On Thu, Sep 7 2023 at 12:55:03 PM +0200, Fabio Valentini
wrote:
Sure, but that means it will still be started on Fedora with default
configuration, unless I misunderstand something?
It will. D-Bus services are a little weird because they often ship
systemd services but they're still
On Thu, Sep 7, 2023 at 12:53 PM Richard Hughes wrote:
>
> On Thu, 7 Sept 2023 at 11:36, Fabio Valentini wrote:
> > All systemd services that have an "enabled by default" preset need to do
> > that:
> >
On Thu, 7 Sept 2023 at 11:36, Fabio Valentini wrote:
> All systemd services that have an "enabled by default" preset need to do that:
> https://docs.fedoraproject.org/en-US/packaging-guidelines/DefaultServices/#_enabling_services_by_default
It's not exactly enabled by default -- it's autostarted
On Thu, Sep 7, 2023 at 8:00 AM Richard Hughes wrote:
>
> On Wed, 6 Sept 2023 at 21:32, Adam Williamson
> wrote:
> > There is no -lib package split in Fedora currently. The 'passim'
> > package provides the libraries.
> > https://koji.fedoraproject.org/koji/buildinfo?buildID=2278800 - there
> >
On Wed, Sep 06, 2023 at 07:58:24AM -0400, Stephen Smoogen wrote:
> On Fri, 25 Aug 2023 at 13:31, Richard Hughes wrote:
>
> > On Fri, 25 Aug 2023 at 16:27, Stephen Smoogen wrote:
> > > It depends on the scanning from ports open to unknown shared files to
> > 'why did our network costs go up so
On Wed, 6 Sept 2023 at 21:32, Adam Williamson
wrote:
> There is no -lib package split in Fedora currently. The 'passim'
> package provides the libraries.
> https://koji.fedoraproject.org/koji/buildinfo?buildID=2278800 - there
> is no 'passim-libs'.
Oopa, sorry. I pushed the commit:
commit
On Wed, 2023-09-06 at 22:33 +0100, Jonathan Dieter wrote:
> On Fri, 2023-08-25 at 12:42 +0100, Richard Hughes wrote:
> > The tl;dr: is I want to add a mDNS server that reshares the public
> > firmware update metadata from the LVFS on your LAN. The idea is that
> > rather than 25 users in an office
On Fri, 2023-08-25 at 12:42 +0100, Richard Hughes wrote:
> The tl;dr: is I want to add a mDNS server that reshares the public
> firmware update metadata from the LVFS on your LAN. The idea is that
> rather than 25 users in an office downloading the same ~2MB file from
> the CDN every day, the
On Wed, 2023-09-06 at 21:09 +0100, Richard Hughes wrote:
> On Wed, 6 Sept 2023 at 19:12, Adam Williamson
> wrote:
> > This message says you're "thinking of adding Passim", but in point of
> > fact, it appears to have been added to the package set already, and as
> > of fwupd-1.9.5-2.fc40 (built
On Wed, 6 Sept 2023 at 18:36, Adam Williamson
wrote:
> NetworkManager has the same concept, but of course it depends on apps
> that use data *caring* about it.
It sounds like passimd should be a thing that cares too --
https://github.com/hughsie/passim/issues/13
Richard.
On Wed, 6 Sept 2023 at 19:12, Adam Williamson
wrote:
> This message says you're "thinking of adding Passim", but in point of
> fact, it appears to have been added to the package set already, and as
> of fwupd-1.9.5-2.fc40 (built two days ago), fwupd hard requires it,
It hard requires the -lib --
On Fri, 2023-08-25 at 12:42 +0100, Richard Hughes wrote:
> Hi all,
>
> I was thinking of adding Passim as a default-installed and
> default-enabled dep of fwupd in the Fedora 40 release. Before I create
> lots of unnecessary drama, is there any early feedback on what's
> described in
On Wed, 2023-09-06 at 07:58 -0400, Stephen Smoogen wrote:
> On Fri, 25 Aug 2023 at 13:31, Richard Hughes wrote:
>
> > On Fri, 25 Aug 2023 at 16:27, Stephen Smoogen wrote:
> > > It depends on the scanning from ports open to unknown shared files to
> > 'why did our network costs go up so much?'
>
On Wed, Sep 6, 2023 at 12:58 PM Stephen Smoogen wrote:
>
>
>
> On Fri, 25 Aug 2023 at 13:31, Richard Hughes wrote:
>>
>> On Fri, 25 Aug 2023 at 16:27, Stephen Smoogen wrote:
>> > It depends on the scanning from ports open to unknown shared files to 'why
>> > did our network costs go up so
On Fri, 25 Aug 2023 at 13:31, Richard Hughes wrote:
> On Fri, 25 Aug 2023 at 16:27, Stephen Smoogen wrote:
> > It depends on the scanning from ports open to unknown shared files to
> 'why did our network costs go up so much?'
>
> Surely if you're on a local network with bandwidth costs you'd
On Thu, 31 Aug 2023 at 23:13, Marcus Müller wrote:
> - using avahi for local peer discovery, how does this compare to good ole
> bittorrent with
> Protocol/Message Stream Encryption turned on, and DHT instead of a tracker?
I think more than a few places would ban/block/report bittorrent
traffic
Just realized:
- using avahi for local peer discovery, how does this compare to good ole bittorrent with
Protocol/Message Stream Encryption turned on, and DHT instead of a tracker?
- I guess the "self-signed certificate" discussion stems from the fact that TLS assumes
you have certificates –
On Thu, 31 Aug 2023 at 22:05, Björn Persson wrote:
> The document doesn't say what design decisions were made based on the
> assumption of a friendly network.
Well, I can certainly add them -- this discussion was started so I can
add any missing information.
> All of those design decisions need
Richard Hughes wrote:
> I was thinking of adding Passim as a default-installed and
> default-enabled dep of fwupd in the Fedora 40 release. Before I create
> lots of unnecessary drama, is there any early feedback on what's
> described in https://github.com/hughsie/passim/blob/main/README.md
>
On Wed, 2023-08-30 at 09:11 +0100, Peter Robinson wrote:
> On Mon, Aug 28, 2023 at 9:50 PM Simo Sorce wrote:
> >
> > On Mon, 2023-08-28 at 15:14 -0500, Chris Adams wrote:
> > > Once upon a time, Richard Hughes said:
> > > > On Mon, 28 Aug 2023 at 16:27, Leon Fauster via devel
> > > > wrote:
>
Once upon a time, Peter Robinson said:
> What about integration with Let's Encypt as an option, the cert
> registration/renewal process is then pretty automated.
Since this is about desktop systems on internal networks, they probably
won't have pre-existing DNS entries, so Let's Encrypt is not
On Mon, Aug 28, 2023 at 9:50 PM Simo Sorce wrote:
>
> On Mon, 2023-08-28 at 15:14 -0500, Chris Adams wrote:
> > Once upon a time, Richard Hughes said:
> > > On Mon, 28 Aug 2023 at 16:27, Leon Fauster via devel
> > > wrote:
> > > > whats the benefit of this "self-signed TLS certificate" (as it
On Tue, 29 Aug 2023 at 21:03, Simo Sorce wrote:
> You could have deltas, so that clients will not get the whole thing
> every day, but deltas compared to what they have already (which would
> be 0 bytes if thy are up to date).
I'm trying to reduce the number of CDN accesses and the number of
On Tue, 2023-08-29 at 20:07 +0100, Richard Hughes wrote:
> On Tue, 29 Aug 2023 at 18:54, Simo Sorce wrote:
> > That depends on how you are going to handle re-installs of peers in the
> > network where the certificate will start mismatching ...
>
> In event of a mismatch I was going to ignore the
On Tue, 2023-08-29 at 20:05 +0100, Richard Hughes wrote:
> On Tue, 29 Aug 2023 at 17:06, Vít Ondruch wrote:
> > The point was that `fwupdmgr get-devices` lists ~32 devices for my LP. I
> > can't imagine that the metadata for these 32 devices would take 2 MBs.
> > That is more likely data for all
On Tue, 29 Aug 2023 at 18:54, Simo Sorce wrote:
> That depends on how you are going to handle re-installs of peers in the
> network where the certificate will start mismatching ...
In event of a mismatch I was going to ignore the peer; in most home
networks there'll be dozens of devices all
On Tue, 29 Aug 2023 at 17:06, Vít Ondruch wrote:
> The point was that `fwupdmgr get-devices` lists ~32 devices for my LP. I
> can't imagine that the metadata for these 32 devices would take 2 MBs.
> That is more likely data for all devices ever supported.
It is the metadata for every device --
On Monday, 28 August 2023 22:07:50 BST Richard Hughes wrote:
> On Mon, 28 Aug 2023 at 21:50, Simo Sorce wrote:
>
> > It could be improved by using TOFU, so that the window of impersonation
> > is small, but requires clients to cache an association and then has
> > weird failure modes to be dealt
On Mon, 2023-08-28 at 22:07 +0100, Richard Hughes wrote:
> On Mon, 28 Aug 2023 at 21:50, Simo Sorce wrote:
> > It could be improved by using TOFU, so that the window of impersonation
> > is small, but requires clients to cache an association and then has
> > weird failure modes to be dealt with
Dne 28. 08. 23 v 21:52 Richard Hughes napsal(a):
On Mon, 28 Aug 2023 at 15:53, Vít Ondruch wrote:
Sorry, I am probably missing something, but how this would help my
computer (or three I have at home)?
One computer downloads the 2MB from the CDN and the other two download
it from the first
On Mon, 28 Aug 2023 at 16:02, Richard Hughes wrote:
> On Fri, 25 Aug 2023 at 12:42, Richard Hughes wrote:
> > I was thinking of adding Passim as a default-installed and
> > default-enabled dep of fwupd in the Fedora 40 release. Before I create
> > lots of unnecessary drama, is there any early
On Mon, 28 Aug 2023 at 21:50, Simo Sorce wrote:
> It could be improved by using TOFU, so that the window of impersonation
> is small, but requires clients to cache an association and then has
> weird failure modes to be dealt with if one of the actors get re-imaged
> or changes the cert for any
On Mon, 28 Aug 2023 at 21:14, Chris Adams wrote:
> Without identification though, it doesn't do that, because there's no
> way for client B to know it is really talking to client A - it could be
> talking to client C with a man-in-the-middle attack and a different
> self-signed cert pretending to
On Mon, 2023-08-28 at 15:14 -0500, Chris Adams wrote:
> Once upon a time, Richard Hughes said:
> > On Mon, 28 Aug 2023 at 16:27, Leon Fauster via devel
> > wrote:
> > > whats the benefit of this "self-signed TLS certificate" (as it does
> > > not provide any "security")? Is this stub for
Once upon a time, Richard Hughes said:
> On Mon, 28 Aug 2023 at 16:27, Leon Fauster via devel
> wrote:
> > whats the benefit of this "self-signed TLS certificate" (as it does
> > not provide any "security")? Is this stub for something later ... ?
>
> It's a good question. It provides encryption
On Fri, 25 Aug 2023 at 12:42, Richard Hughes wrote:
> I was thinking of adding Passim as a default-installed and
> default-enabled dep of fwupd in the Fedora 40 release. Before I create
> lots of unnecessary drama, is there any early feedback on what's
> described in
On Mon, 28 Aug 2023 at 11:05, Petr Pisar wrote:
> V Fri, Aug 25, 2023 at 07:34:35PM +0100, Richard Hughes napsal(a):
> > you need to reboot into the new firmware before the published firmware gets
> > shared;
> Won't this suppress an effeciency of the local sharing?
Yes -- but it's a compromise
On Mon, 28 Aug 2023 at 16:27, Leon Fauster via devel
wrote:
> whats the benefit of this "self-signed TLS certificate" (as it does
> not provide any "security")? Is this stub for something later ... ?
It's a good question. It provides encryption (so client A can provide
the file to client B
On Mon, 28 Aug 2023 at 15:53, Vít Ondruch wrote:
> Sorry, I am probably missing something, but how this would help my
> computer (or three I have at home)?
One computer downloads the 2MB from the CDN and the other two download
it from the first computer. This saves you 4MB in bandwidth, and
Hi Richard,
Am 25.08.23 um 13:42 schrieb Richard Hughes:
Hi all,
I was thinking of adding Passim as a default-installed and
default-enabled dep of fwupd in the Fedora 40 release. Before I create
lots of unnecessary drama, is there any early feedback on what's
described in
Sorry, I am probably missing something, but how this would help my
computer (or three I have at home)? Why should I have anything like this
installed on my computer(s)? Why they should talk "secretly" to each other?
And why there is need to download ~2 MB of data every day? My laptop has
just
Le samedi 26 août 2023 à 15:14 +0100, Peter Robinson a écrit :
>
> In a lot of corporate datacentre networks the "users" on the network
> would know what the network is comprised of, and often on these
> networks they will have 10s, 100s of even 1000s of identical devices
> where being able to do
V Fri, Aug 25, 2023 at 07:34:35PM +0100, Richard Hughes napsal(a):
> you need to reboot into the new firmware before the published firmware gets
> shared;
Won't this suppress an effeciency of the local sharing? If a typical period
between a download and the reboot is significantly longer than a
On Fri, Aug 25, 2023 at 7:35 PM Richard Hughes wrote:
>
> On Fri, 25 Aug 2023 at 19:26, Marcus Müller wrote:
> > I fully agree with that assessment. "Here's a knob you turn that has the
> > potential to make
> > your firmware update 2s faster and is generally good for the ecosystem, but
> >
On 8/25/23 20:24, Richard Hughes wrote:
> On Fri, 25 Aug 2023 at 16:00, Benson Muite wrote:
>> Better as optional rather than default-enabled. It would likely be
>> helpful for computers in an institutional setting where the LAN is well
>> controlled.
>
> So that's the thing; if it's default
That sounds very good, and having a libs package desirable anyway should more
consumers pop up
On 25.08.23 20:43, Richard Hughes wrote:
On Fri, 25 Aug 2023 at 19:34, Richard Hughes wrote:
Yes, that's what I have right now. I do need to split out a
passim-libs so that you can remove the
On Fri, 25 Aug 2023 at 19:34, Richard Hughes wrote:
> Yes, that's what I have right now. I do need to split out a
> passim-libs so that you can remove the daemon and just leave the tiny
> client library.
Something like this perhaps?
diff --git a/passim.spec b/passim.spec
index bc51e57..3ad7ccc
On Fri, 25 Aug 2023 at 19:26, Marcus Müller wrote:
> I fully agree with that assessment. "Here's a knob you turn that has the
> potential to make
> your firmware update 2s faster and is generally good for the ecosystem, but
> you will have
> set it on every machine you set up" will not lead to
Hi Richard,
On 25.08.23 19:24, Richard Hughes wrote:
So that's the thing; if it's default disabled then I can say with
certainty that almost nobody will use it and we won't see any
reduction in network traffic at all.
I fully agree with that assessment. "Here's a knob you turn that has the
On Fri, 25 Aug 2023 at 15:59, Peter Robinson wrote:
> Is this something where you could enable it on one specific device and
> have a systemd time to pull the metadata and it advertises it to the
> network so you can designate a single device to run the service?
Yes, not a bad idea at all. Can
On Fri, 25 Aug 2023 at 16:27, Stephen Smoogen wrote:
> It depends on the scanning from ports open to unknown shared files to 'why
> did our network costs go up so much?'
Surely if you're on a local network with bandwidth costs you'd turn
off avahi or lock down the firewall? Lots of stuff blasts
On Fri, 25 Aug 2023 at 16:00, Benson Muite wrote:
> Better as optional rather than default-enabled. It would likely be
> helpful for computers in an institutional setting where the LAN is well
> controlled.
So that's the thing; if it's default disabled then I can say with
certainty that almost
On Fri, 25 Aug 2023 at 10:31, Richard Hughes wrote:
> On Fri, 25 Aug 2023 at 13:19, Stephen Smoogen wrote:
> > My understanding was that Microsoft found their own 'share updates' not
> working as much as expected
>
> Hmm, I heard the opposite; can you give any more info? They have way
>
No, I
On 8/25/23 14:42, Richard Hughes wrote:
> Hi all,
>
> I was thinking of adding Passim as a default-installed and
> default-enabled dep of fwupd in the Fedora 40 release. Before I create
> lots of unnecessary drama, is there any early feedback on what's
> described in
On Fri, Aug 25, 2023 at 12:43 PM Richard Hughes wrote:
>
> Hi all,
>
> I was thinking of adding Passim as a default-installed and
> default-enabled dep of fwupd in the Fedora 40 release. Before I create
> lots of unnecessary drama, is there any early feedback on what's
> described in
On Fri, 25 Aug 2023 at 13:19, Stephen Smoogen wrote:
> My understanding was that Microsoft found their own 'share updates' not
> working as much as expected
Hmm, I heard the opposite; can you give any more info? They have way
more telemetry than we do, and I was told it would not "be feasible"
On Fri, Aug 25, 2023, at 7:42 AM, Richard Hughes wrote:
> Hi all,
>
> I was thinking of adding Passim as a default-installed and
> default-enabled dep of fwupd in the Fedora 40 release. Before I create
> lots of unnecessary drama, is there any early feedback on what's
> described in
On Fri, 25 Aug 2023 at 13:50, Petr Pisar wrote:
> I see you wrote "metadata". It's not the firmware itself .Sill my concert is
> the same: what's a license of the metada? Can I redistribute them?
The metadata is explicitly CC0-1.0 -- but even if we later did
firmware one of the things I
V Fri, Aug 25, 2023 at 12:42:34PM +0100, Richard Hughes napsal(a):
> The tl;dr: is I want to add a mDNS server that reshares the public
> firmware update metadata from the LVFS on your LAN.
I see you wrote "metadata". It's not the firmware itself .Sill my concert is
the same: what's a license of
On Fri, 25 Aug 2023 at 07:44, Richard Hughes wrote:
> Hi all,
>
> I was thinking of adding Passim as a default-installed and
> default-enabled dep of fwupd in the Fedora 40 release. Before I create
> lots of unnecessary drama, is there any early feedback on what's
> described in
Hi all,
I was thinking of adding Passim as a default-installed and
default-enabled dep of fwupd in the Fedora 40 release. Before I create
lots of unnecessary drama, is there any early feedback on what's
described in https://github.com/hughsie/passim/blob/main/README.md
please.
The tl;dr: is I
61 matches
Mail list logo
