XSF Discussion - 2021-01-19


  1. MattJ

    larma: nice to know!

  2. flow

    larma, that's some nice history, sounds very plausible and is somehow what I also assumed (that they used gtalk for push at some point). But may I ask if you have a reference to this information or was it passed to you privately?

  3. larma

    flow: Reverse engineering and reimplementing google android software for years gives you a lot of insights ;)

  4. wurstsalat

    nice to see some familiar constants :) https://source.chromium.org/chromium/chromium/src/+/master:third_party/libjingle_xmpp/xmpp/constants.h

  5. flow

    larma, ahh microg, of course

  6. flow

    larma, you don't happen to have written down version of the insights you gained somehwere?

  7. larma

    Uhm I still have all the source code

  8. larma

    My first implementation of the push service was using smack ;)

  9. larma

    I also remember experimenting with using a non-google XMPP server as a push backend at some point accidentally had me Google Talk use my non-google XMPP account and it actually worked :D

  10. dwd

    larma, I think it'd be more accurate to say that WebRTC started as Jingle without the XMPP. WebRTC hasn't ever used XMPP, but the signalling owes more to Jingle/GoogleTalk than it does to SDP. But fippo would probably know more than me (and correct me where I'm wrong).

  11. Kev

    I believe that to be correct.

  12. Kev

    They started off with releasing libjingle, which was the xmppish bits in gtalk needed for (the Google ancestor of) Jingle. Then the media bits of that got used in chrome, morphing into libwebrtc in the process.

  13. Kev

    (Source: early access to libjingle and integrating it into Psi as a PoC back in the day, and occassionally looking at it since, so not not exactly Fippo levels of intimacy)

  14. larma

    Kev: but wasn't one of the first usages of WebRTC at Google to replace the Google Talk browser extension so GMail could do video calls via Google Talk without that extension?

  15. larma

    Certainly WebRTC does not include the XMPP parts, as in the GMail video call case, that was already done on the server, but signalling still happened using XMPP and probably Jingle.

  16. Kev

    > but wasn't one of the first usages of WebRTC at Google to replace the Google Talk browser extension so GMail could do video calls via Google Talk without that extension I believe that to be correct.

  17. mathieui

    thanks mdosch, I indeed forgot about the abstract

  18. mdosch

    Yw. Just these days you talked about this thing and now it's already written down as protoxep. Pretty impressive. 😃

  19. mathieui

    content is still mostly up for discussion, but at least that provides a clear example of the features I would like to include

  20. Link Mauve

    Even back in 2015-2017 I had to use the Hangouts extension for Firefox in order to do video calls at work, and it was spouting JIDs and MUC and Jingle references.

  21. stp

    Hi, I just joined this chat to get to know if there ever was some sort of campaign or something like that to stop the use of "branding" XMPP as Jabber. IMO it's really detremental due to it confusing interested non-tech people.

  22. moparisthebest

    it's a touchy subject, many people in here think we should only use "Jabber" or use it more

  23. moparisthebest

    in theory, the XSF has the right to sublicense the trademark as it sees fit, however, all the public documents I've seen that say this expired almost 20 years ago, so I can't say for sure if we do or don't

  24. moparisthebest

    regardless, "jabber" is used by cisco to describe a super crappy product we probably don't want to be associated with anyway, I'd like to never use the word again ever, personally

  25. Zash

    In theory? All that should still be valid.

  26. Zash

    Info should be at https://xmpp.org/about/xsf/jabber-trademark/

  27. moparisthebest

    https://xmpp.org/about/xsf/jabber-trademark/background has some PDFs of agreements between JINC and JSF , neither of which exist anymore

  28. moparisthebest

    what do you think the chances are Cisco attorneys even know the XSF exists, why even risk it

  29. Zash

    Regardless, few modern things use the Jabber trademark these days.

  30. moparisthebest

    even if it is all valid, Cisco can terminate it "in the event of a material breach that remains uncured for 30 days" , the XSF can't do anything in 30 days, especially anything requiring lawyers in who-knows-where jurisdiction

  31. stp

    moparisthebest, Though I'd be fine with Jabber as an overlaying brand name just for the word itself, I would also much prefer XMPP due to those possible legal issues and the fact that one thing shouldn't have two names. Overall a firm decision everybody adheres to is long overdue and very important. XMPP.org at least doesn't mention Jabber on it's front page though for example Conversations still does multiple times which I don't get. It even calls the ID "Jabber ID". XMPP is difficult enough to figure out for new comers especially non-IT people on it's own, so it having two names is pretty catastrophic.

  32. Zash

    "Jabber ID" is the technicall term for it, for historical reasons.

  33. moparisthebest

    I just tell people to use Snikket nowadays :) (or Quicksy for the super-nontechnical)

  34. stp

    Conversations.im website I was writing about, I'm currently not sure about in-app usage of the Jabber term.

  35. Zash

    Says "XMPP address" here, as per https://docs.modernxmpp.org/ recommendations

  36. stp

    moparisthebest, Quicksy, blabber.im: helpful for non-IT people though what I read about Snikket I think it adds to the confusion.

  37. stp

    Zash, thanks, I'll quickly check that.

  38. stp

    Zash, you meant the Conversations app says "XMPP address"?

  39. Zash

    "XMPP-adress" actually, but locale 🤷

  40. stp

    Zash, ok thanks. So it's just the website that's still behind the times.

  41. stp

    Thanks guys! I just wanted to make sure that I don't give false recommendations to XMPP operators when I tell them to better use the XMPP term.

  42. wurstsalat

    stp, https://docs.modernxmpp.org/terminology/

  43. stp

    wurstsalat, yes thanks. I'll bookmark it now. Before joining this MUC is wasn't sure if that website had any authority in regards to the XSF. Btw looking at that site now the only remaining reminiscence of Jabber is the J in JID :-)

  44. Zash

    The XSF is primarily concerned with shepherding the protocol specifications. The Modern XMPP project was meant to target developers in order to foster coherent UI/UX.

  45. moparisthebest

    stp, to be fair the XSF doesn't have any authority over developers anyway

  46. SamWhited

    They don't have authority, but they have influence.

  47. stp

    moparisthebest, true.

  48. stp

    SamWhited, yes that's the word I should've used.

  49. jonas’

    stp, one reason for pushing towards the term Jabber is that it is an established brand (be it good or bad). XMPP is a technical term, and you wouldn’t advertise for Email as "IMAP" or "SMTP".

  50. jonas’

    (which is essentially what "XMPP" is, in terms of terms)

  51. jonas’

    but as long as we don’t have clients which could be marketed as "Jabber" (or whatever else term) on all platforms, I think the entire discussion is moot-ish anyway

  52. moparisthebest

    Which is why you market products to end users, not protocols

  53. moparisthebest

    No one says, hey contact me with XMPP over protobuf with fairy dust, they say, contact me with signal

  54. jonas’

    which is why using the term XMPP outside of developer and maybe *some* operator (if it’s e.g. for applications where XMPP specifically is a requirement, think IoT) documentation is not really a good idea

  55. moparisthebest

    If end users want a signal replacement I recommend Quicksy, if more technical people want to run something so they can chat with their family/friends, I recommend Snikket

  56. stp

    jonas’, It's not an established brand in the general public and it's not the case anymore in IT-environments. In that regard it's in the same state where brand names like ICQ are. The comparison to E-Mail is lacking since that has multiple underlying protocols as you rightly said.

  57. Zash

    "contact me on Conversations"

  58. stp

    moparisthebest, Comparison to Signal also can't be made since since that system has only one server and one client.

  59. moparisthebest

    "but you can run your own" (so they say, not really though)

  60. stp

    moparisthebest, yes, but they don't market it like that (quite the contrary in fact) and anybody who would run his own Signal service wouldn't call it Signal then.

  61. Zash

    "You can run your own" isn't a talking point (?) on any of the other silo messengers

  62. Zash

    Not like E2EE

  63. stp

    Zash, Conversations isn't the best of names judging by me experience with non-IT people.

  64. stp

    Zash, Conversations isn't the best of names judging by my experience with non-IT people.

  65. Zash

    🤷️

  66. Zash

    Haven't heard any issues with the name.

  67. SamWhited

    stp: Oh really? I'd be curious to hear what happened, I always thought that was a rather good name (but also I am not the best at marketing)

  68. stp

    SamWhited, Maybe it's easier to speak and remember for the English speaking audience, but for Germans not so much. The directly corresponding German word Konversationen is also not a word you'd often use in speech.

  69. SamWhited

    Interesting; I wouldn't have thought that since Daniel is german

  70. stp

    What I personally like about it though is that it's a unique name and very well describes what the app is all about.

  71. stp

    SamWhited, Yes, I think he was going for what I wrote in my second comment I just posted.

  72. mdosch

    Also devs use a lot more english than average Joe and therefore might be more used to the term.

  73. Ge0rG

    https://twitter.com/fedora/status/1351154894363877378 why can't we have this sort of advertising

  74. mathieui

    Ge0rG, but XMPP is boring, everyone knows how to setup a server!

  75. stp

    Hello again! Does XEP-0392 cover the use case in which a user sets a certain color for his own JID and the XMPP client then colors the contact icons on that JID's roster in the same color? And also is there a possibility to sync the color set for a JID across clients?

  76. mathieui

    stp, I am pretty sure it does not

  77. MattJ

    No, the colour is predetermined, the point is that there is nothing to sync

  78. mathieui

    (but I am probably not the best person to answer that)

  79. stp

    MattJ, mathieui ok, thanks for you replies. In all usecases I can imagine (don't know if that is representative) it would make more sense to predetermine colors based on the local user's JID so that a user could identifiy by colors on his roster/open chats list to which of his own JIDs/rosters every contact belongs. The way it's currently done, giving contacts the same color across clients is pretty useless especially since the resulting colors often look very similar even on color calibrated monitors. Your thoughts?

  80. mathieui

    not sure what you mean, colors are generally distinct, even on poezio which has a much smaller range of them available

  81. Ge0rG

    stp: as somebody who's caused this to happen, I think there is merit to see the nicknames of your contacts in the same color on all of your clients

  82. Ge0rG

    stp: when reading a message, is it more important to know on which account it arrives, or from which contact?

  83. moparisthebest

    I'm partially colorblind and most of the xep-0392 generated colors are almost the same in my IRC channels (biboumi), but then again, when each client did it's own thing it wasn't better so :)

  84. Ge0rG

    That said, a multi account client surely can use a different colored element to identify your account

  85. stp

    Ge0rG, Which contact takes priority of course, though that is pretty well indicated by the big letter and the nickname set for the contact. Even a color blind person can use those to indentify the contact.

  86. Ge0rG

    The challenge with the color palette is that you really can only use different colors on the color wheel, keeping hue and brightness constant

  87. stp

    Ge0rG, A client could do that, but if there's no sync happening as mentioned before by mathieui than that's not good, too.

  88. Ge0rG

    Google has eight mostly distinct colors for different contacts, so they repeat often

  89. mathieui

    stp, afaik 392 does not specify any communication

  90. stp

    moparisthebest, thanks for that input. I'm on a color calibrated display on PC and half the color of my contacts look nearly the same.

  91. Ge0rG

    stp: there is no need to sync, all 0392 clients arrive at the same color tone

  92. Ge0rG

    https://upload.yax.im/upload/JwK5FiUz-bpJdaWu/Screenshot_20210119-220031_yaxim.png

  93. Zash

    stp, half the colors are the same for everyone, for some reason.

  94. stp

    Ge0rG, yes, but sync would be needed if quote of yours "multi account client surely can use a different colored element to identify your account". At least that's how I understand it.

  95. mathieui

    (I have nearly the same colors as Ge0rG on my poezio, FYI, so it works quite well)

  96. mdosch

    > stp: there is no need to sync, all 0392 clients arrive at the same color tone Except profanity. 😂

  97. stp

    Zash, I'm not sure if I get what you mean, but currently on my open chat list four accounts are slightly different greens (wouldn't be visible I guess on low-end, uncalibrated displays) and three are purple/pink-ish.

  98. Zash

    You should have seen early versions when everyone was pink.

  99. stp

    Ge0rG, on that screenshot you posted nearly every contact has green applied. That's useless, don't you agree?

  100. mathieui

    Zash, still no sync required!

  101. Zash

    Huh? 2/11 is nearly everyone?

  102. stp

    Zash, XMPP barbie edition.

  103. Zash

    Birthday attacks on your color perception! 🙂

  104. Ge0rG

    stp: removing much green will reduce contrast

  105. mathieui

    stp, https://pix.mathieui.net/o/kBuvK.png "everyone is green" is a bit of an overstatement

  106. Ge0rG

    Complain to your ancestors for living in the prairie

  107. stp

    I have me best friend on the roster of two of my several JIDs. So far I can never make out on which account "I click him" (difficult to explain that for me in English). So my idea was my own JIDs would get certain colors, decided upon by myself and synched, or rather after this discussion predetermined by XEP-0392 and all contacts belonging to a certain of my own JIDs would get colored the same.

  108. stp

    mathieui, I was purely refering to Ge0rG's screenshot. Your screenshot though demonstrates that there in no obvious concept by which the colors are given. Looking at it my guess would be the algorithm is just trying to not have the same color for contacts showing up next to each other though looking at the contact names "matkor" and "Matthew" on the bottom of the list it even can't be that or fails at it.

  109. Ge0rG

    stp: some email clients allow color coding different accounts. I think what you ask for is sufficient to implement as a local client feature. Adding an account is much harder than assigning a color code to it, and manual assignment makes sense there

  110. moparisthebest

    stp, how sure are you that you aren't a bit colorblind too? :P I didn't know until a casual conversation when I was 21 when I was talking about my favorite green shirt and the other person said "you mean the brown one???" lol

  111. stp

    Ge0rG, Yes, I use that via an addon in Thunderbird. To understand your opinion please tell me how the current behavior of giving colors to contacts helps you in instantly recognizing that contact even though the possiblity is very high to have several contacts showing up with the (very nearly) exact same color. I don't get that.

  112. Ge0rG

    stp: I'm using it for nicknames in rooms

  113. SamWhited

    Because there's also text (whether it's the first letter on the avatar or the full name like the screenshots from earlier). If you're scanning the list for your friend your eye will get used to jumping between just the green ones instead of having to look through the whole list, which shortens the search time a lot for a long list.

  114. stp

    moparisthebest, I hope you're joking.

  115. SamWhited

    (or whatever color the person you're looking for gets set to)

  116. stp

    Ge0rG, Aaah, that's a usecase I didn't think about. So even if my proposal would bet applied there would need to be a different (the current) behavior for the room members lists.

  117. moparisthebest

    stp, I'm absolutely not joking :)

  118. stp

    SamWhited, for that to work you'd need to memorize each contact's color though which is a flawd concept.

  119. moparisthebest

    colors from a 36 person IRC channel from Dino: https://burtrum.org/up/ae72c078-b1ae-4289-97a4-b482d52c63ca/open-screeny-181319.png + https://burtrum.org/up/c9c66a61-7f20-4de4-b3fb-45fca61f9c6e/open-screeny-181444.png + https://burtrum.org/up/dc882354-d570-482d-9288-bf590bfd0e38/open-screeny-181575.png + https://burtrum.org/up/26df34d2-5439-4889-8913-bb54401ebe74/open-screeny-181738.png

  120. stp

    moparisthebest, Do you think I'd color calibrate displays if I couldn't differentiate between colors?

  121. moparisthebest

    there are 2 R's that look to be the same green

  122. moparisthebest

    stp, yes, I think I can see all colors but some brown things appear army green

  123. SamWhited

    stp: no you don't, your brain does it automatically for you after a while.

  124. moparisthebest

    like I said, I went 21 years without realizing I was a bit colorblind, since then I've learned to ask questions like "what color does this look like to you" and noticed it more :)

  125. stp

    moparisthebest, I think it would already be better if all R's for example would receive the same color.

  126. moparisthebest

    392 isn't perfect, but so far, I've yet to see perfect, it's not worse than anything else I've seen, and consistency is nice

  127. moparisthebest

    any different ideas though I'd be interested in trying

  128. stp

    SamWhited, My brain didn't and I don't think I'm the only case. When there is zero natural connection between a color and a contact name I think it's a bad concept. If it was a good one it would work immediately instead of only after using it for some days.

  129. stp

    moparisthebest, I'm not against XEP-392, but used the way it gets used.

  130. Zash

    What's a "good connection between contact and color"?

  131. stp

    *just the way it gets used.

  132. moparisthebest

    good connection is what avatars are for imho

  133. SamWhited

    Sure, it's not perfect for 100% of people, but it fills in the gap if you can't have avatars like moparisthebest just said.

  134. moparisthebest

    have you seen, for example, how gitea generates avatars for people?

  135. moparisthebest

    I don't know the word, but they look like a kaleidoscope ?

  136. Zash

    It's far above what you get if everyone gets exactly the same avatar.

  137. SamWhited

    It's like syntax highlighting in source code. You can't disable the words and just have colored blocks and still read the code, it just provides more info. You *can* probably spot a bug when a string is green instead of the red it normally is (or whatever) though even though if asked you probably wouldn't be able to tell people "strings are normally red", something just feels off because you've picked up visual cues after using it for a while.

  138. moparisthebest

    like this: https://burtrum.org/up/42c14e99-8f25-4f1e-9665-f47f7e600a72/open-screeny-183339.png

  139. SamWhited

    moparisthebest: identicons are what gravatar calls them IIRC

  140. SamWhited

    I do think that's a better visual hash than just colors personally.

  141. moparisthebest

    maybe it's better for other people, but I find this not better at all compared to color+initial

  142. Zash

    moparisthebest, http://sv.gravatar.com/site/implement/images/#default-image

  143. moparisthebest

    like I can't differentiate them very well, I don't really learn to recognize them the way I can an avatar

  144. moparisthebest

    if they had a letter inside them too, maybe that'd be ok?

  145. stp

    Zash, in human nature it would be a skin or hair color of your contact (provided you know how a contact looks). Disconnecting it from the personal connotations we could for example make it strict that A is always blue.

  146. Zash

    You're proposing always using the first character? Or what?

  147. moparisthebest

    but that's super bad when you have 20% of your channel starting with an M

  148. stp

    SamWhited, Avatars don't tell the user on which of your JIDs they are. I currently see my best friends avatar twice in my Gajim client.

  149. SamWhited

    stp: that's not the point of this XEP, and that's something clients can do themselves.

  150. SamWhited

    Add a tag like conversations does (I think) or set a little color icon somewhere else on the line. That's a visual cue that doesn't need to be consistent between clients. Scanning a big list and having hints about where in the list to look though is nice and works better if it's the same between clients.

  151. SamWhited

    Or don't merge the lists, put them in a tree underneath a thing representing the account or something. There are lots of ways to do what you want.

  152. stp

    SamWhited, clients could do, but if there's no sync between them it would lead the concept to absurdity.

  153. SamWhited

    stp: why is that? I don't understand why I would want that to be the same on all clients.

  154. moparisthebest

    stp, propose a XEP to color as you like and another XEP to sync avatar generation method across clients?

  155. SamWhited

    Dino doesn't use the same visual language as Conversations or Gajim, for example, so why would they indicate which account something belongs to in the same way?

  156. SamWhited

    I guess if you wanted to do something with colors there it would make sense to have the colors be the same, but I suspect different clients will or won't use colors and the actual mechanism for conveying that information would be different

  157. moparisthebest

    enough clients are similar-ish enough UI-wise where it'd make sense to have them share that

  158. Zash

    Publish yourself a 1x1px avatar? Invent yourself a way to publish personal avatars for your contacts? (There's an unused (I think) thing for publishing contact details.)

  159. stp

    SamWhited, let's say I I would set purple for my jabber.at JID and green for my blabber.im JID. The whole thing would fail if I set different color for those JIDs in another client by accident or in case a client limits the selection to certain colors.

  160. SamWhited

    stp: I agree with that, if for example Conversations used colored icons to indicate which account something was on and Yaxim did too, but Gajim used a tree instead it would still make sense to have Conversations and Yaxim use the same colors. I just don't know if it makes sense to say "all clients have to use colors for this".

  161. SamWhited

    But yah, write an XEP for it, this XEP does something totally different and isn't for that, but that doesn't mean there can't be one for personal avatars or something. It could even use consistent color hashing to get the account colors that it uses. These two things aren't competitors.

  162. stp

    SamWhited, yes, no need to force clients to use colors, but at least if colors are used make sure the concept behind it is a good one and consistent. XEP-0392 tries the latter successfully judging just by Conversations and Gajim on my devices but fails at the former in *my* opinion.

  163. moparisthebest

    yea "good" will always be an opinion I guess

  164. moparisthebest

    it's good enough that many client devs decided it was better than what they hand-rolled

  165. moparisthebest

    but please, define a different way to do it, if it's better, people will switch

  166. SamWhited

    I tend to think it's a good concept, but we can use colors for multiple things; at the risk of possibly making a false equivalence (though I don't think so, the only similarity is that both concepts use color) this is like you're saying "XEP-0045 defines multi-user chat but I want it to define colors for my account JID". That's just not ht epurpose of XEP-0045, and if you don't want to support MUCs but you do want to support custom avatars you can create a new thing to define how custom avatars work and some clients can implement one and some the other, and some can even implement both.

  167. stp

    SamWhited, XEP-0392 is very difficult to read for me but on top of the page it says "The string can be a nickname, a JID or any other piece of information." Reading that again I'm even more confused and I now think it's too losely worded but I may understand the whole thing wrong.

  168. stp

    moparisthebest, it's better than before that's 100% certain.

  169. SamWhited

    stp: I'm not sure if this helps or hurts your understanding, but all 0392 does is define a hash function where the output is a color. How you use it is up to you.

  170. SamWhited

    jonas’: maybe some of this confusion would be cleared up if the use cases section was filled out in a future version? Right now it sort of outlines the algorithm, which isn't really a use case

  171. SamWhited

    stp's idea could even just be one of the use cases actually, maybe it doesn't even need a separate XEP (although me personally I think the way everybody uses it now helps me a lot more, but I only have one account and may be in the minority).

  172. stp

    SamWhited, yes that helped. I know think the Gajim devs claimed XEP-0392 would do more than it actually does or I didn't understand what they meant.

  173. SamWhited

    This further makes me think we need an organization that publises a series of documents that are UI related instead of technical and define how consistency works across clients in the public network. Maybe Modern XMPP is that place, I'm not sure.

  174. stp

    SamWhited, YES, that's what I wanted to express before - that XEP-0392 could be used for what I propose.

  175. Zash

    PR it to Modern XMPP? 🤷

  176. stp

    SamWhited, also you're right that with just one personal account my proposal wouldn't be needed.

  177. SamWhited

    That's why I think setting the color of every single person in the list to an account color is a bit overkill, maybe just have a colored square next to them or something and use the avatar or contact name color in the way clients mostly use it now.

  178. SamWhited

    Or do an expandable tree so they show up under the account itself

  179. Zash

    Colored border? Or a corner? Based on the account JID?

  180. Zash

    (I said something like that in gajim@)

  181. SamWhited

    That sounds nice looking

  182. stp

    SamWhited, Gajim has the expendable tree but also an "open chats"-tablist and that's were the issue came up.

  183. stp

    Zash, I had proposed colored border around avatar to Gajim earlier today when I discussed this with them.

  184. stp

    Zash, ah you wrote @gajim too :-)

  185. SamWhited

    ah yes, I see. I +1 Zash's idea.

  186. Zash

    Dunno if I was following.

  187. Zash

    Dunno if I saw you write that. Maybe I did.

  188. Zash

    I've discovered that ideas don't matter all that much unless they're implemented 😉

  189. SamWhited

    Or do like Slack et al do where you have to switch between accounts and you don't see them all merged. I kind of like that personally. Helps me keep work separate from personal stuff and keeps the open chats list shorter since it's not combined between all accounts)

  190. SamWhited

    Although I hated it at first; can't remember why now. I guess it grew on me.

  191. emus

    Hi guys, here we go: I wrote this comment on the current situation in the media. As you can read, my focus is the decision making for people. I wrote a general article, but added also direct example or "statements" within the article. In the end I also provided some links for further reading. Please review, but please consider that I invested several hours in this incl. a Pull Request for improving the XSF Client page. Please provide further links if you think it suits! Tweet etc. will follow.

  192. Zash

    Sounds like that makes a whole lot of multi-account problems go away.

  193. emus

    https://github.com/xsf/xmpp.org/pull/871

  194. stp

    SamWhited, A long time ago I proposed left/right swipe to switch accounts in Conversations once when my Conversations chat list got longer and it's a mix between 1:1s and MUCs.

  195. SamWhited

    Reading through modernxmpp's terminology now; mostly I love it, but seeing "Roster" go away makes me kind of sad :) that was the one XMPP term I thought was better than all the others, because it always made me think of baseball. But something simple and obvious like "contact list" is probably way better in general

  196. stp

    I liked roster to. I knew the word from the second NBA Live game. Ok I'm away now. Good discussion, I hope I encouraged some thinking about such UI/UX improvements for XMPP at least. :-)

  197. SamWhited

    Maybe it would be good to take multi-account discussion to the ModernXMPP chat room? (/cc zash who might know if this has been discussed already)

  198. MattJ

    SamWhited, it's not that "roster" is an unused term over here, but consider that we don't have the baseball connection and you'll find that most people find it a weird word to use in place of what everyone else calls it... a contact list

  199. Zash

    MattJ! Board member and native English speaker and even awake, could you take a look at emus post?

  200. Zash

    emus: Good work.

  201. emus

    > Zash escribió: > MattJ! Board member and native English speaker and even awake, could you take a look at emus post? Please 🙈

  202. emus

    Zash: thanks

  203. Zash

    I wonder if that worked as an anti-bystander effect thing.

  204. emus

    Zash: the blog post? what do you mean by anti-by.. ?

  205. Zash

    Hrm, trying to transfer the text to my e-reader thing, but suddenly so many yaks in the way and I'm so sleepy.

  206. Zash

    emus: https://en.wikipedia.org/wiki/Bystander_effect

  207. Zash

    When you think "someone else will do it" but everyone else thinks the same thing and nobody does anything.

  208. emus

    Ok, I understand the term now, but not the context, so what you want to say

  209. SamWhited

    MattJ: indeed, makes perfect sense

  210. Zash

    To quote http://www.hpmor.com/chapter/88 > The standard remedy for bystander apathy was to focus on a single individual. I.e. "You, there, help me!"

  211. Zash

    Might not work in a volonteer organization tho.

  212. moparisthebest

    I don't think I realized roster was baseball related until just now

  213. emus

    Zash: Now I understand. Its late, lets wait for tomorrow

  214. Zash

    Indeed.

  215. emus

    at least in EU

  216. moparisthebest

    emus: First off, great work! I think it should be published ASAP, wording can be tweaked later, but waiting until perfection is not required

  217. moparisthebest

    Second, a few nitpicks:

  218. moparisthebest

    > It mostly does not matter how your website look like, but everyone can interact with it.

  219. moparisthebest

    Maybe "It mostly does not matter what your website looks like, but that everyone can interact with it." ?

  220. moparisthebest

    Or "It mostly does not matter how your website looks, but that everyone can interact with it."

  221. moparisthebest

    And then, this is duplicated:

  222. moparisthebest

    > most people and organizations out there – and if not yet, it is open for your modification in the future - and if not yet, it is open for you to make modifications in the future to make it do what you need.

  223. emus

    moparisthebest: Thanks, yes the example quote were a quick shot because one recommended to give some simple or clear samples. Happy to hear better comparisons can you post the changes requests int the PR?

  224. emus

    Yeah, but lets not rush, we are late anyway. some review is fine

  225. emus

    Gonna sleep now. good night