XMPP Service Operators - 2024-02-09


  1. roughnecks

    sorry for the OT, but is it bad practice to remove a website binding to port 80 instead of having a redirect to 443?

  2. MattJ

    No, you can do that. I think many/some browsers will try 443 now if 80 is closed (you might want to confirm how prevalent that behaviour is)

  3. MattJ

    Also be sure you don't need port 80 for Let's Encrypt challenges

  4. roughnecks

    I'll leave port 80 open, but without nginx listening to that (this was the idea)

  5. Madiator2011

    Hi there! Anyone uses ejabberd cause I have some questions.

  6. Guus

    Madiator2011 there's a dedicated ejabberd chatroom: xmpp:ejabberd@conference.process-one.net?join

  7. f_

    Trung: "Connection from trung.fun to conference.vitali64.duckdns.org failed !" It was my server

  8. f_

    Fixed it

  9. Polarian

    f_: hmmm... I know a `f_`

  10. Polarian

    and they have XMPP...

  11. Polarian

    correlation]

  12. Polarian

    correlation?

  13. f_

    Is it the one on libera #librespeech?

  14. Polarian

    > Is it the one on libera #librespeech? yes

  15. f_

    It's me

  16. Polarian

    small world eh?

  17. f_

    nah

  18. Polarian

    There's only one Polarian out there... and its me :)

  19. f_

    Nice to see you here

  20. Polarian is everywhere

  21. Trung

    > Fixed it Beautiful ! Happy (Lunar) New Year !

  22. chunk

    Happy Friday :D

  23. f_

    :D

  24. Polarian

    > Happy Friday :D Friday is never happy... actually most days aren't

  25. Polarian

    maybe after a pint the day might be a little better :D

  26. wb

    why does the xmpp providers repo use kde invent >:(

  27. wb

    isnt this like an official thing now?

  28. Licaon_Kter

    wb: what's the issue?

  29. Polarian

    wait Licaon_Kter do you even host a XMPP server, don't you use conversations's server?

  30. Licaon_Kter

    Why would anyone?

  31. Licaon_Kter

    ...do either

  32. cal0pteryx

    wb: what's the matter with it?

  33. wb

    I simply want to add my instance to the list but I need to make another account which very much wants my full name just for a single issue.

  34. wb

    If it was an XMPP git instance, I would be willing since I actually use XMPP. I don't touch KDE.

  35. cal0pteryx

    I see, and I understand. Which provider would that be?

  36. wb

    worlio.com

  37. cal0pteryx

    wb: I'll add an issue for you

  38. wb

    ty ty

  39. Guus

    wb: maybe add support for https://xmppnetwork.goodbytes.im while you're at it?

  40. wb

    I had never seen this page before and seeing the pure blue circle as I load the page is amazing.

  41. Guus

    Super Secret work in progress 3d representation https://xmppnetwork.goodbytes.im/3d.html

  42. wb

    My computer cries as it loads each frame once a second.

  43. Polarian

    > Super Secret work in progress 3d representation https://xmppnetwork.goodbytes.im/3d.html funny enough had a lecture on graphs at uni today xD

  44. wb

    The most demanding video games don't break a sweat on this machine, but XMPP graphs? That's the gym for this sucker.

  45. Guus

    Polarian: oh so now you're qualified to improve this!

  46. Menel

    My mobile can do 2-3 fps ๐Ÿ™‚

  47. Polarian

    > Polarian: oh so now you're qualified to improve this! fuck no :)

  48. Guus

    Performance varies dramatically. I've seen phones on which it runs flawlessly

  49. Guus

    I'm just a crap frontend developer ๐Ÿ˜

  50. Polarian

    better than me :)

  51. Guus

    I did not create the 3d variant mind

  52. cal0pteryx

    The 3D one should be fine after it cooled down (5s)

  53. cal0pteryx

    The movement is what kills performance. Once it's stable, framerate goes up :)

  54. Guus

    He did ๐Ÿ˜

  55. Polarian

    Guus: why not render the "main" nodes, then click on them to see all the branches

  56. Polarian

    so any node with less than 50 edges gets hidden

  57. Polarian

    until "expanded"

  58. wb

    Make that optional. This is a good WebGL stresstest.

  59. Polarian

    > Make that optional. This is a good WebGL stresstest. more like a good way to kill a laptop

  60. wb

    Well I imagine the optimal default will be to hide.

  61. cal0pteryx

    > so any node with less than 50 edges gets hidden /me likes the wow effect when seeing everything at once

  62. Polarian

    >> so any node with less than 50 edges gets hidden > /me likes the wow effect when seeing everything at once and the sound of your laptop/phone/desktop dying

  63. Polarian

    buy a crypto mining setup to render the graphs

  64. cal0pteryx

    well, it very much depends on hardware

  65. Polarian

    and the browser

  66. Polarian

    Firefox is substantially slower

  67. Menel

    it's super smooth on phone and laptop for me after the initial few seconds. using firefox

  68. Polarian

    > it's super smooth on phone and laptop for me after the initial few seconds. using firefox what phone do you have?

  69. Menel

    only on the phone its only 1/4 of the screen, just like the 2d opengl version

  70. Menel

    pixel 7

  71. cal0pteryx

    I don't see a difference when using firefox/chrome. depends on whether gpu acceleration is invoked I guess

  72. Polarian

    > pixel 7 that's a modern phone, no wonder!

  73. Polarian

    I just realised all the messages in this room are logged :)

  74. Polarian gotta try not to seem stupid

  75. chunk

    lol

  76. chunk

    logged and searchable on the web

  77. Polarian

    > Super Secret work in progress 3d representation https://xmppnetwork.goodbytes.im/3d.html what's your future plans for all of this? get the XEP accepted

  78. Polarian

    > Super Secret work in progress 3d representation https://xmppnetwork.goodbytes.im/3d.html what's your future plans for all of this? get the XEP accepted?

  79. Guus

    No future plans other than messing about, see if we can make it better.

  80. Guus

    The XEP was accepted this week I think

  81. Polarian

    > The XEP was accepted this week I think ah cool

  82. wb

    Guus: The WebGL variant of this graph has an interesting bug on my browser. I cannot click without it being stuck in move mode.

  83. wb

    I added Pubsub to my instance though so it's on the list now.

  84. Bob Evans

    > logged and searchable on the web chunk: where.

  85. Menel

    Look at the room topic

  86. chunk

    Bob Evans: I can't remember, maybe an XSF site page, or something, maybe i'm wrong even and it's not this chat but a different xSF page

  87. chunk

    > Look at the room topic or this LoL

  88. Menel

    Ah, not in the topic https://xmpp.org/community/chat/

  89. Menel

    Community - - > logs

  90. chunk

    candy for the spies

  91. wb

    oooh these logs are nice

  92. Bob Evans

    Thank you.

  93. Bob Evans

    For the candy.

  94. wb

    even supports the replies

  95. Bob Evans

    Direct link: https://logs.xmpp.org/operators/

  96. wb

    I'm stealing ur chats

  97. wb

    piping them into ai

  98. Bob Evans

    > A general off-topic chat can be found at conversations-offtopic-reloaded@conference.trashserver.net. This statement should come with a Content Warning.

  99. Licaon_Kter

    Bob Evans: be the first to warning

  100. Polarian

    > This statement should come with a Content Warning. Why?

  101. Bob Evans

    The logs are in realtime. And includes join/part.

  102. Polarian

    its funny how there is a mailing list, yet nobody uses it xD

  103. Polarian

    Hm... maybe that should be a XEP

  104. wb

    I'm subscribed to them incase anybody does.

  105. Polarian

    a server should advertise that it is logging and a client MUST warn the user that the channel is logging

  106. wb

    I just think if you are worried about being logged, you should be careful with what you say.

  107. wb

    Any user can join a muc and just save the history as they lurk.

  108. Polarian

    true

  109. Polarian

    but it would still be nice to have a heads up that the server is logging :)

  110. Polarian

    welp I guess I will subscribe to the mailing list too

  111. wb

    And how would this be controlled? What is to say the owner of the muc or the server operator can't just lie.

  112. Polarian

    true

  113. Polarian

    I guess the mailing list would be useful though

  114. wb

    I think perhaps if we want it to be known, it could be added to the topic IMO.

  115. Polarian

    say if a server can't peer with this channel and need help from other operators

  116. Polarian

    then the mailing list would be a good fallback, no?

  117. wb

    I started getting more involved just a few days ago and the first thing I did was join the mailing lists that suited me.

  118. Menel

    I think at the xsf room they were discussing exactly that some days ago. A way rooms can warn the users about policies like this

  119. Bob Evans

    > Any user can join a muc and just save the history as they lurk. But the user doesn't publicly publish logs to the Internet.

  120. wb

    They could.

  121. wb

    Nothing is stopping them.

  122. Bob Evans

    Of course.

  123. Bob Evans

    I am more uneasy of join/part information being public than logs.

  124. wb

    I just feel like adding any sort of indication (in protocol) of logging can be misused to provide a false sense of security and privacy.

  125. neutrino

    The one slightly interesting thing about matrix: view.matrix.org . Where is similar for xmpp public mucs?

  126. wb

    ยป [15:51:23] <neutrino> The one slightly interesting thing about matrix: view.matrix.org . Where is similar for xmpp public mucs? Is that just a room search or something?

  127. wb

    https://search.jabber.network/

  128. neutrino

    >> [15:51:23] <neutrino> The one slightly interesting thing about matrix: view.matrix.org . Where is similar for xmpp public mucs? > Is that just a room search or something? It links to logs of each one.

  129. wb

    Oh.

  130. Bob Evans

    Logging of a MUC can't be prevented.

  131. Bob Evans

    But users should know when they're published to the Internet.

  132. neutrino

    Right, so all should be. Where are the others?

  133. wb

    In a perfect world, they would.

  134. Menel

    >> Is that just a room search or something? > It links to logs of each one. *most* xmpp rooms don't have public logging. Actually the only one I'm aware of are the xsf rooms and this

  135. Menel

    Xmpp people are more shy about their data and metadata then matrix people. That's also why most public rooms a semi-anon

  136. neutrino

    > In a perfect world, they would. In a perfect world, people would know to assume they are.

  137. Licaon_Kter

    > a server should advertise that it is logging and a client MUST warn the user that the channel is logging Bugs... https://issues.prosody.im/1424

  138. Bob Evans

    > I am more uneasy of join/part information being public than logs. Because those who don't speak are part of this published data, and yields no future value.

  139. Bob Evans

    Logs of a XEP discussion are valuable.

  140. wb

    You knew it's a public chat before you joined.

  141. wb

    If you do not take in consideration of logging then, the muc or network should not be to blame.

  142. Bob Evans

    Not join/part logs of lurkers who never speak.

  143. wb

    "before you joined"

  144. wb

    I do not see how that is any different than logging a mesage considering it only happens if you joined in the first place.

  145. Bob Evans

    Would you like me to sit in a vehicle on the public way, in front of your residence, and publish public events to the Internet.

  146. Bob Evans

    Of your leaving or arriving at your residence.

  147. wb

    My residence is not public.

  148. wb

    This muc is.

  149. Bob Evans

    Your presence on the public way is observable by any in public view.

  150. wb

    Letting people know a muc is logging doesn't make you less creepy.

  151. wb

    This is the Internet, your real life scenarios just can't match up with what is happening in a chatroom either.

  152. wb

    When people join or leave this muc, it's not then walking in or out of their house, it's them entering/leaving the park or something.

  153. Bob Evans

    > This is the Internet, your real life scenarios just can't match up with what is happening in a chatroom either. I find this to be the root of many privacy debates.

  154. Bob Evans

    That there's a distinct separation.

  155. neutrino

    If IPs were also published the join parts could be more interesting, like for locating bad connections.

  156. wb

    If you are concerned about privacy, Mr. Evans, may I suggest not using your name as your nick in a public muc that clearly shows it is logged on pages where it is advertised?

  157. wb

    And don't go spreading out identifyable information in public mucs either.

  158. Bob Evans

    > When people join or leave this muc, it's not then walking in or out of their house, it's them entering/leaving the park or something. Of course. That was my intent. Any place in public that you may be observed at.

  159. neutrino

    IRL some jurisdictions have rules against stalking people.

  160. wb

    My neighborhood is my network, and people drive by it and live on it. Watching someone on any neighborhood is creepy and stalking whether virtual or physical.

  161. chunk

    lmao

  162. chunk

    it's the internet, as soon as you ventured passed your LAN demarc point, you were subject to... the internet.

  163. chunk dies

  164. neutrino

    As if your ISP doesn't log your LAN join part from their connection.

  165. Bob Evans

    We're veering too far from the topic of the MUC.

  166. Bob Evans

    I'll leave it alone.

  167. Bob Evans

    In closing, logging is not an issue.

  168. Bob Evans

    Publishing it is, without notification.

  169. Bob Evans

    In my opinion.

  170. Martin

    Maybe put a note and a link to the logs into the description to make it clear to everyone joining.

  171. kapad

    a bad think with the logs, is that contains presences/pings log as just time+nick, no reason to be there

  172. kapad

    have see also in weblogs in ejabberd think also psi at first, and is not good

  173. kapad

    also make reading quiet difficult

  174. kapad

    hmm, that is chatlogs

  175. kapad

    also agree with Bob Evans' closing

  176. kapad

    maybe not an issue, but a matter

  177. Polarian

    > *most* xmpp rooms don't have public logging. Actually the only one I'm aware of are the xsf rooms and this Gajim is another one

  178. kapad both ejabberd's and psi's were very usefull. especially the ancient pages !

  179. Polarian

    > As if your ISP doesn't log your LAN join part from their connection. They do though...

  180. Polarian

    If you run ISP routers they withhold the right to access your home network, I don't even think they need permission to do so in some countries

  181. Polarian

    but I wouldn't trust off the shelf routers either...

  182. chunk

    paranoia detected

  183. chunk

    loljks

  184. kapad

    you smell it, eh ?

  185. Polarian

    I am pretty sure a large number of users in this MUC are also paranoid, you OMEMO fish huggers ๐Ÿ™ƒ

  186. chunk

    lmao

  187. kapad

    https://ua3.anondns.net/im-storage/c12ecf930e5b40178c396859b5a281f9c56430fe/bH7OHFQfIPjMeYSHvLWuLfB5lCAjE8KTTL4NvrsBJwyUD/shell-class.jpg, ``Shell Classes`` - that my last project ( some min ago ). took 20 secs to write the code, but lot to reach that speed

  188. kapad

    :)

  189. kapad

    ^ also the icon is wrong, cause wlx... is the wifi

  190. kapad nobody's perfect .

  191. Polarian

    kapad, what OS is the router running?

  192. kapad

    debian stable

  193. kapad

    this is my home pc, but also everything else

  194. kapad

    ` cat -- /proc/cpuinfo|grep name model name: Intel(R) Core(TM)2 Duo CPU E4500 @ 2.20GHz (x2)