Just started a list of compatible devices for #tangara on the forum. There's also an issue open asking for some place to add information about the tangara audio player.

I never owned or used an Apple iPod for listening to music and I really like the Tangara playing as an device independent of my phone.

If you're wondering about the FSF starting to work on Android and are interested in an opinion about the technical options they have about their endeavor, please read Caseys thread about it.

This is the insight I do not have, because I'm no developer.

If you do not follow her already on the fediverse or elsewhere: my impression is that she is bringing up devices with PostmarketOS at an unmatched frequency . This makes her opinion exceptional interesting, because she probably knows very well about which blobs we're talking.

After thinking a bit more about the situation and the idea of cleanroom reversing I wrote down this thought.

Original post deleted, because I accidentally used the wrong pronoun - sorry, I hope I didn't hurt your feelings, @cas@treehouse.systems! Thanks for making me aware, @kate@treehouse.systems!

https://chrichri.ween.de/o/cffc6733c9a74ae2a3ad3ccc8c8a0923

@fsf@hostux.social @fsfe@media.fsfe.org

Every few hours my thoughts start cycling around the idea to free Android.

About reversing blobs in a cleanroom I had the following thought:

I pretty well know the SAMBA project, because the company I work for delivers support for SAMBA and takes part in its development.

In my opinion SAMBA is an example for a very successful attempt of reverse engineering closed source. In the end Microsoft had to give in and started to work together with the SAMBA community in some kind of manner and everywhere you find smb you'll find any device (nas, workstation, whatever) running SAMBA or anything based on the SAMBA code.

This success took part in a very slow moving environment. Microsoft designed SMB and Active Directory for huge networks running in a stable and very slow moving manner. This gave the SAMBA team the time to get market shares for their FLOSS solution. They had the time to make it work to a point that people looking to talk to a Windows Network without having to run Windows and/or pay license fees started using SAMBA (and paying SAMBA developers to enhance and adopt the code).

Compared to the environment the blobs for Android devices live in it becomes clear that this SAMBA success factor - having time - will not apply for the FSFs project to reverse these blobs.

Developers of these reversed blobs will not be payed by any market participant to enhance or alter there blob replacement - I'm not able to think about any good reason. Companies wanting to sell a product for profit will just base their effort on Android or iOS. Why should they have interest to replace any of the working mobile ecosystems?

I'm still wondering why people at FSF opted for this direction instead of helping to build a new ecosystem running at our own speed making us independent of decisions taken elsewhere.

https://librephone.fsf.org/

Free the last bits of "anything Android"?

My positive thought about this is that #PostmarketOS and #MobileLinux in general will profit from any public knowledge about hardware it could run on.

If I understood correctly the money for doing the work on the #Android blobs is donated by John Gilmore. His Money, his decision where to put it. And there is a positive effect, but there is also a negative one:

Android is based on ideas by #Google. To free it we'd need to fork it and adopt it to different ideals and goals. Android is designed to maximize the profit of Google.

It is not designed with the users well-being and interests as the primary goal.

Just replacing blobs in Android keeps the ecosystem the same, promotes Google and their goals and leaves the control over design decisions for Android in Googles hands.

Once a device is freed by hard work of a few engineers it will be old, it will be uninteresting for people looking at Android and the latest shiny hardware running it.

But still - Mobile Linux will make good use of those devices as free OSes in general do when it comes to hardware left behind by commercial OSes.

@fsf@hostux.social @fsfe@media.fsfe.org

https://chrichri.ween.de/o/1d2b0563ff3045ebbe36f8cb1a90aa41

… so far so good. The mechanical part seemed to have gone well and #ModemManager recognizes a #5G connection. No tests, yet.

#ShotOnLibrem5

…let's see where this leads…

#shotonlibrem5

Just read this about hosting ones own email server:

https://mako.cc/copyrighteous/google-has-most-of-my-email-because-it-has-all-of-yours

One of the reasons to host my own email is that I do not want the big surveillance capitalists storing my emails and using them for their own good. Well, then...

Starting to learn #Proxmox. I thought I'd just replace the virt-manager on my desktop to get some hands on in an environment I only use for testing.

I failed.

The following I found out digging through documentation, forum and code:

• There needs to be a domain defined by the dns lookup for the local IP. Just a hostname alone doesn't seem to be sufficient. hostname -f should return a domain part.
• The node name (hostname) needs to be the same in various places
  • /etc/pve/.members (created from corosync.conf)
  • rrd file in /var/lib/rrdcached/db/pve2-node
• corosync.conf exists two times - in /etc/pve and in /etc/corosync - which one is used under which circumstances I didn't find out, yet
  • /etc/pve/corosync.conf is generated from the sqlite database in /var/lib/pve-cluster/config.db

For debugging I looked at this code to understand where the Status: unknown pop-over on the host does come from.

I added some logging there which I could trigger from the cli using pvesh ls nodes.

The variables used there where filled like this in the not working state:

member=node1
get_rrd_key=pve2-node/PureBlackSoul
status=unknown

The function get's the nodename $node shown on the gui from dns for the local IP I suspect. The $member hash seems to be filled with the data from /etc/pve/.members created by the corosync.service from - in my case - /etc/corosync/corosync.conf.

The value from $node is used to find the rrd data for the host.

The condition for online being:

There needs to be rrd-data that can be found by using the content of the $node argument to extract_node_stats and if %$members (/etc/pve/.members) is not empty it needs to contain a member named $node being marked online.

Short: if you host is Status: unknown dig into the following places:

• members hash: /etc/corosync/corosync.conf, /etc/pve/corosync.conf, /var/lib/pve-cluster/config.db
• dns hostname: nslookup <local IP not being 127.0.0.0/8 or interface lo>, nameserver or /etc/hosts, hostname -i to find IPs, hostname -f to check whether there is a domain part
• rrd files: /var/lib/rrdcached/db/pve2-node/ - filename should be present as the hostname without the domain

So let us establish our bearings. Uncertainty is not chaos but rather the necessary habitat of the present tense. We choose the fallibility of shared promises and problem solving over the certain tyranny imposed by a dominant power or plan because this is the price we pay for the freedom to will, which founds our right to the future tense

The Age of Surveillance Capitalism: The Fight for a Human Future at the New Frontier of Power
by Shoshana Zuboff

#SurveillanceCapitalism #privacy