kernels beeing not packaged in a .deb. I think about other way to perform this task: using config/chroot_local-includes putting there kernel and modules. Is it right? Will be it working? For a live CD? yes. probably. For an installer: no. -- Tzafrir Cohen icq#16849755jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ Debian-live-devel mailing list Debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: custom kernel
On Tue, Jun 12, 2007 at 05:11:51PM +0400, Vladimir Stavrinov wrote: Typical problem I am solving with custom kernel for years is installation system onto hardware not supported by Debian. In most cases this is disk (raid) controllers supported by drivers in binary form only (no open source) available for limited list of distributions like suse or redaht. Few years ago I have written the script, generating bootable iso image with custom kernel (e.g. taken from radhat) and as minimum running base system from debian, boot loader and some extra packages. With this image I install base system manually using debootstrap. All is OK, but whole system running from big ram disk, at least 128 MB in size. That is why I think how to adapt debian live for this task. Here is way I think about, but not testing yet: 1. Make package with custom kernel, e.g. kernel-2.6.22-custom.deb 2. Put kernel package to config/chroot_local-packages 3. Set configuration parameters: LIVE_LINUX_PACKAGES=kernel-2.6.22 LIVE_LINUX_FLAVOURS=custom LIVE_CHROOT_FILESYSTEM=plain Why do you need a plain (rather than squashfs) filesystem? Is it because your custom kernel does not support squashfs? 4. Build debian live as usual. -- Tzafrir Cohen icq#16849755jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ Debian-live-devel mailing list Debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: custom kernel
On Wed, Jun 13, 2007 at 12:04:22AM +0400, Vladimir Stavrinov wrote: On Tue, Jun 12, 2007 at 09:24:49PM +0200, Daniel Baumann wrote: i do not understand. you don't need to make kernel packages yourself, debian offers binary packages for the linux kernel. what /excately/ do you want? Daniel, please read whole this thread from beginning. If we are using debian kernel package, what about we talking? Custom kernel is not debian kernel nigher binary nor source. If they are an rpm package, look into alien to convert them into a deb package. Yes, it is far from perfect. It will likely have broken dependencies and such. But at least it will have basic accounting of files. And it is one file to carry arround, as opposed to a whole bunch. You may need to mkinitrd manually for that kernel. If you do, be sure to re-run update-grub. -- Tzafrir Cohen icq#16849755jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ Debian-live-devel mailing list Debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: customize?
On Mon, Jun 11, 2007 at 12:52:06PM +0100, [EMAIL PROTECTED] wrote: hello one time i have build the image , how can i customize the aparence of the image?. The image or the installed system? -- Tzafrir Cohen icq#16849755jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ Debian-live-devel mailing list Debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: How to customize isolinux.cfg/menu.lst?
On Tue, May 15, 2007 at 05:49:03PM +0200, Goran wrote: Hi all, is there a way to edit isolinux.cfg/menu.lst? I want to change the settings of default boot settings. With hook scripts I had no effect. binary_localincludes ? What changes are you really after? There are several different cases here: 1. Installed system (in this case it would also be nice to have a boot splash. not so tricky) 2. Live CD with isolinux 3. Live CD with grub. Which did you have in mind? -- Tzafrir Cohen icq#16849755jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ Debian-live-devel mailing list Debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: How to customize isolinux.cfg/menu.lst?
On Wed, May 16, 2007 at 09:49:42AM +0200, Goran wrote: Hi is there a way to edit isolinux.cfg/menu.lst? I want to change the settings of default boot settings. With hook scripts I had no effect. binary_localincludes ? What changes are you really after? I just want to change/delete some files from /isolinux. E.g. f4.txt should be deleted and noted in isolinux.cfg editet. Further It would be nice to add/change/delete the boot labels. You need to chnge files. You don't really need to delete f4.txt if noone looks at it. I blieve you can do with the extra 3 or 4 blocks. -- Tzafrir Cohen icq#16849755jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ Debian-live-devel mailing list Debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: Missing bits and pieces in make-live?
On Sun, May 06, 2007 at 12:55:03AM +0200, Daniel Baumann wrote: [EMAIL PROTECTED] wrote: - there is no documentation how to write what to the USB stick dd if=binary.img of=/dev/$your_device And when we have a non-Linux user? http://www.chrysocome.net/dd is rumoured to work. -- Tzafrir Cohen icq#16849755jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ Debian-live-devel mailing list Debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
Re: Install on HD button
On Thu, May 03, 2007 at 10:03:51PM +0200, Daniel Baumann wrote: Mathieu GELI wrote: Not at the moment AFAIK. debian-installer integration is a work in progress. well, debian-installer and 'install-to-hdd' are two different things. debian-installer installs the generic debian system on the harddisk. 'install-to-hdd' (let's call it live-installer) does install that very live system to the harddisk by formating and mounting a partition and uncompressing the squashfs image to it. debian-installer integration is nearly done, there is some fuckup with the repository population needed on the media. What does it take to add this to a CD? How does the user start a real installation? -- Tzafrir Cohen icq#16849755jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ Debian-live-devel mailing list Debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel
export includes from svn
Hi I'm using debian-live to build a live CD based on Debian-Live that aims at installing a fully-working Asterisk-based system. The CD itself could be found at http://updates.xorcom.com/ I thought it was a good time to start feeding my patches. Here's the first one: I maintain the config directory in a subversion repository. Thus when config/chroot_localincludes and config/binary_localincludes arecopied, the .svn directoies in them are copied as well. I decided to add an optional variable LIVE_EXPORT_COMMAND, and added the following as /usr/share/live-helper/functions/copy_local.sh : CopyLocalDir() { COPY_SRC=${1} COPY_DST=${2} # must be a relative path if [ ${LIVE_EXPORT_COMMAND} != '' ]; then COPY_TMP=cache/export rm -rf ${COPY_TMP} ${LIVE_EXPORT_COMMAND} ${COPY_SRC} ${COPY_TMP} COPY_SRC=${COPY_TMP} fi CURR_DIR=${PWD} (cd ${COPY_SRC}; find . | cpio -dmpu ${CURR_DIR}/${COPY_DST} ) } With respective changes in /usr/bin/lh_{chroot,binary}_localincludes . This should be useful to anybody who maintains the config directory under a version control system. I figure that for cvs, bzr, git, arch or whatever there is an equivalent syntax. This also allows the use of SVN keywords, such as $Id$ and $Date$ in the included content. -- Tzafrir Cohen icq#16849755jabber:[EMAIL PROTECTED] +972-50-7952406 mailto:[EMAIL PROTECTED] http://www.xorcom.com iax:[EMAIL PROTECTED]/tzafrir ___ Debian-live-devel mailing list Debian-live-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/debian-live-devel