XSF Discussion - 2020-09-24


  1. pep.

    https://runyourown.social

  2. pep.

    “How to run a small social network site for your friends”

  3. pep.

    Does pubsub allow a node to be marked/tagged/namespaced to say "here that's the kind of content you'll find and how you're supposed to handle it", or is it just a non-namespaced name?

  4. pep.

    pubsub#type ? :/

  5. pep.

    (node configuration)

  6. pep.

    unrelated, is it possible to have xmpp: NSs that actually point to the document?

  7. ralphm bangs gavel

  8. ralphm

    0. Welcome + Agenda

  9. pep.

    !

  10. ralphm

    Hi!

  11. ralphm

    I just sent the agenda, but it seems stuck somewhere.

  12. ralphm

    0) Welcome 1) Minute taker 2) txxmpp client on client list? 3) Elections 4) AOB 5) Date of Next 6) Close

  13. ralphm

    MattJ, Guus, Seve?

  14. Seve says hello!

  15. MattJ

    Hey

  16. guus.der.kinderen

    o/

  17. ralphm

    1. Minute taker

  18. pep.

    I'll do it

  19. ralphm

    Thanks pep.

  20. ralphm

    2. txxmpp client

  21. ralphm

    From what I understand, this is a command line tool, rather than an interactive chat client.

  22. Seve

    Corect

  23. Seve

    Correct

  24. ralphm

    I don't particularly see a problem with adding this, if the description just notes this point.

  25. guus.der.kinderen

    Is this a matter for board?

  26. pep.

    I also don't see any issue with that

  27. pep.

    I don't think so

  28. ralphm

    guus.der.kinderen, good question, probably not

  29. guus.der.kinderen

    We don't show descriptions of clients on the site, by the way

  30. pep.

    Even better :)

  31. ralphm

    I mean on the page linked

  32. Seve

    Guus well, it is an open PR I would like to close and I believe applying my only opinion is not fair. Because I would not have accepted the PR at this moment, because I personally feel it does not fit there.

  33. guus.der.kinderen

    I'm with you on that Seve,

  34. pep.

    Seve, what do you think fits there then?

  35. guus.der.kinderen

    The page linked is source code

  36. guus.der.kinderen

    there's no documentation.

  37. guus.der.kinderen

    I'd stick it with 'libraries'rather then 'clients' to not confuse 99% of the people on our website looking at that list

  38. pep.

    Is all this stuff referenced somewhere in our criteria? (do we have any public criteria?)

  39. guus.der.kinderen

    but to be honest, I don't care much either way

  40. guus.der.kinderen

    No, we do not.

  41. guus.der.kinderen

    although...

  42. pep.

    Well then it probably fits into no criteria

  43. guus.der.kinderen

    there's a bit of documentation in the stuff that processes the data files, I think.

  44. guus.der.kinderen

    anyway, let's not make a big deal out of this

  45. pep.

    sure

  46. guus.der.kinderen

    Happy to see webteam move forward on this how they see fit.

  47. ralphm

    +1

  48. guus.der.kinderen

    I'd stick it with libraries, but no-one made me king of anything.

  49. Seve

    Alright then

  50. pep.

    We should really get rid of these lists someday..

  51. pep.

    Anyway..

  52. ralphm

    To replace it with something else?

  53. pep.

    nothing

  54. pep.

    Not on xmpp.org at least

  55. ralphm

    Well, ok. That might be an interesting topic some other time, venue.

  56. pep.

    Yep

  57. ralphm

    3. Elections

  58. ralphm

    As you've seen Alex has put out the call for our upcoming Board and Council elections.

  59. ralphm

    Submission deadline is November 8.

  60. ralphm

    Please consider who you'd like to ping to serve on Board and/or if you want to stand (again) yourself.

  61. ralphm

    I don't have anything further on this myself.

  62. ralphm

    4. AOB

  63. ralphm

    Any OB?

  64. Seve

    None here 🙂

  65. pep.

    None from me

  66. guus.der.kinderen

    neither

  67. guus.der.kinderen

    maybe we can review some feedback/backburner topics next meeting?

  68. MattJ

    None here

  69. guus.der.kinderen

    in an effort to clean up the plate a bit

  70. ralphm

    Yeah, I already did a bit of clean up, but there's more that may be so obsolete that it can be removed.

  71. ralphm

    If you all could have a look at Trello, that'd be great. I'll add an agenda item for this

  72. ralphm

    6. Date of Next

  73. ralphm

    +1W

  74. ralphm

    7. Close

  75. ralphm

    Thanks all!

  76. MattJ

    Thanks!

  77. guus.der.kinderen

    Cheers

  78. ralphm bangs gavel

  79. Seve

    Super 🙂

  80. pep.

    Thanks

  81. pep.

    jonas’, vanitasvitae, https://xmpp.org/extensions/xep-0277.html#location there's already some escaping in URIs being done here btw

  82. lovetox

    pep., "Outside of native XMPP systems"

  83. lovetox

    of course you have to escape urls if you want them to be used by other applications

  84. pep.

    Is that an issue?

  85. lovetox

    if you have not allowed chars in your uri you have to escape them

  86. lovetox

    it does not matter if it happens in the context of xmpp or whatever

  87. lovetox

    otherwise its not a valid uri

  88. lovetox

    i dont see how this has anything to do with the question if you are allowed to put an escaped uri into a node attribute

  89. pep.

    I don't think we're talking about the same thing

  90. pep.

    are we

  91. pep.

    I'm talking about this: https://github.com/xsf/xeps/pull/983 and council's comments

  92. pep.

    Unrelated, I regret 0277 doesn't specify pubsub#type to be set to 'urn:xmpp:microblog:0' when not in PEP.. Any reason why? miss?

  93. pep.

    “A person's microblog SHOULD be located at a personal eventing (PEP) node named "urn:xmpp:microblog:0" but MAY be located at a generic publish-subscribe node that is not attached to a user's IM account.”

  94. pep.

    Would it make sense to change it now? :/

  95. pep.

    TIL microblog is not even Draft

  96. flow

    pep., sounds sensible to recommend that pubsub#type is set

  97. flow

    or even require that. but I don't want to get into "recommend it" vs "require it" discussion right nwo

  98. flow

    or even require that. but I don't want to get into "recommend it" vs "require it" discussion right now

  99. pep.

    I'd also require it, but I think I'm not gonna like what that implies to change in the document

  100. flow

    point is, recommending something that is sensible makes always sense, no matter the XEP state

  101. flow

    what does it imply? and what do you not like?

  102. pep.

    Well, recommending it would be better than rn, but still quite useless because it's not something I can use to detect 277 nodes in bare PubSub

  103. pep.

    So I'd rather require it.

  104. pep.

    Now..

  105. pep.

    Requiring it certainly means breaking change

  106. flow

    make a PR and put it on council's plate :)

  107. pep.

    heh

  108. pep.

    Do we have stats of what client is implementing microblog? Movim, Sàt, anybody else? What about libraries?

  109. pep.

    Link Mauve, any idea? (with the DOAP stuff you've been playing)

  110. pep.

    I'd also recommend using atom:generator in 277 maybe

  111. pep.

    Even though just using pubsub#type is already a step up

  112. pep.

    https://github.com/xsf/xeps/pull/986 !

  113. Zash

    Hm, searching for `pubsub#type` yields mostly examples and a changelog entry about clarifying the field.

  114. pep.

    Yeah I've seen that :/

  115. pep.

    "clarifying", always the same word. Though tbh it's still as loose

  116. pep.

    "payload type" doesn't mean much

  117. Zash

    https://xmpp.org/extensions/xep-0060.html#registrar-formtypes-config has the longest description I could find

  118. Zash

    > The type of node data, usually specified by the namespace of the payload (if any)

  119. pep.

    Ok, so that looks appropriate

  120. pep.

    There are two example that have "http://www.w3.org/2005/Atom" as value, but that's quite handwavy to me tbh

  121. pep.

    "Yeah it's atom, do whatever you want however you want"

  122. pep.

    Which is not exactly what 0277 is

  123. pep.

    And even if it were "just atom on pubsub", there's no possibility of evolution when done this way (when pubsub#type is set to Atom)

  124. pep.

    Where would it make sense to add an "Integration with Data Forms" section in 0060? Similar to https://xmpp.org/extensions/xep-0137.html#usecase.xdata

  125. pep.

    I'm thinking somewhere in §12 or before or after

  126. pep.

    https://xmpp.org/extensions/xep-0060.html#impl

  127. pep.

    How do I link from a XEP to an anchor in another XEP?

  128. pep.

    Just a normal html link?

  129. pep.

    I mean, <link/> with a normal html url?

  130. pep.

    xep-1234.html#foo ?

  131. emus

    On todays meeting: ralphm pep. guus.der.kinderen Seve Even you don't see your responsibility there (but I support Seve asking for opinion, because I also couldn't evaluate myself): You all would likely put txxmpp into the library list than client? Or do you not care to put it into the client list and one need to find out himself? In general I think not haveing a client list is very bad + I think XSF should try to maintain a list. Because here is the most knowledge on XMPP aggregated and that's where one would go first in my understanding and view.

  132. pep.

    For txxmpp specifically, I don't see why it wouldn't go into the client list. There is no requirement that this list must only include clients that provide a binary, or have documentation, or even match any definition of a "chat" application. There is simply no requirements.

  133. pep.

    I'd rather discuss a proposal that adds such a definition. In the meantime I think it perfectly fits the no-criteria requirements for the client list

  134. pep.

    As for the second point about having a client list, I disagree. I think xmpp.org is not the right place for this. Having a generic list of things is at best misleading

  135. emus

    > I'd rather discuss a proposal that adds such a definition. In the meantime I think it perfectly fits the no-criteria requirements for the client list Yes okay, I would keep it just simple as it is, as long as its not completely out of space.

  136. pep.

    What do you mean "simple"?

  137. pep.

    We can't do that, we have to accept all clients that ask to be in the list otherwise we wouldn't be "neutral" :/

  138. emus

    > As for the second point about having a client list, I disagree. I think xmpp.org is not the right place for this. Having a generic list of things is at best misleading I dont see where it is generic? I think not even haveing a list of this anymore is another step disconnecting from the actual applied community. And I think that is not good.

  139. Holger

    xmpp.org is about presenting the protocol. I would've thought it makes some sense to showcase a few implementations as examples. But it's probably mostly irrelevant to end users. They'll search app stores. (In theory. In practice their geek friends tell them to install client $x.)

  140. emus

    > What do you mean "simple"? Just not put any further hurdles in making PRs to a list

  141. pep.

    Holger, not everybody does that unfortunately :/

  142. Holger

    pep.: Does what?

  143. pep.

    For a really long time people would create accounts on xmpp.jp because it's what xmpp.org presented first

  144. emus

    > We can't do that, we have to accept all clients that ask to be in the list otherwise we wouldn't be "neutral" :/ Yes

  145. Holger

    Ah.

  146. emus

    > We can't do that, we have to accept all clients that ask to be in the list otherwise we wouldn't be "neutral" :/ Yes, I mean all XMPP

  147. emus

    I dont see why XSF cannot present what people are actually doing the the protocol

  148. pep.

    emus, the issue with having is list is multiple. First we can't even get people to agree on criteria, so we have as little as possible. Some are tring to fight against having unmaintained implementations while some other want their implementation to be able to make it up the list so they're rather not be too specific.

  149. pep.

    That's a first point

  150. pep.

    Then presenting a list not having any idea who the target is and what they're looking for, how they ended up here. No description of clients, what kind of clients they are (I discovered today some think there should only be "chat" clients in there?), etc.

  151. pep.

    As a user I'd just be lost with so much choice and no directions

  152. pep.

    I don't want to end up here

  153. emus

    Hmm, yes, but if there is no agreement so far its better having one list, all or no one, than have no list in my view. However, I know now what you mean my criteria, I thought about something different.

  154. emus

    Hmm, yes, but if there is no agreement so far its better having one list, all or no one, than have no list in my view. However, I know now what you mean with criteria, I thought about something different.

  155. lovetox

    maybe an idea would be to create another organization which defines own goals for the xmpp world

  156. lovetox

    instead of trying to make the xsf into it

  157. Zash

    A lobby organization, charged with promotion of XMPP? Mmmmmm

  158. emus

    Yeah, there you see ... I only see fights between those possible two organisations from the current situation

  159. pep.

    Holger, I'd rather showcase in XEPs with things like DOAP for example

  160. emus

    pep.: Sorry whats DOAP again?

  161. Zash

    Of course you don't need a separate organization to make a web page with a list of clients presented in a nice format

  162. pep.

    A document in which implementations say what XEP they support etc.

  163. lovetox

    Zash, yeah completly separate

  164. lovetox

    own homepage, own logo

  165. Zash

    cf modernxmpp

  166. lovetox

    for example

  167. emus

    yes, thats also fine

  168. lovetox

    and if enough client/server developers back this organisation

  169. pep.

    emus, https://lab.louiz.org/poezio/poezio/-/blob/main/data/doap.xml

  170. lovetox

    then it basically becomes the law

  171. lovetox

    we dont need the xsf to make a compliance xep

  172. Zash

    Huh?

  173. pep.

    lovetox, "the law"?

  174. pep.

    you mean "a possibility" rather? :)

  175. lovetox

    bad term, but i mean the standard

  176. lovetox

    the org that says whats good and bad

  177. Zash

    AIUI modernxmpp is meant to be more like extended implementation notes and guidelines for stuff other than the protocol itself

  178. lovetox

    i mean thats what most of you want anyway or not :)?

  179. pep.

    lovetox, sure. With implementations still having a choice to follow it or not

  180. Zash

    point being you can have a group/org/webpage with a different focus, and let the XSF focus on herding XEs

  181. emus

    --> I did not intend to say to loose the focus on on the protocol ☝🙂

  182. emus

    --> I did not intend to say to loose the focus on the protocol ☝🙂

  183. lovetox

    Zash, exactly, and i guess people want to make the XSF into it, because the XSF has some kind of official authority

  184. lovetox

    my point was, if enough client/server devs back this new org, it has enough authority

  185. Zash

    ... for what?

  186. lovetox

    for example to push compliance regulations

  187. Zash

    you don't even need that, you just need a popular client

  188. lovetox

    or to decide to *not* showcase unmaintained clients

  189. lovetox

    just an idea, i dont care too much, just see people fighting over this XSF direction stuff for a long time

  190. lovetox

    and i think people need to ask themself why they so badly need the XSF to do it instead of some other org

  191. lovetox

    i would understand if the XSF had an agenda or direction it was pushing, but it seems XSF says we dont care, we are neutral we just publish XEPs

  192. pep.

    https://github.com/Ppjet6/xeps/commit/dcbb771034bff841fd4f8ad4fb68173f93e9c414 gonna submit something like this. Thoughts? before I push the PR button

  193. lovetox

    then this is perfect, for another org to not jujst push XEPs :)

  194. lovetox

    its not a competition then

  195. pep.

    The XSF does care. People really don't get that. #NeutralityIsALie.

  196. pep.

    Wanting to showcase all clients is as much a direction than wanting to showcase a few clients

  197. jonas’

    lovetox, people want to have a separate org, but lack the resources to do it

  198. pep.

    MattJ maybe? ^ the commit above

  199. Zash

    pep.: Based on https://xmpp.org/extensions/xep-0122.html#usecases-datatypes I'm thinking it should have a prefix of some sort

  200. pep.

    pubsub:?

  201. pep.

    Or.. do we want this more generic?

  202. Zash

    > Start with a prefix registered with the XMPP Registrar [3] Is there any such thing?

  203. Zash

    Ah, https://xmpp.org/registrar/xdv-prefixes.html

  204. pep.

    0137 has a sipub: prefix

  205. pep.

    right

  206. Zash

    So I guess pubsub: would follow this

  207. pep.

    Where do I add that

  208. pep.

    I'm always confused with registrar things

  209. pep.

    "2005-08-26 Added sipub: prefix specified in XEP-0137. (psa)", in the XEP directly? Same as the datatype?

  210. pep.

    https://xmpp.org/extensions/xep-0137.html#registrar.xdata-validate "The XMPP Registrar includes 'sipub:' in its registry of Data Forms Validation Datatype Prefixes."

  211. pep.

    Does that mean this XEP defines it. I'm failing at english, or is it spec-language I don't understand

  212. Zash

    That's Editor procedural stuff that's done on Draft or so?

  213. pep.

    https://github.com/xsf/xeps/pull/988 here. With a few questions

  214. emus

    > lovetox, people want to have a separate org, but lack the resources to do it

  215. emus

    I don't see that there are actually resources in XSF either

  216. emus

    Still, I personally would see fewer disconnect and not have another additional organisation

  217. emus

    Still, I personally would like to see fewer disconnect and not have another additional organisation

  218. MattJ

    The problem is that the XSF consistently fails at anything much further beyond protocol development

  219. pep.

    I think it's rather that there's no will to do more by a majority. Be it for the (expected?) lack of resources or anything else, so the few attempts often fail indeed

  220. emus

    I think the ressources problem could be dealt with

  221. Zash

    emus, recruiting? 🙂

  222. emus

    Thats not what I thought of. but is sonething we should think about in general, too. I thought more into the direction, as pep. mentioned that there is a majority not wanting this, so who wants to work against that. If there is general support and motivation that will bring volunteers, I personally think.

  223. emus

    Thats not what I thought of. but is something we should think about in general, too. I thought more into the direction, as pep. mentioned that there is a majority not wanting this, so who wants to work against that. If there is general support and motivation that will bring volunteers, I personally think.

  224. emus

    Btw, Seve lets merge the txxmpp PR then?

  225. emus

    oh wait, I think there is one build error