Re: lilo about to be dropped?
Samuel Thibault samuel.thiba...@ens-lyon.org writes: Giacomo Catenazzi, le Wed 08 Apr 2009 19:47:55 +0200, a écrit : Samuel Thibault wrote: I installed grub (and Debian). Trying the Windows hidden partition (to install windows), grub stopped working (it was rescue mode, but without capability to rescue something). Also rescue disk + reconfiguring + update-grub did nothing. Err, did you re-run install-grub? No ;-) only update-grub and dpkg-reconfigure -plow grub-pc grub-common Then little wonder. update-grub only updates menu.lst. I was expecting that reconfigure will do the right things, grub maintainers considered that it's a bad thing to automatically reinstall things in a MBR. You need to re-run grub-install to do that. Is there some easy way to find out what is installed where? Trying to use grub/grub2, I've often wondered: Is my grub installation on /dev/sda a) complete? or did some other OS/RAID controller/whatever overwrite parts of it? b) uptodate? or did I forget to run grub-install after upgrading the grub package c) identical to the /dev/sdb mirror? or did I forget to run grub-install after replacing /dev/sda? Some scripts answering these questions would really be helpful. Yes, I know. Send patches. Might do. Or just go for extlinux, which seems to DTRT for me. Bjørn -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: lilo about to be dropped?
Giacomo A. Catenazzi c...@debian.org writes: Ferenc Wagner wrote: Giacomo A. Catenazzi c...@debian.org writes: Does grub use the unallocated disk space near the MBR? Yes. As far as I know, even grub2 does so, but pls. correct me. So next question: why does windoze installation write to these block (but not to MBR)? Ah, ok the windoze in question is already the answer ;-) Linux installation (with grub, that is, in most of the cases) writes there as well. On what basis do you blame Windows for the same? http://lwn.net/Articles/322777/ Thanks, that was a very interesting read. -- Cheers, Feri. -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: lilo about to be dropped?
Samuel Thibault wrote: Giacomo Catenazzi, le Wed 08 Apr 2009 19:47:55 +0200, a écrit : Samuel Thibault wrote: I installed grub (and Debian). Trying the Windows hidden partition (to install windows), grub stopped working (it was rescue mode, but without capability to rescue something). Also rescue disk + reconfiguring + update-grub did nothing. Err, did you re-run install-grub? No ;-) only update-grub and dpkg-reconfigure -plow grub-pc grub-common Then little wonder. update-grub only updates menu.lst. I was expecting that reconfigure will do the right things, grub maintainers considered that it's a bad thing to automatically reinstall things in a MBR. You need to re-run grub-install to do that. but MBR was fine, i.e. I had the grub rescue prompt. But no way to boot from rescue prompt. I know to few about grub... So I wonder what was overwriten. Does grub use the unallocated disk space near the MBR? This is very bad: I think in future we will use also this space. The cylinder notation in partition table has no more physical meaning, so also the partition boundary could change (but there was a discussion about poor performance if partitions/filesystem was not aligned to the physical block sectors.) ciao cate -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: lilo about to be dropped?
Giacomo A. Catenazzi c...@debian.org writes: Does grub use the unallocated disk space near the MBR? Yes. As far as I know, even grub2 does so, but pls. correct me. there was a discussion about poor performance if partitions/ filesystem was not aligned to the physical block sectors. I also heard hpa mention this. But this doesn't mean he is a big fan of using this space. Which discussion do you refer to? -- Feri. -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: lilo about to be dropped?
Ferenc Wagner wrote: Giacomo A. Catenazzi c...@debian.org writes: Does grub use the unallocated disk space near the MBR? Yes. As far as I know, even grub2 does so, but pls. correct me. So next question: why does windoze installation write to these block (but not to MBR)? Ah, ok the windoze in question is already the answer ;-) there was a discussion about poor performance if partitions/ filesystem was not aligned to the physical block sectors. I also heard hpa mention this. But this doesn't mean he is a big fan of using this space. Which discussion do you refer to? I think this: http://lwn.net/Articles/322777/ (note: windoze align partition at 1MB according comments). ciao cate -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: lilo about to be dropped?
Nenolod: sorry for the other mail. William Pitcock wrote: On Tue, 2009-04-07 at 13:06 -0300, Otavio Salvador wrote: On Mon, Apr 6, 2009 at 9:21 PM, William Pitcock neno...@sacredspiral.co.uk wrote: Lilo upstream is dead (no release in quite a while), but the lilo maintainer has also been seen as saying in various mailing lists etc, that since Debian patches lilo that he has no interest in helping to fix problems in our version. Ok but you could try to push those patches upstream. This is how grub has been improved and also parted. This works most of time. This way we reduce the amount of patches we keep in Debian and also you could try to get in touch with other distros to share the load and avoid reworking at same things. lilo is officially unmaintained now. The canonical website of lilo now points to a 404 error page, see http://lilo.go.dyndns.org/ . as grub was not really maintained. Also grub2 doesn't seems so fast in development. I think these kind of project have difficult to maintain motivated maintainer. What do the other distributions? extlinux seems the real alternative: the maintainer is active in kernel boot since a lot of years, he has a good knowledge of lilo (thus is not the usual: do a new project because I cannot read/understand the old code). OTOH hpa test always the boot changes in kernel, and lilo is always tested, so in this regards, he take also care about lilo. I think we need a discussion of the fate of lilo at DebConf. I volunteer to check and give you technical details of the main boot loaders for i386/amd64 architecture, so that we can decide better (and give inputs to upstream on what they miss). Any interest in such talk? BTW: my new laptop was saved by lilo ;-) I installed grub (and Debian). Trying the Windows hidden partition (to install windows), grub stopped working (it was rescue mode, but without capability to rescue something). Also rescue disk + reconfiguring + update-grub did nothing. Installing lilo gave me a know boot environment, and it worked at first try. So: lilo should live! ciao cate -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: lilo about to be dropped?
On Wed, Apr 8, 2009 at 10:22 PM, Giacomo A. Catenazzi wrote: as grub was not really maintained. Also grub2 doesn't seems so fast in development. I think these kind of project have difficult to maintain motivated maintainer. I would really love to own a computer that used coreboot, Linux and a userland bootloader like kboot/kexec-loader/runnix to boot. -- bye, pabs http://wiki.debian.org/PaulWise -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: lilo about to be dropped?
As the silent co-maintainer of lilo I believe I should now voice my thoughts on this I too believe that lilo should belive that lilo should be remove *at some point* but now is not the time. So I restate my willingness to take over fully publicly. Upstream made a release of a bootloader in 2007 a bootloader is quite different from an internet facing service or a desktop app, so it is possible that upstream hasn't made a release because they haven't felt a need to existed. From this thread there still appears to be use cases for lilo and it seems to be meeting the needs of the people that need it. Unless there is a security hole or show stopping bug that makes the package totally unusable why remove it. There will eventually be that case and when such a time comes we will reexamine the issue but why fix what is working for people. Again I will take over the package if you (nenolod) don't want it anymore. I An RM seems overkill when a line in the package description will do nicely On Wed, Apr 8, 2009 at 10:22 AM, Giacomo A. Catenazzi c...@debian.org wrote: Nenolod: sorry for the other mail. William Pitcock wrote: On Tue, 2009-04-07 at 13:06 -0300, Otavio Salvador wrote: On Mon, Apr 6, 2009 at 9:21 PM, William Pitcock neno...@sacredspiral.co.uk wrote: Lilo upstream is dead (no release in quite a while), but the lilo maintainer has also been seen as saying in various mailing lists etc, that since Debian patches lilo that he has no interest in helping to fix problems in our version. Ok but you could try to push those patches upstream. This is how grub has been improved and also parted. This works most of time. This way we reduce the amount of patches we keep in Debian and also you could try to get in touch with other distros to share the load and avoid reworking at same things. lilo is officially unmaintained now. The canonical website of lilo now points to a 404 error page, see http://lilo.go.dyndns.org/ . as grub was not really maintained. Also grub2 doesn't seems so fast in development. I think these kind of project have difficult to maintain motivated maintainer. What do the other distributions? extlinux seems the real alternative: the maintainer is active in kernel boot since a lot of years, he has a good knowledge of lilo (thus is not the usual: do a new project because I cannot read/understand the old code). OTOH hpa test always the boot changes in kernel, and lilo is always tested, so in this regards, he take also care about lilo. I think we need a discussion of the fate of lilo at DebConf. I volunteer to check and give you technical details of the main boot loaders for i386/amd64 architecture, so that we can decide better (and give inputs to upstream on what they miss). Any interest in such talk? BTW: my new laptop was saved by lilo ;-) I installed grub (and Debian). Trying the Windows hidden partition (to install windows), grub stopped working (it was rescue mode, but without capability to rescue something). Also rescue disk + reconfiguring + update-grub did nothing. Installing lilo gave me a know boot environment, and it worked at first try. So: lilo should live! ciao cate -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: lilo about to be dropped?
On Wed, 2009-04-08 at 11:05 -0400, Matt Arnold wrote: As the silent co-maintainer of lilo I believe I should now voice my thoughts on this I too believe that lilo should belive that lilo should be remove *at some point* but now is not the time. So I restate my willingness to take over fully publicly. Upstream made a release of a bootloader in 2007 a bootloader is quite different from an internet facing service or a desktop app, so it is possible that upstream hasn't made a release because they haven't felt a need to existed. From this thread there still appears to be use cases for lilo and it seems to be meeting the needs of the people that need it. Unless there is a security hole or show stopping bug that makes the package totally unusable why remove it. There will eventually be that case and when such a time comes we will reexamine the issue but why fix what is working for people. Again I will take over the package if you (nenolod) don't want it anymore. I An RM seems overkill when a line in the package description will do nicely Does this mean that you will become lilo upstream as well? Are you *qualified* to become lilo upstream? Do you know assembly language? (tip: most of the important parts are assembly language.) If not, then stop talking now. Anything less is unhealthy as it will just become another XMMS with lots of patches ontop of it to fix bugs. William signature.asc Description: This is a digitally signed message part
Re: lilo about to be dropped?
On Wed, 2009-04-08 at 16:22 +0200, Giacomo A. Catenazzi wrote: Nenolod: sorry for the other mail. William Pitcock wrote: On Tue, 2009-04-07 at 13:06 -0300, Otavio Salvador wrote: On Mon, Apr 6, 2009 at 9:21 PM, William Pitcock neno...@sacredspiral.co.uk wrote: Lilo upstream is dead (no release in quite a while), but the lilo maintainer has also been seen as saying in various mailing lists etc, that since Debian patches lilo that he has no interest in helping to fix problems in our version. Ok but you could try to push those patches upstream. This is how grub has been improved and also parted. This works most of time. This way we reduce the amount of patches we keep in Debian and also you could try to get in touch with other distros to share the load and avoid reworking at same things. lilo is officially unmaintained now. The canonical website of lilo now points to a 404 error page, see http://lilo.go.dyndns.org/ . as grub was not really maintained. Also grub2 doesn't seems so fast in development. I think these kind of project have difficult to maintain motivated maintainer. What do the other distributions? I've seen a few smaller distributions looking into extlinux as an alternative to lilo. Not sure what the redhat/fedora/centos/etc camp are doing though. extlinux seems the real alternative: the maintainer is active in kernel boot since a lot of years, he has a good knowledge of lilo (thus is not the usual: do a new project because I cannot read/understand the old code). OTOH hpa test always the boot changes in kernel, and lilo is always tested, so in this regards, he take also care about lilo. I think we need a discussion of the fate of lilo at DebConf. I volunteer to check and give you technical details of the main boot loaders for i386/amd64 architecture, so that we can decide better (and give inputs to upstream on what they miss). Any interest in such talk? It would be a good topic for discussion if I can make it to DebConf this year (which is probable, just a matter of getting a good deal on plane tickets). BTW: my new laptop was saved by lilo ;-) One of my newer servers was also saved by lilo (fucking adaptec SAS controllers...). However, the current health of lilo is still something to be concerned about. I installed grub (and Debian). Trying the Windows hidden partition (to install windows), grub stopped working (it was rescue mode, but without capability to rescue something). Also rescue disk + reconfiguring + update-grub did nothing. Installing lilo gave me a know boot environment, and it worked at first try. So: lilo should live! That's because grub does a number of things incorrectly as well. I don't think extlinux repeats those mistakes though, at least from what I have seen in production. William signature.asc Description: This is a digitally signed message part
Re: lilo about to be dropped?
I installed grub (and Debian). Trying the Windows hidden partition (to install windows), grub stopped working (it was rescue mode, but without capability to rescue something). Also rescue disk + reconfiguring + update-grub did nothing. Err, did you re-run install-grub? Samuel -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: lilo about to be dropped?
Samuel Thibault wrote: I installed grub (and Debian). Trying the Windows hidden partition (to install windows), grub stopped working (it was rescue mode, but without capability to rescue something). Also rescue disk + reconfiguring + update-grub did nothing. Err, did you re-run install-grub? No ;-) only update-grub and dpkg-reconfigure -plow grub-pc grub-common I was expecting that reconfigure will do the right things, but on the other side, without a good rescue CD (64-bit), I just renounced after 2 tries, not to redo the long d-i rescue procedure ciao cate -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: lilo about to be dropped?
On Wed, Apr 08, 2009 at 11:21:12AM -0500, William Pitcock wrote: Does this mean that you will become lilo upstream as well? Are you *qualified* to become lilo upstream? Do you know assembly language? (tip: most of the important parts are assembly language.) If not, then stop talking now. Anything less is unhealthy as it will just become another XMMS with lots of patches ontop of it to fix bugs. No, it'll become another grub1 with lots of patches on top of it to fix bugs. Oh wait, it already is that. -- Steve Langasek Give me a lever long enough and a Free OS Debian Developer to set it on, and I can move the world. Ubuntu Developerhttp://www.debian.org/ slanga...@ubuntu.com vor...@debian.org -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: lilo about to be dropped?
On Mon, Apr 06, 2009 at 11:05:43AM -0500, William Pitcock wrote:
On Mon, 2009-04-06 at 08:53 -0700, Steve Langasek wrote:
On Mon, Apr 06, 2009 at 10:13:32AM -0500, William Pitcock wrote:
I agree here too. I think these install paths could be replaced by
ext2linux as well, if that is what is needed to be done.
And why in the world is it useful to transition these use cases to ext2linux
when we already have a lilo package that suits these needs perfectly well?
Because it does not.
That's not for you to say. There are clearly a large number of users who
are using lilo (3388 who also enable popcon - and if they're running popcon,
I guess that means lilo is working for them what with that whole booting
thing). So lilo *is* meeting the needs of these users, notwithstanding your
dissatisfaction with the use case coverage.
The LVM support in LILO is hideously broken, so these arguments do not
really matter. It only works in certain conditions and is known to break
horribly if you have say, a kernel spanning multiple PVs.
They matter to the users who are *using* lilo this way, whether or not you
happen to find the implementation to your liking.
I don't use lilo. I have gradually transitioned all my old installs over to
grub, delayed only by the need to accomodate the risks of downtime. That
doesn't mean I think it's acceptable to drop lilo on the floor for squeeze,
when it's still being offered as an installation option for *two* supported
Debian releases, in some cases by default, and there doesn't appear to be an
actual transition plan for those users who currently have lilo installed,
whether that's by necessity or choice.
Only a true idiot boots off an LVM volume anyway, since there is risk of
metadata corruption, etc.
Bullshit.
But, you will. Infact, you told me yesterday on IRC that your intention
is to take over lilo maintenance to score points with DDs and that you
just needed it for a few months. This isn't the right issue to score
points on, as lack of proper maintenance is WORSE than not having it in
Debian at all.
No - *bad* maintenance is worse than not having it in Debian at all. But
having the package in Debian on autopilot is *better* than leaving those
currently using it out in the cold, or giving them a poorly-implemented
transition.
Insisting that we drop lilo from the archive before any work has been done
to make a transition to grub{1,2} possible is putting the cart before the
horse.
--
Steve Langasek Give me a lever long enough and a Free OS
Debian Developer to set it on, and I can move the world.
Ubuntu Developerhttp://www.debian.org/
slanga...@ubuntu.com vor...@debian.org
--
To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: lilo about to be dropped?
Giacomo Catenazzi, le Wed 08 Apr 2009 19:47:55 +0200, a écrit : Samuel Thibault wrote: I installed grub (and Debian). Trying the Windows hidden partition (to install windows), grub stopped working (it was rescue mode, but without capability to rescue something). Also rescue disk + reconfiguring + update-grub did nothing. Err, did you re-run install-grub? No ;-) only update-grub and dpkg-reconfigure -plow grub-pc grub-common Then little wonder. update-grub only updates menu.lst. I was expecting that reconfigure will do the right things, grub maintainers considered that it's a bad thing to automatically reinstall things in a MBR. You need to re-run grub-install to do that. Samuel -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: lilo about to be dropped?
On Mon, 2009-04-06 at 22:20 +0100, Stephen Gran wrote: This one time, at band camp, William Pitcock said: The only way it is feasible to do so is to drop all of the Debian patches. Without this, upstream is not cooperative with us. Why is this? See my other mail, basically, lilo upstream view is that our patches broke it and that we have to fix it ourselves. I've seen him on various threads saying basically that over the years. But regardless, a lilo release has not been made in some time. William signature.asc Description: This is a digitally signed message part
Re: lilo about to be dropped?
On Mon, 2009-04-06 at 18:46 +0200, Christian Perrier wrote: Quoting Frans Pop (elen...@planet.nl): I'm not sure where the original mail comes from, but IMO this should be From lilo package BTS which I was tracking for l10n purposes. So I just happened to notice William's answer to a bug report and thought it would be good for this to be discussed in public. Clearly, I didn't choose the right place to discuss and the topic has wider implications than just D-I, as the followups show. Good thing that you made the discussion wider. Don't we have some install paths that still depend on LILO? Yes: /boot on LVM is the main one. Anyway, even if we don't, I think we should track that lilo removal and coordinate with William, in order to stop providing lilo-installer. And, I think this should be mentioned as a release goal (dropping lilo). Either high priority if we have install paths depending on lilo, or normal priority otherwise. D-I release goal or Debian release goal [1]? Clearly Debian release goal. IMO the latter could well be justified as there will also need to be some kind of upgrade strategy for existing users that does not make uncontrolled changes on their hard disk or loses them the ability to boot alternative OSes on dual (or multi) boot systems. Which might be very tricky But, as William mentioned in his original mail, upstream activity seems to be low so we need to figure out if we want to keep yet another unmaintained software in Debian. What later puzzled me if the mention in non collaboratve upstream *if we don't drop Debian patches*. That's not exactly inactive upstream so it would be good to clarify the situation of lilo upstream. Lilo upstream is dead (no release in quite a while), but the lilo maintainer has also been seen as saying in various mailing lists etc, that since Debian patches lilo that he has no interest in helping to fix problems in our version. William signature.asc Description: This is a digitally signed message part
Re: lilo about to be dropped?
Stephen Gran wrote: This one time, at band camp, William Pitcock said: The only way it is feasible to do so is to drop all of the Debian patches. Without this, upstream is not cooperative with us. Why is this? I think because of William Pitcock with: - his very strong words, - his attitude: perfect or nothing (in design, in management, ...), - his lack to listen upstreams and their needs: needs of other distributions, old compatibility needs, or simply time constrain and limited interest of upstream. ciao nenolod ;-) ciao cate -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: lilo about to be dropped?
On Mon, Apr 6, 2009 at 9:21 PM, William Pitcock neno...@sacredspiral.co.uk wrote: Lilo upstream is dead (no release in quite a while), but the lilo maintainer has also been seen as saying in various mailing lists etc, that since Debian patches lilo that he has no interest in helping to fix problems in our version. Ok but you could try to push those patches upstream. This is how grub has been improved and also parted. This works most of time. This way we reduce the amount of patches we keep in Debian and also you could try to get in touch with other distros to share the load and avoid reworking at same things. -- Otavio Salvador O.S. Systems E-mail: ota...@ossystems.com.br http://www.ossystems.com.br Mobile: +55 53 9981-7854 http://projetos.ossystems.com.br -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: lilo about to be dropped?
On Tue, 2009-04-07 at 10:52 +0200, Giacomo A. Catenazzi wrote: Stephen Gran wrote: This one time, at band camp, William Pitcock said: The only way it is feasible to do so is to drop all of the Debian patches. Without this, upstream is not cooperative with us. Why is this? I think because of William Pitcock with: - his very strong words, - his attitude: perfect or nothing (in design, in management, ...), - his lack to listen upstreams and their needs: needs of other distributions, old compatibility needs, or simply time constrain and limited interest of upstream. ciao nenolod ;-) Actually, the damage was done years ago, long before I ever maintained lilo. But thanks for the flame. William signature.asc Description: This is a digitally signed message part
Re: lilo about to be dropped?
On Tue, 2009-04-07 at 13:06 -0300, Otavio Salvador wrote: On Mon, Apr 6, 2009 at 9:21 PM, William Pitcock neno...@sacredspiral.co.uk wrote: Lilo upstream is dead (no release in quite a while), but the lilo maintainer has also been seen as saying in various mailing lists etc, that since Debian patches lilo that he has no interest in helping to fix problems in our version. Ok but you could try to push those patches upstream. This is how grub has been improved and also parted. This works most of time. This way we reduce the amount of patches we keep in Debian and also you could try to get in touch with other distros to share the load and avoid reworking at same things. lilo is officially unmaintained now. The canonical website of lilo now points to a 404 error page, see http://lilo.go.dyndns.org/ . So at this point, our only option seems to be taking over upstream lilo maintainance ourselves (which could be a good thing in some ways, I am not denying that), or find a way to transition these use-cases to grub/grub2/extlinux. However, if we are to maintain lilo ourselves, then we need to flesh out exactly what usecases we're going to be using it for. I recommend if we go that route that we come up with a list of improvements that we want to see and get to hacking. If some of the people who like lilo a lot got around to helping with a fork, we could create a much less buggy bootloader than the current lilo. Alternatively, we can just leave it and let it become another XMMS. I don't like this solution very much. William signature.asc Description: This is a digitally signed message part
Re: lilo about to be dropped?
On Monday 06 April 2009, Christian Perrier wrote: Quoting William Pitcock (neno...@dereferenced.org): lilo is due for removal anyway due to being unmaintained upstream and the widespread availability of alternatives. I think that last part is debatable. I do not have time to manage the removal at this point, but it will be gone by June. Has the package already been offered for adoption? Preferably with an overview of its current (upstream) status and main issues. I'd say that if there's anybody willing to (actively) maintain it, it should not be removed. This is a heads up mail for the D-I team. I'm not sure where the original mail comes from, but IMO this should be discussed on d-devel, especially since it impacts more than just D-I. I suspect there are quite a few packages that make some sort of provisions for lilo. There are also significant numbers of people still using lilo for, at least for them, very good reasons. Anyone remember the fairly big upset when lilo was removed from testing around D-I Lenny Beta2? Don't we have some install paths that still depend on LILO? Yes: /boot on LVM is the main one. Anyway, even if we don't, I think we should track that lilo removal and coordinate with William, in order to stop providing lilo-installer. And, I think this should be mentioned as a release goal (dropping lilo). Either high priority if we have install paths depending on lilo, or normal priority otherwise. D-I release goal or Debian release goal [1]? IMO the latter could well be justified as there will also need to be some kind of upgrade strategy for existing users that does not make uncontrolled changes on their hard disk or loses them the ability to boot alternative OSes on dual (or multi) boot systems. Cheers, FJP [1] goal is a somewhat strange term here... signature.asc Description: This is a digitally signed message part.
Re: lilo about to be dropped?
On Mon Apr 06 08:55, Frans Pop wrote: This is a heads up mail for the D-I team. I'm not sure where the original mail comes from, but IMO this should be discussed on d-devel, especially since it impacts more than just D-I. I suspect there are quite a few packages that make some sort of provisions for lilo. There are also significant numbers of people still using lilo for, at least for them, very good reasons. Yes, please do discuss it here. I am one of those users, grub didn't work on one of my machines for some reason. Anyway, isn't grub1 equally unmaintained upstream? I thought they were only working on grub2 (which isn't ready for use yet, or is it?) Don't we have some install paths that still depend on LILO? Yes: /boot on LVM is the main one. We _certainly_ shouldn't throw it out if there are _known_ situations for which it's required. By all means print large warnings or only make it available in expert mode, or whatever, but please don't break existing functionality. Matt -- Matthew Johnson signature.asc Description: Digital signature
Re: lilo about to be dropped?
Matthew Johnson mj...@debian.org writes: On Mon Apr 06 08:55, Frans Pop wrote: This is a heads up mail for the D-I team. I'm not sure where the original mail comes from, but IMO this should be discussed on d-devel, especially since it impacts more than just D-I. I suspect there are quite a few packages that make some sort of provisions for lilo. There are also significant numbers of people still using lilo for, at least for them, very good reasons. Yes, please do discuss it here. I am one of those users, grub didn't work on one of my machines for some reason. Anyway, isn't grub1 equally unmaintained upstream? I thought they were only working on grub2 (which isn't ready for use yet, or is it?) So lets get grub2 working everywhere. :) A worthy goal. Don't we have some install paths that still depend on LILO? Yes: /boot on LVM is the main one. We _certainly_ shouldn't throw it out if there are _known_ situations for which it's required. We just shouldn't have /boot on lvm. At least there should be one place outside lvm to store /etc/lvm/archive and /etc/lvm/backup so that in the case lvm breaks (gets broken by the user) one can repair it. Linking them to /boot/lvm/archive and /boot/lvm/backup with /boot outside lvm seem like a good idea. The problem with /boot on lvm is that moving or resizing it can break it. So I always found it a good partition to keep outside lvm. By all means print large warnings or only make it available in expert mode, or whatever, but please don't break existing functionality. Matt -- Matthew Johnson MfG Goswin -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: lilo about to be dropped?
On Mon Apr 06 11:07, Goswin von Brederlow wrote: So lets get grub2 working everywhere. :) A worthy goal. Sure, but don't remove lilo until we're happy that grub2 does work everywhere. Matt -- Matthew Johnson signature.asc Description: Digital signature
Re: lilo about to be dropped?
Frans Pop elen...@planet.nl writes: On Monday 06 April 2009, Christian Perrier wrote: [...] I do not have time to manage the removal at this point, but it will be gone by June. Has the package already been offered for adoption? Preferably with an overview of its current (upstream) status and main issues. I'd say that if there's anybody willing to (actively) maintain it, it should not be removed. Fully agree; it should be properly offered for adoption. This is a heads up mail for the D-I team. I'm not sure where the original mail comes from, but IMO this should be discussed on d-devel, especially since it impacts more than just D-I. I suspect there are quite a few packages that make some sort of provisions for lilo. There are also significant numbers of people still using lilo for, at least for them, very good reasons. Anyone remember the fairly big upset when lilo was removed from testing around D-I Lenny Beta2? I also share the feeling that a lot of people still uses LILO; if possible I do belive it should be kept. [...] -- O T A V I OS A L V A D O R - E-mail: ota...@debian.org UIN: 5906116 GNU/Linux User: 239058 GPG ID: 49A5F855 Home Page: http://otavio.ossystems.com.br - Microsoft sells you Windows ... Linux gives you the whole house. -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: lilo about to be dropped?
Frans Pop wrote:
On Monday 06 April 2009, Christian Perrier wrote:
Quoting William Pitcock (neno...@dereferenced.org):
lilo is due for removal anyway due to being unmaintained upstream and
the widespread availability of alternatives.
I think that last part is debatable.
I do not have time to manage the removal at this point, but it will
be gone by June.
Has the package already been offered for adoption? Preferably with an
overview of its current (upstream) status and main issues. I'd say that
if there's anybody willing to (actively) maintain it, it should not be
removed.
This is a heads up mail for the D-I team.
I'm not sure where the original mail comes from, but IMO this should be
discussed on d-devel, especially since it impacts more than just D-I. I
suspect there are quite a few packages that make some sort of provisions
for lilo.
There are also significant numbers of people still using lilo for, at
least for them, very good reasons.
I totally agree.
But I think that lilo package description must be changed, warning new
users that lilo have several limits (thus not all kernel within debian
are bootable with lilo).
Maybe we could also require grub{,2} when installing lilo (chained
as other in lilo, for emergency, new debian kernel policies, etc),
but I don't know if it is feasible (e.g. when lilo is not in MBR).
ciao
cate
--
To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: lilo about to be dropped?
On Mon, 2009-04-06 at 06:42 +0200, Christian Perrier wrote: Quoting William Pitcock (neno...@dereferenced.org): lilo is due for removal anyway due to being unmaintained upstream and the widespread availability of alternatives. I do not have time to manage the removal at this point, but it will be gone by June. This is a heads up mail for the D-I team. Don't we have some install paths that still depend on LILO? We do, in lilo-installer. Such as booting directly from LVM. But with increasing maturity of grub2, this does not seem necessary anymore. Anyway, even if we don't, I think we should track that lilo removal and coordinate with William, in order to stop providing lilo-installer. I agree. And, I think this should be mentioned as a release goal (dropping lilo). Either high priority if we have install paths depending on lilo, or normal priority otherwise. I agree here too. I think these install paths could be replaced by ext2linux as well, if that is what is needed to be done. William -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: lilo about to be dropped?
On Mon, 2009-04-06 at 10:44 -0300, Otavio Salvador wrote: Frans Pop elen...@planet.nl writes: On Monday 06 April 2009, Christian Perrier wrote: [...] I do not have time to manage the removal at this point, but it will be gone by June. Has the package already been offered for adoption? Preferably with an overview of its current (upstream) status and main issues. I'd say that if there's anybody willing to (actively) maintain it, it should not be removed. Fully agree; it should be properly offered for adoption. This is a heads up mail for the D-I team. I'm not sure where the original mail comes from, but IMO this should be discussed on d-devel, especially since it impacts more than just D-I. I suspect there are quite a few packages that make some sort of provisions for lilo. There are also significant numbers of people still using lilo for, at least for them, very good reasons. Anyone remember the fairly big upset when lilo was removed from testing around D-I Lenny Beta2? I also share the feeling that a lot of people still uses LILO; if possible I do belive it should be kept. The only way it is feasible to do so is to drop all of the Debian patches. Without this, upstream is not cooperative with us. However, I think ext2linux is a feasible upgrade path and that lilo will become unnecessary by the release of squeeze. William -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: lilo about to be dropped?
William Pitcock wrote:
On Mon, 2009-04-06 at 17:26 +0200, Giacomo A. Catenazzi wrote:
Frans Pop wrote:
On Monday 06 April 2009, Christian Perrier wrote:
Quoting William Pitcock (neno...@dereferenced.org):
lilo is due for removal anyway due to being unmaintained upstream and
the widespread availability of alternatives.
I think that last part is debatable.
I do not have time to manage the removal at this point, but it will
be gone by June.
Has the package already been offered for adoption? Preferably with an
overview of its current (upstream) status and main issues. I'd say that
if there's anybody willing to (actively) maintain it, it should not be
removed.
This is a heads up mail for the D-I team.
I'm not sure where the original mail comes from, but IMO this should be
discussed on d-devel, especially since it impacts more than just D-I. I
suspect there are quite a few packages that make some sort of provisions
for lilo.
There are also significant numbers of people still using lilo for, at
least for them, very good reasons.
I totally agree.
But I think that lilo package description must be changed, warning new
users that lilo have several limits (thus not all kernel within debian
are bootable with lilo).
Maybe we could also require grub{,2} when installing lilo (chained
as other in lilo, for emergency, new debian kernel policies, etc),
but I don't know if it is feasible (e.g. when lilo is not in MBR).
chainloader will work with lilo, but lilo is only kept around for the
people who are crazy and booting off LVMs as it is.
Yes, but it works if you have an additional partition (for boot
record). I don't know if they could live in the same partition
(with some magic).
But IIRC lilo fails also in other cases: some xen immages, on very big
images (which can be reached in some initram).
Booting off LVMs is supported directly by grub2 and ext2linux could
probably be modified to support it in a much better way than lilo does
it, so this is not really a compelling argument for keeping it.
What is ext2linux? packages.d.o and google doesn't give me relevant
informations.
ciao
cate
William
--
To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: lilo about to be dropped?
On Mon, 2009-04-06 at 17:26 +0200, Giacomo A. Catenazzi wrote:
Frans Pop wrote:
On Monday 06 April 2009, Christian Perrier wrote:
Quoting William Pitcock (neno...@dereferenced.org):
lilo is due for removal anyway due to being unmaintained upstream and
the widespread availability of alternatives.
I think that last part is debatable.
I do not have time to manage the removal at this point, but it will
be gone by June.
Has the package already been offered for adoption? Preferably with an
overview of its current (upstream) status and main issues. I'd say that
if there's anybody willing to (actively) maintain it, it should not be
removed.
This is a heads up mail for the D-I team.
I'm not sure where the original mail comes from, but IMO this should be
discussed on d-devel, especially since it impacts more than just D-I. I
suspect there are quite a few packages that make some sort of provisions
for lilo.
There are also significant numbers of people still using lilo for, at
least for them, very good reasons.
I totally agree.
But I think that lilo package description must be changed, warning new
users that lilo have several limits (thus not all kernel within debian
are bootable with lilo).
Maybe we could also require grub{,2} when installing lilo (chained
as other in lilo, for emergency, new debian kernel policies, etc),
but I don't know if it is feasible (e.g. when lilo is not in MBR).
chainloader will work with lilo, but lilo is only kept around for the
people who are crazy and booting off LVMs as it is.
Booting off LVMs is supported directly by grub2 and ext2linux could
probably be modified to support it in a much better way than lilo does
it, so this is not really a compelling argument for keeping it.
William
--
To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: lilo about to be dropped?
On Mon, Apr 06, 2009 at 10:13:32AM -0500, William Pitcock wrote: And, I think this should be mentioned as a release goal (dropping lilo). Either high priority if we have install paths depending on lilo, or normal priority otherwise. I agree here too. I think these install paths could be replaced by ext2linux as well, if that is what is needed to be done. And why in the world is it useful to transition these use cases to ext2linux when we already have a lilo package that suits these needs perfectly well? That sounds like pointless churn to me. If we're not able to migrate these users to grub (1 or 2), we ought not mess with it. -- Steve Langasek Give me a lever long enough and a Free OS Debian Developer to set it on, and I can move the world. Ubuntu Developerhttp://www.debian.org/ slanga...@ubuntu.com vor...@debian.org -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: lilo about to be dropped?
On Mon, 2009-04-06 at 17:40 +0200, Giacomo A. Catenazzi wrote:
William Pitcock wrote:
On Mon, 2009-04-06 at 17:26 +0200, Giacomo A. Catenazzi wrote:
Frans Pop wrote:
On Monday 06 April 2009, Christian Perrier wrote:
Quoting William Pitcock (neno...@dereferenced.org):
lilo is due for removal anyway due to being unmaintained upstream and
the widespread availability of alternatives.
I think that last part is debatable.
I do not have time to manage the removal at this point, but it will
be gone by June.
Has the package already been offered for adoption? Preferably with an
overview of its current (upstream) status and main issues. I'd say that
if there's anybody willing to (actively) maintain it, it should not be
removed.
This is a heads up mail for the D-I team.
I'm not sure where the original mail comes from, but IMO this should be
discussed on d-devel, especially since it impacts more than just D-I. I
suspect there are quite a few packages that make some sort of provisions
for lilo.
There are also significant numbers of people still using lilo for, at
least for them, very good reasons.
I totally agree.
But I think that lilo package description must be changed, warning new
users that lilo have several limits (thus not all kernel within debian
are bootable with lilo).
Maybe we could also require grub{,2} when installing lilo (chained
as other in lilo, for emergency, new debian kernel policies, etc),
but I don't know if it is feasible (e.g. when lilo is not in MBR).
chainloader will work with lilo, but lilo is only kept around for the
people who are crazy and booting off LVMs as it is.
Yes, but it works if you have an additional partition (for boot
record). I don't know if they could live in the same partition
(with some magic).
But IIRC lilo fails also in other cases: some xen immages, on very big
images (which can be reached in some initram).
Booting off LVMs is supported directly by grub2 and ext2linux could
probably be modified to support it in a much better way than lilo does
it, so this is not really a compelling argument for keeping it.
What is ext2linux? packages.d.o and google doesn't give me relevant
informations.
Oops. It is extlinux. It's syslinux except it boots off a hard-disk
instead of a floppy or CD. Quite similar to lilo in featureset.
William
--
To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: lilo about to be dropped?
On Mon, 2009-04-06 at 08:53 -0700, Steve Langasek wrote: On Mon, Apr 06, 2009 at 10:13:32AM -0500, William Pitcock wrote: And, I think this should be mentioned as a release goal (dropping lilo). Either high priority if we have install paths depending on lilo, or normal priority otherwise. I agree here too. I think these install paths could be replaced by ext2linux as well, if that is what is needed to be done. And why in the world is it useful to transition these use cases to ext2linux when we already have a lilo package that suits these needs perfectly well? Because it does not. That sounds like pointless churn to me. If we're not able to migrate these users to grub (1 or 2), we ought not mess with it. We can migrate them to grub2 in my opinion, extlinux is just another option worth considering for lilo diehards. I migrated my lilo boxes to grub2 a couple of weeks ago, and it seems to behave as expected. William -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: lilo about to be dropped?
Quoting Frans Pop (elen...@planet.nl): I'm not sure where the original mail comes from, but IMO this should be From lilo package BTS which I was tracking for l10n purposes. So I just happened to notice William's answer to a bug report and thought it would be good for this to be discussed in public. Clearly, I didn't choose the right place to discuss and the topic has wider implications than just D-I, as the followups show. Good thing that you made the discussion wider. Don't we have some install paths that still depend on LILO? Yes: /boot on LVM is the main one. Anyway, even if we don't, I think we should track that lilo removal and coordinate with William, in order to stop providing lilo-installer. And, I think this should be mentioned as a release goal (dropping lilo). Either high priority if we have install paths depending on lilo, or normal priority otherwise. D-I release goal or Debian release goal [1]? Clearly Debian release goal. IMO the latter could well be justified as there will also need to be some kind of upgrade strategy for existing users that does not make uncontrolled changes on their hard disk or loses them the ability to boot alternative OSes on dual (or multi) boot systems. Which might be very tricky But, as William mentioned in his original mail, upstream activity seems to be low so we need to figure out if we want to keep yet another unmaintained software in Debian. What later puzzled me if the mention in non collaboratve upstream *if we don't drop Debian patches*. That's not exactly inactive upstream so it would be good to clarify the situation of lilo upstream. signature.asc Description: Digital signature
Re: lilo about to be dropped?
This one time, at band camp, William Pitcock said: The only way it is feasible to do so is to drop all of the Debian patches. Without this, upstream is not cooperative with us. Why is this? -- - | ,''`.Stephen Gran | | : :' :sg...@debian.org | | `. `'Debian user, admin, and developer | |`- http://www.debian.org | - signature.asc Description: Digital signature
Re: lilo about to be dropped?
On Mon, 06 Apr 2009, Matthew Johnson wrote: On Mon Apr 06 11:07, Goswin von Brederlow wrote: So lets get grub2 working everywhere. :) A worthy goal. Sure, but don't remove lilo until we're happy that grub2 does work everywhere. And that we have something resembling acceptable, up-to-date documentation for it. -- One disk to rule them all, One disk to find them. One disk to bring them all and in the darkness grind them. In the Land of Redmond where the shadows lie. -- The Silicon Valley Tarot Henrique Holschuh -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
