Slidge - XMPP gateways - 2025-02-16


  1. Orjan

    Peter, deuill might have meant that any new connected device, be it the app for smartphone or desktop, is considered by WA as _primary_ therefore never gets untied and doesn't lose connection with the ''host'' — no matter if it's switched on or off.

  2. Orjan

    To make slidged wa2xmpp accounts continuously work we still need to open official WA at least once in a fortnight.

  3. ThUnD3r|Gr33n

    https://emevth.no-ip.biz/file_share/kcNPhyJDJUQUhWshSO3gn4lg/zb2rhYdi1D4KtLBa31RqzGYXziSArdGa9Mox7i5jz69j7SAnr.jpg

  4. ThUnD3r|Gr33n

    Say hello to my new lover

  5. ThUnD3r|Gr33n

    65 kmh

  6. ThUnD3r|Gr33n

    😎🤞🤣💪

  7. nicoco

    > nicoco: If you wish I can try registering for you using a genuine iPhone, that will probably work and then you can try to overtake the number back using the WhatsApp message OTP instead of the phone verification. Thanks for the offer but I have a stock android at hand too.

  8. c3p0

    nicoco pushed 6 commit(s) to slidge - fix(avatar): detached instance error on startup (nicoco) - compat(cheogram): allow group creation (nicoco) - style(commitlint): add 'compat' and 'cfix' headers (nicoco) - fix(avatar): do not crash on startup if component avatar is unavailable (nicoco) - fix(avatar): do not attempt to read content of failed downloads (nicoco) https://codeberg.org/slidge/slidge/compare/3bd3b11d9..044cce6b2

  9. ThUnD3r|Gr33n

    What is conpat?

  10. nicoco

    A workaround for a cheogram bug that we'll remove once cheogram fixes it

  11. gllmhyt

    Hi, good morning everyone

  12. gllmhyt

    I'm wondering: I use the Debian packages at the moment

  13. gllmhyt

    And I don't want to do the work to switch over to podman

  14. gllmhyt

    Is it best to stick to the packages or to switch to pipx?

  15. ThUnD3r|Gr33n

    gllmhyt: just use docker

  16. gllmhyt

    Using podman or docker is the same for me. I don't like it and I don't want to switch to it 😉

  17. gllmhyt

    (like, a whole new brain gymnastics to learn, for the (t)rusty ol' server I manage, is kinda too much)

  18. ThUnD3r|Gr33n

    > Using podman or docker is the same for me. I don't like it and I don't want to switch to it 😉 With docker you will never have package issues .. compile issues ..just run and enjoy

  19. ThUnD3r|Gr33n

    Nothing complicated to learn about

  20. Menel

    What's wrong with the debain package? I use it too

  21. Menel

    Works

  22. gllmhyt

    Oh, nothing wrong, just wondering if it's recommended

  23. Menel recommends

  24. deuill

    AFAIK PIP/Wheel, Debian, and Docker/Podman are all supported methods of installation

  25. deuill

    If any of these doesn't work, we should/will fix

  26. gllmhyt

    OK thank you 🙂

  27. nicoco

    Debian is a bit awkward because there are official slidge and slidge-matridge packages in debian main (testing and unstable), and we also provide a different slidge package (bundling all working bridge).

  28. nicoco

    Also builds aren't automated anymore for this one, I need to move the thing to codeberg too.

  29. lissine

    https://slidge.im is missing links to slidgnal, slidcord, and sleamdge

  30. lissine

    Also, you could sort the slidges from best-working to least best-working, for example like the sorting here https://www.nicoco.fr/blog/2024/04/09/slidge-first-release/#slidge-legacy-modules

  31. lissine

    And a minor note: the codeberg repos for skidge and matridge (and slidge) are missing the slidge label: https://codeberg.org/explore/repos?q=slidge&topic=1

  32. nicoco

    slidgnal is completely broken now, but slidcord and sleamdge should be added yes

  33. nicoco

    the reason they aren't is because deuill worked on the landing page before they were moved to codeberg :)

  34. nicoco

    I'll add the labels

  35. nicoco

    thanks

    🙏 1
  36. nicoco

    The website is https://codeberg.org/slidge/pages if you want to do some HTML 👀

  37. lissine

    You could also add a favicon for https://slidge.im

  38. gllmhyt

    > Debian is a bit awkward because there are official slidge and slidge-matridge packages in debian main (testing and unstable), and we also provide a different slidge package (bundling all working bridge). Oh didn't know they were in the debian repositories, I just went to the slidge.im repo

  39. nicoco

    if you're on stable and/or want other bridges than matridge, I'd say our repo is better.

  40. gllmhyt

    I am considering moving my server to Arch because I might love to live dangerously

    🎩 1
  41. gllmhyt

    But yeah, debian stable atm

  42. nicoco

    ideally someone could look at how slidge-matridge is packaged and do the rest ^^

  43. gllmhyt

    The package has been made by debacle@movim.eu

  44. nicoco

    Yes, cool dude, he's often around ;)

  45. Hugo

    Re slidgnal: I think there's two approaches with presage: write Python binding for presage, or write Rust bindings for slidge.

  46. nicoco

    approach 2 sounds awkward because it would make more sense for slidge core to be written in high perf rust with bindings in other languages, but hey, it sounds fun anyway!

  47. nicoco

    Also I suspect there are more use cases for a good python lib for signal than there are easy to use rust libs for various crappy silos.

    👍 1
  48. Hugo

    Are you suggesting rewriting slide in Rust? 🚀

  49. nicoco

    So probably investing time for python bindings for presage is interesting outside of slidge too

  50. nicoco

    > Are you suggesting rewriting slide in Rust? 🚀 well, once xmpp-rs and the rest is top notch and, small details, I understand the first thing about rust, why not? xD

  51. nicoco

    > Are you suggesting rewriting slide in Rust? 🚀 well, once xmpp-rs and the rest is top notch and, small detail, I understand the first thing about rust, why not? xD

  52. Link Mauve

    Which features are you still missing from xmpp-rs?

  53. nicoco

    I was kidding, the "small detail" about me not understanding much of rust is actually the reason why that won't happen. It would be super cool if someone did a slidge-equivalent in rust, though :)

  54. Hugo

    So `aiosignald.SignaldAPI` runs its own event loop or whatever in background, and dispatches async events via methods like handle_IncomingMessage, right? Essentially, you could say that it's based on async callbacks.

  55. nicoco

    Hugo: I think it's the same event loop. Most of aiosignald is generated from the schema https://signald.org/protocol.json

  56. c3p0

    nicoco pushed 2 commit(s) to slidge - feat: customisable component name and welcome message (nicoco) - docs(config): grammar (nicoco) https://codeberg.org/slidge/slidge/compare/044cce6b2..3c69945ca

  57. Link Mauve

    nicoco, wanna work on it together?

  58. nicoco

    It is tempting but probably a recipe for a burnout (from me), especially since I start a new job (the kind that pays for food and rent and boring stuff) very soon…

  59. Hugo

    Right, not its own, the same event loop. Presage is async too, and pyo3 allows running rust async jobs in python's asyncio event loop.

  60. Hugo

    So I think that the bindings could expose a function to start the task that handles events both ways.

  61. Hugo

    Presage does keep its own message store. I think for our use we can trim it as soon as messages are relayed to the XMPP gateway, yes?

  62. nicoco

    yes, 1:1 message archives are handled by the XMPP server, and group archives are handled by slidge core

  63. nicoco

    Hugo, I had no idea you'd be interested in looking into this, this may have been a good fit for a google summer of code thing, but I think it's too late now :(

  64. Hugo

    I can't promise I'll have the time to see the whole thing through. It's definitely interesting tho.

  65. nicoco

    Anyway, if there are presage python bindings, I'm willing to do the work of rewriting slidgnal for it (unless you want to do both obv)

  66. Hugo

    The slidgnal part would be much faster if done by someone already familiar with the codebase and with slidge.

  67. nicoco

    yup, that's me lol

  68. nicoco

    but having someone try to use slidge the lib with no prior knowledge would probably yield some interesting feedback about counter-intuitive or awkward or surprising parts of its API (spoiler alert: there are all of those)

    🤣 1
  69. Hugo

    I'm not that excited about Python these days (though I don't hate it and it's been my career for like a decade). Go and Rust have spoiled me with static types and development tools that catch silly mistakes :P

  70. Hugo

    I'll peek further at presage; I don't think the bindings should be that tricky, but I've definitely said that before

  71. Hugo

    Not too hyped about presage having 540 dependencies 🙄

  72. nicoco

    SavagePeanut had a look at it a while ago (he's MIA now) and said the tricky part is that presage is designed as single user, so it might require some extra plumbing to support multiuser use. maybe just a matter of pointing to different "stores" or something like that

  73. nicoco

    however most slidge deployments are probably single user, so some people would probably use it anyway :)

  74. Hugo

    > Which features are you still missing from xmpp-rs? Can it do OMEMO? I would be super interesting to have gateways that can do e2e encryption (and then you run those gateways at home or something).

  75. nicoco

    (I know I would)

  76. gllmhyt

    > https://codeberg.org/slidge/matridge/issues/10 I searched around and there's no icon with a solid background. Could the COMPONENT_AVATAR be a base64 encoding?

  77. Hugo

    > SavagePeanut had a look at it a while ago (he's MIA now) and said the tricky part is that presage is designed as single user, so it might require some extra plumbing to support multiuser use. maybe just a matter of pointing to different "stores" or something like that I missed this detail! I think that presage would expose a single-user API, and you'd need a separate async task (handling its own connection) for each user.

  78. nicoco

    Hugo: that does not sound like a massive problem _a priori_

  79. nicoco

    about e2ee gateway, have a look at https://xmpp.org/extensions/inbox/gateway-relayed-encryption.html

  80. nicoco

    would be super cool to have, but most likely very tricky, and you _know_ that we'll have "cryptographers" all around shitting on any implementation because "the only real thing is signal and all the rest is amateur work"

  81. nicoco

    > I searched around and there's no icon with a solid background. Could the COMPONENT_AVATAR be a base64 encoding? no but it can be a path directly in the repo

  82. gllmhyt

    https://upload.postblue.info:5281/file_share/067b1f15-e7f4-7534-8bce-64a8d0eb860c/matrix.png

  83. gllmhyt

    Here's one made with the favicon.svg

  84. nicoco

    I hope [matrix] won't sue us for copyright violation or something if we include a png of their logo in the matridge repo

  85. gllmhyt

    About it, check https://commons.wikimedia.org/wiki/File:Matrix_icon.svg

  86. nicoco

    more seriously, we probably want to check with debacle if the license of this [m] is compatible with debian, I don't want to make his life harder

  87. gllmhyt

    "This logo image consists only of simple geometric shapes or text. It does not meet the threshold of originality needed for copyright protection, and is therefore in the public domain. Although it is free of copyright restrictions, this image may still be subject to other restrictions. See WP:PD § Fonts and typefaces or Template talk:PD-textlogo for more information."

  88. nicoco

    oh ok cool then

  89. nicoco

    gllmhyt, want to do the commits for matridge? I have a feeling you are about to become a rockstar python dev very soon

  90. gllmhyt

    I don't even understand how to fork on codeberg x)

  91. gllmhyt

    But yeah I can try a commit

  92. nicoco

    I replied with a link to edit with the codeberg web ui on the issue about the component name

  93. Hugo

    > about e2ee gateway, have a look at https://xmpp.org/extensions/inbox/gateway-relayed-encryption.html I feared this was headed into "just use wasm encrypters" territory. Not sure how I feel about that.

  94. gllmhyt

    OK!

  95. gllmhyt

    La page que vous essayez d'atteindre n'existe pas ou vous n'êtes pas autorisé à la voir.

  96. gllmhyt

    🙁

  97. gllmhyt

    I'm trying smthing

  98. Hugo

    There doesn't seem to be a usage policy on the Matrix logo yet: https://github.com/matrix-org/matrix.org/issues/2353#issuecomment-2163508975

    👍 1
  99. nicoco

    > La page que vous essayez d'atteindre n'existe pas ou vous n'êtes pas autorisé à la voir. oh right, you need to fork in your namespace first apparently, I was hoping the link would do it all

  100. nicoco

    you probably need to use https://codeberg.org/slidge/matridge/fork first then the link will work?

  101. gllmhyt

    Ah that's what I was looking for thx

  102. nicoco

    «bifurcation» in French

  103. c3p0

    postblue opened a pull request for matridge Fix component name https://codeberg.org/slidge/matridge/pulls/11

  104. gllmhyt

    Writing in english, reading the UI in french while listening to a live coverage in dutch, all is good

    🤣 1
  105. Kris

    >> about e2ee gateway, have a look at https://xmpp.org/extensions/inbox/gateway-relayed-encryption.html > I feared this was headed into "just use wasm encrypters" territory. Not sure how I feel about that. For now this is basically only useful for pgp, so Deltachat and some xmpp clients.

  106. nicoco

    Why wouldn't it work for signal too?

  107. nicoco

    Hugo: I think the only 2 other options are (a) a standardized encryption and formatter scheme (not sure this the correct terminoly) or (b) XMPP clients implementing encryption used in other networks. What's wrong with "wasm encrypters" approach? (I don't know much about wasm and security and encryption in general)

  108. gllmhyt

    https://upload.postblue.info:5281/file_share/067b1f72-13e4-7894-92de-21878429cb75/RNYpUkl4Q6eAyKO-YAeRag.jpg

  109. gllmhyt

    https://upload.postblue.info:5281/file_share/067b1f72-15af-74e4-b631-d16d619014ea/c1jchNXFT8ulgMizhBpdOw.jpg

  110. gllmhyt

    Neither is good, the white background without borders is an eye sore

  111. nicoco

    well…

  112. nicoco

    also, gajim makes all avatars round

  113. nicoco

    wow you use matteridge too

  114. nicoco

    I thought I was the only user lol

  115. c3p0

    nicoco pushed 1 commit(s) to matridge - style: harmonise component name with other slidge-based gateways (Guillaume Hayot) https://codeberg.org/slidge/matridge/compare/0d72ce090..65640313f

    🥳 1
  116. gllmhyt

    The UX of framateam is so painful, yeah it was the first i configured

  117. gllmhyt

    > nicoco pushed 1 commit(s) to matridge > - style: harmonise component name with other slidge-based gateways (Guillaume Hayot) > > https://codeberg.org/slidge/matridge/compare/0d72ce090..65640313f The description is longer than the fix :^p

  118. c3p0

    nicoco closed a pull request for matridge Fix component name https://codeberg.org/slidge/matridge/pulls/11

  119. nicoco

    > The description is longer than the fix :^p Yes but it allowed me to check that CI on PRs and c3p0 work as intended, great! Although maybe I should add a check for the commit message in CI, but I know some people hate that sort of things and don't want to discourage contributions for something I can easily reword after the fact…

  120. nicoco

    So all in all, thanks gllmhyt! <3

    🥰 1
  121. gllmhyt goes back to flemish dudes riding road bikes in the mud

  122. nicoco

    > The UX of framateam is so painful, yeah it was the first i configured does it work well for you? I only Inria's mattermost (a French research institute) and haven't managed yet to get to a state where I don't need to restart it every few days/week. But Inria uses authentication via gitlab and expiring tokens, I suspect most mattermost deployments are more simple.

  123. nicoco

    > The UX of framateam is so painful, yeah it was the first i configured does it work well for you? I only use Inria's mattermost (a French research institute) and haven't managed yet to get to a state where I don't need to restart it every few days/week. But Inria uses authentication via gitlab and expiring tokens, I suspect most mattermost deployments are more simple.

  124. gllmhyt

    I use it read-only at the moment and it works

  125. gllmhyt

    But I think last time I tried to send a message it failed in some capacity

  126. gllmhyt

    OK, it just kicked me

    🤣 1
  127. Peter

    Friends, you do realize that purple-presage is a thing?

    🤯 1
  128. c3p0 removed by nicoco

    you weren't supposed to do that, man

  129. Peter

    https://github.com/hoehermann/purple-presage

    🤣 1
  130. Peter

    Just saying, maybe making Slidpurple would be a more sane approach.

  131. nicoco

    gllmhyt, oh well, thanks for the feedback. do not hesitate to open tickets! enjoy your mud thing ;)

  132. c3p0

    c3p0-slidge published a release for slidge v0.2.5 https://codeberg.org/slidge/slidge/releases/tag/v0.2.5

  133. Peter

    It would alsoAlso, you wouldn't have to rust for that I guess.

  134. lissine

    > Friends, you do realize that purple-presage is a thing? It does not support the modern™ features supported by Slidge https://github.com/hoehermann/purple-presage?tab=readme-ov-file#missing

    ™ 1
  135. nicoco

    Peter, rust to C to python through libpurple does not exactly sound sane to me :)

  136. a moderator removed a message

    you weren't supposed to do that, man

  137. gllmhyt

    nicoco: c'est retombé en marche

  138. gllmhyt

    https://upload.postblue.info:5281/file_share/067b1fbf-155c-7a88-a74f-4af9a390f410/_Mx7JHNvThmzOXXrXMLCeQ.jpg

  139. nicoco

    after a restart or not even required?

  140. gllmhyt

    Unregistered, restarted, the whole lot

  141. gllmhyt

    Lets see when it fails again

  142. Peter

    nicoco: Yeah well it is bullshit from a technical perspective, I thought more like to protect more people from becoming rusty.

  143. Peter

    Also, it wouldn't be that bad because while it is a technical disaster to have those 3 languages in this case, it would also open the world for a teams bridge.

  144. nicoco

    I agree that a slidge-purple thing would be nice to have to benefit from all the reverse-engineering work the pidgin/libpurple folks do. But part of the motivation for starting slidge was to enjoy modern chat features that libpurple does not provide. Also, I think XMPP can already be seen a "chat API" and I think it's nice to avoid going through the libpurple layer. About MS Teams I have bad news, I'm no longer interested in working on it myself as I got fired from the job where they used it. 🤷‍♂️

  145. nicoco

    slurple. That's a good name.

  146. nicoco

    s/seen a/seen as a/

  147. ThUnD3r|Gr33n

    https://emevth.no-ip.biz/upload/sWQHh153IyrtFDLDP6bvGcVo/chat_image.png

  148. ThUnD3r|Gr33n

    readme on codeberg shows still docker.io

  149. ThUnD3r|Gr33n

    which tag? codeberg.org/slidge/slidge-whatsapp: ? master ?

  150. nicoco

    oops, will fix. it's codeberg.org/slidge/slidge-whatsapp

  151. ThUnD3r|Gr33n

    tag?

  152. nicoco

    tag :latest by default, latest tagged release

  153. ThUnD3r|Gr33n

    latest? Breaking the habit ? Breaking the bridge?

  154. nicoco

    or :main if you are crazy in your head

  155. ThUnD3r|Gr33n

    main is bloody edge ?

  156. nicoco

    bleeding edge yes

  157. ThUnD3r|Gr33n

    I love t he bleeding edge

    🩸 1
  158. ThUnD3r|Gr33n

    hope you implemented some nice breaking stuff

  159. ThUnD3r|Gr33n

    like .... deactivated messaging

  160. ThUnD3r|Gr33n

    or.... deactivate ansering :P

  161. deuill

    nicoco: I'll review your change tonight, thanks for pushing. I think I might pick up a SIM card to test groups etc. with as well

    ❤ 1
  162. deuill

    Group affiliations and communities

  163. nicoco

    yes, affiliations changes in groups are broken, well they don't show up until restart

  164. ThUnD3r|Gr33n

    > oops, will fix. it's codeberg.org/slidge/slidge-whatsapp so basicall the url is always: codeberg.org/slidge/BRIDGE ?

    👍 1
  165. nicoco

    I'd say do not mess with changing affiliations until deuill reviews what he just said he'd do :)

  166. ThUnD3r|Gr33n

    > yes, affiliations changes in groups are broken, well they don't show up until restart I saw bleeding pacth / fix

  167. nicoco

    I'm the one who broke it, so blame me. It might have been what caused my spam flag (it got lifted so it's not that bad)

  168. ThUnD3r|Gr33n

    so latest is always the way to not break the neck ?

  169. nicoco

    well, in an ideal work, yes

  170. ThUnD3r|Gr33n

    well.. we are dealing with you .. maaster of breaking stuff so ....

  171. ThUnD3r|Gr33n

    ill pull latest :)

  172. ThUnD3r|Gr33n

    if it breaks... je vais de tenir responsable

  173. nicoco

    but in practice, right now :latest has this affiliation change issue that might have caused my temporary ban, although it may also not be related (it might have been adding a non-contact to a group that triggered it, we'll probably never know)

  174. ThUnD3r|Gr33n

    is that the fix to go ?

  175. nicoco

    > nicoco: I'll review your change tonight, thanks for pushing. I think I might pick up a SIM card to test groups etc. with as well well I'm the one that fucked it up so thank you for not firing me!

  176. ThUnD3r|Gr33n

    https://emevth.no-ip.biz/upload/whXKC6JMwVJ8py0ZGLSQDemP/chat_image.png

  177. ThUnD3r|Gr33n

    don't wanna get banned like an idiot on wa :P

  178. nicoco

    ThUnD3r|Gr33n, yes you can pull that tag if you want to experiment with adding and removing members to group at your own risk, please report back to tell us if it works :)

  179. ThUnD3r|Gr33n

    fuck off no! I will not do anything but just make groups stuff in legacy app :P

  180. ThUnD3r|Gr33n

    let's try and break and get bannnneeedddd

  181. ThUnD3r|Gr33n

    FULL RISK

  182. ThUnD3r|Gr33n

    it's updating the docker image .....

  183. c3p0

    nicoco opened a pull request for slidge-whatsapp docs(readme): fix container repo https://codeberg.org/slidge/slidge-whatsapp/pulls/17

  184. ThUnD3r|Gr33n

    I fear my wa spare phone might take fire and shoots down a Starlink satellite

  185. ThUnD3r|Gr33n

    :P

  186. ThUnD3r|Gr33n

    I registered on codeberg but cannot distribute... may i also have the right to make PR ?

  187. nicoco

    anybody can make pull requests, what did you try and where did it fail?

  188. ThUnD3r|Gr33n

    codeberg.org/slidge/matridge NOT FOUND?

  189. jsb

    Works for me

  190. ThUnD3r|Gr33n

    i think i fucked up :P

  191. jsb

    Well, I guess it depends on if you're talking about a docker image or the codeberg repo. I tapped on that URL and it worked :)

  192. ThUnD3r|Gr33n

    i use portainer and did not use advanced mode to change from docker.io to codeberg

  193. c3p0

    nicoco pushed 2 commit(s) to matridge - docs(e2ee): mention that it is broken (nicoco) - chore: update lockfile (nicoco) https://codeberg.org/slidge/matridge/compare/65640313f..35b99524f

  194. procrastinator

    hello

  195. procrastinator

    how do i get the slidgram binary ?

  196. procrastinator

    in slidgram repo, i did: ``` . ./venv/bin/activate pip install . ```

  197. procrastinator

    it only creates a slidge binary but no slidgram binary

  198. ThUnD3r|Gr33n

    slidgram is still on slidge 0.2.2 iso 0.2.4 ?

  199. ThUnD3r|Gr33n

    I am shocked

  200. ThUnD3r|Gr33n

    BUG or FEATURE? trying to join a WHatsapp group via adhoc will end up in: Unknow XMPP address.. with normal commands i can jin.

  201. Hugo

    > Hugo: I think the only 2 other options are (a) a standardized encryption and formatter scheme (not sure this the correct terminoly) or (b) XMPP clients implementing encryption used in other networks. What's wrong with "wasm encrypters" approach? (I don't know much about wasm and security and encryption in general) wasm runs on its own VM/interpreter, so now your clients either need to bundle a whole wasm runtime, or execute one (assuming that it is pre-installed) to encrypt messages. It's not terrible, but it does put a burden on client devs.

  202. lissine

    procrastinator, the easiest thing would be to install using pipx: ``` pipx install slidgram==0.0.0.dev0 \ --pip-args='--extra-index-url https://codeberg.org/api/packages/slidge/pypi/simple/ ``` If you want to edit the source code, you'll find it under $HOME/.local/share/pipx/venvs/slidgram/lib/python3.1?/site-packages/slidgram/

  203. lissine

    Since python is an interpreted language, editing the code and restarting slidgram is all you need for your changes to take effect

  204. ThUnD3r|Gr33n

    has someone a sip server for testing ?

  205. procrastinator

    lissine, thanks! i actually want to understand the build process and thus wanted to know how it works --- previously i just did pip install . and that created a slidgram binary however, i am curious about the other steps that's needed now

  206. procrastinator

    ok, looks like ```python3 -m slidgram -c config``` is what's supposed to work

  207. procrastinator

    there's a slidge binary which is just ``` from slidgram import main ... sys.exit(main()) ```

  208. procrastinator

    ok i'm dumb, it's just slidge -c config or, if your config doesn't define legacy-module, slidge --legacy-module slidgram -c config

  209. ThUnD3r|Gr33n

    does signal work again ?? did not follow up

  210. lissine

    No

  211. ThUnD3r|Gr33n

    ... found the anser

  212. ThUnD3r|Gr33n

    what the henk is going on.. we need fixing the workd !

  213. lissine

    Be change you want in the world!

  214. jsb

    Signal is (in)famously hostile to third party clients

  215. jsb

    signald, as far as I can tell, was basically done by trying to tease out the networking portion of the signal android client and that's not very sustainable

  216. ThUnD3r|Gr33n

    i agree...

  217. Hugo

    I think that presage relies on some of the same rust libs used by Signal upstream.

  218. Hugo

    Which makes it _slightly_ more sustainable.

  219. jsb

    When did they start using rust? Does it coincide with signald's death?

  220. jsb

    Of course, I never got signald to work right anyway

  221. Hugo

    In theory they might have great crypto, but their client is electron crap built with binaries from a bunch of various sources. It doesn't even encrypt credentials in a keyring (the claim that the problem is "impossible to solve", despite other clients having done so for years).

  222. jsb

    They invented the great crypto. But everyone else is using it now so I'm not sure if there's a clear advantage to signal. And I'm still bitter about them dumping e2ee SMS

  223. jsb

    Even though I don't use SMS anymore

  224. Hugo

    Everyone else uses this great crypto and then uploads backups of your conversatoins plaintext onto google drive or something like that.

  225. Hugo

    If only people used xmpp more

  226. debacle

    > more seriously, we probably want to check with debacle if the license of this [m] is compatible with debian, I don't want to make his life harder I could check, if the logo is maybe already used and packaged in Debian somewhere, i.e. as part of one of the packaged Matrix clients. In this case, I assume that any copyright questions are already answered :-))