Jeremiah wrote:
> There are some folks who're will to mirror our PureOS distro which is
> very generous of them.
Alas, I fear this is a bigger question that just rsync daemon
settings. :)
Whilst extremely kind of them, how will this work for our users?
Ask them to alter their sources.list? Are we going to commit to
providing a CDN of some description?
Even a round-robin DNS solution has issues in that we have no
immediate, insight or control over whether any "non-official"
mirror is up-to-date, something obviously important from a user
point of view.
Thoughts?
Best wishes,
--
Chris Lamb
https://puri.sm
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512
There are some folks who're will to mirror our PureOS distro which is
very generous of them. One of them, linux.pizza (
https://mirror.linux.pizza/) recommends we use rsync to support
mirrors. There is another potential mirror in the US who'll work with
rsync too.
What are the recommended locations to point our rsync daemon at so that
other sites might mirror our repos? This is what they mirror for
Debian; https://mirror.linux.pizza/debian/
Regards,
Jeremiah
-----BEGIN PGP SIGNATURE-----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=51Sb
-----END PGP SIGNATURE-----
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512
On Tue, 2019-01-29 at 23:27 +0100, Matthias Klumpp wrote:
> Am Di., 29. Jan. 2019 um 22:22 Uhr schrieb Jeremiah C. Foster
> <jeremiah.foster(a)puri.sm>:
> > [...]
> > Checksums match but it fails to boot. Message is;
> > firmware: failed to load i915/skl_dmc_ver1_27.bin (-2)
> > firmware_class: See https://wiki.debian.org/Firmware for
> > imformation
> > about missing firmware.
> >
> > rm: can't remove '/root/var/crash/crash.init': No such file or
> > directory
>
> All of these are not hard errors. The first issue is just missing
> firmware, the second ones are from us not using any crash reporter in
> PureOS and casper still having some code for that (the messages are
> harmless though).
> Which virtualization solution are you using?
I'm booting on a Librem13 v3.
> On real metal, like the Librems, this issue never appeared though. If
> the issue stems from a Librem, we need to look into what's causing it
> (try removing quiet splash from the kernel command-line, if it boots
> then, it's an indicator Plymouth is to blame).
In the end I was able to install. I tried three times, first time
failed, the second time worked with failsafe, the third time worked
with the regular installation.
Where would I change the command-line? I suppose I need to edit the
image I'm booting from, but where? In grub.cfg?
Regards,
Jeremiah
-----BEGIN PGP SIGNATURE-----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=uqeo
-----END PGP SIGNATURE-----
Hey,
> I think I just inferred a "I need this info as soon as possible
> since it blocks my work" from the ping
Thanks for the feedback; I can relate to that point of view.
> I hope you didn't interpret my statement as rude, I seriously thought
> I missed some information/request to do something in the mails.
Heh, but if there was something to do I would have … *evil grin
spreads slowly over his face* … used the issue tracker!
Thanks. :)
Best wishes,
--
Chris Lamb
https://puri.sm
Hi Matthias,
> I was looking through my notes and I think I rebased the packaging on
> Debian's upstream packaging to make it easier for me to incorporate
> some bigger systemd patches I wrote. No changes that you made should
> have been lost though, or were they?
Without doing a bit of Git archeology it might be difficult to 100%
say but I think everything was merged upstream and, of course, your
package has been in PureOS for bit… let's assume Everything Is Fine.
[…]
> In any case, I think this was my fault, although admittedly I don't
> know what you would like me to do here.
Not asking for any technical action: this was, right from the start,
merely trying to work out what happened as it was somewhat confusing.
Please do let me know how I could have made that clearer for the
future.
Best wishes,
--
Chris Lamb
https://puri.sm
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512
Hi,
In this PureOS issue; https://tracker.pureos.net/T677 it appears that
we have two places to download a PureOS ISO;
- - The PureOS download page (https://pureos.net/download/)
- - And downloads (https://downloads.puri.sm/live/gnome/2018-12-10/)
The downloads page has a more recent release and I wonder if we should
move it to the PureOS "download" page? What is the process?
Regards,
Jeremiah
- --
GnuPG key fingerprint;
798D B834 436A 7BE3 8C97 422D 0DC0 6220 5451 931B
-----BEGIN PGP SIGNATURE-----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=pC+9
-----END PGP SIGNATURE-----
Matthias,
Thank you for your reply.
> > > I just went to refresh my local copy of pureos/core/systemd.git and
> > > noticed that the Git history had completely diverged from my
> > > 238-5pureos1 upload of June 2018.
[…]
> I don't remember changing anything (last commit was by me 5 months
> ago), but the current Git VCS is the exact same systemd uses in
> Debian, which makes merging with the Debian packaging much easier.
> So, this layout is certainly intentional.
Alas, I think you may have misread what I wrote — I had previously
uploaded systemd around June 2018 (based on the very same Debian
packaging you mention), including pushing to our Git repositories
as usual/expected and not doing anything that would prevent merging…
if only to keep the diff minimal and sanity maximal.
However, when I went to do another update in the last few days I
noticed that the history had been entirely scrubbed (see grandfather
mail in this thread.) It is this latter rewriting that made me
curious.
> Didn't you make a systemd upload just recently?
Mm, bhence my "I just went to refresh…" that prompted me to notice
in the first place, the actual upload is not really relevant.
Best wishes,
--
Chris Lamb
https://puri.sm
[Gentle ping on the below?]
§
Matthias,
Thank you for your reply.
> > > I just went to refresh my local copy of pureos/core/systemd.git and
> > > noticed that the Git history had completely diverged from my
> > > 238-5pureos1 upload of June 2018.
[…]
> I don't remember changing anything (last commit was by me 5 months
> ago), but the current Git VCS is the exact same systemd uses in
> Debian, which makes merging with the Debian packaging much easier.
> So, this layout is certainly intentional.
Alas, I think you may have misread what I wrote — I had previously
uploaded systemd around June 2018 (based on the very same Debian
packaging you mention), including pushing to our Git repositories
as usual/expected and not doing anything that would prevent merging…
if only to keep the diff minimal and sanity maximal.
However, when I went to do another update in the last few days I
noticed that the history had been entirely scrubbed (see grandfather
mail in this thread.) It is this latter rewriting that made me
curious.
> Didn't you make a systemd upload just recently?
Mm, bhence my "I just went to refresh…" that prompted me to notice
in the first place, the actual upload is not really relevant.
Best wishes,
--
Chris Lamb
https://puri.sm
Matthias wrote:
> this issue has been fixed in PureOS for 2 days now (since 2019-01-24)
> and in green since yesterday
Mm I understand that, but did you perhaps miss:
> > extra curiosity about the process of expedition in general
^^^^^^^^^^^^^^^^^^^^^
Best wishes,
--
Chris Lamb
https://puri.sm
Hi Jeremiah,
> I will follow up and see where we are. I do know PureOS folks are aware
> of this issue.
JFTR I am part of this PureOS team/folks, hence extra curiosity
about the process of expeditin in general, not specifically for
CVE-2019-3462.
(If we can get this centrally documented for the next person, all
the better...!)
Best wishes,
--
Chris Lamb
https://puri.sm