XSF Discussion - 2018-01-25


  1. moparisthebest

    Whey is memberbot source code

  2. moparisthebest

    Where*

  3. moparisthebest

    I know I've seen it before but can't find it now

  4. moparisthebest

    Is it https://github.com/legastero/memberbot

  5. moparisthebest

    Or is there a newer version

  6. jonasw

    SamWhited, would you be more in favour of #559 if it were a separate XEP?

  7. SaltyBones

    ah, mostly the usual crowd ;)

  8. jonasw

    yupp

  9. jonasw

    consolidating protocol discussion here is good tho

  10. Ge0rG updated the already-complicated MUC self-message matching with another heuristic yesterday.

  11. Ge0rG

    If the sent message begins with the reflected message, followed by a newline, this must be a reflection from a "MUC" that splits multi-line messages. Yikes.

  12. edhelas

    damn, we reached XEP 400

  13. SouL

    :D

  14. edhelas

    https://xmpp.org/extensions/xep-0404.html

  15. Ge0rG

    edhelas: that in-joke won't work for long any more

  16. SouL

    Haha :)

  17. Ge0rG

    unless we can convince the editors to skip over 404

  18. Ge0rG

    Or, even better, somebody prepares a Humorous "XEP-404: XEP Not Found" that will be put into the queue at the right moment

  19. Dave Cridland

    Ge0rG, That's literally what I was in the process of typing.

  20. Ge0rG

    Dave Cridland: that proposal or that XEP?

  21. Dave Cridland

    The idea.

  22. Holger

    .

  23. jonasw

    :

  24. Zash

    ·:

  25. SouL

    :/

  26. vanitasvitae

    v

  27. jonasw

    .

  28. zinid

    ping

  29. jonasw

    pong!

  30. Guus

    🔫 (pang)

  31. SouL

    Holger, look what you have done >:[

  32. zinid

    vanitasvitae knows how to fight spam!

  33. zinid

    https://github.com/processone/ejabberd/issues/2246

  34. Guus

    there's a XEP that describes that...

  35. Ge0rG

    I've asked for that feature a year ago, or maybe two, too :)

  36. zinid

    what feature? rendering captchas?

  37. zinid

    and why a client cannot generate captchas on its own?

  38. Ge0rG

    zinid: why should the client do that? what do you do with multiple clients?

  39. Ge0rG

    zinid: do you have to solve a captcha for each of my clients? I have five or six

  40. Guus

    XEP-0159 I think

  41. zinid

    Ge0rG, ah, I see

  42. Guus

    allowing a server to respond with a challenge to potential spimmers (and keeping track of who is or is not a potential spimmer)

  43. zinid

    well, this is trivial to implement in ejabberd

  44. jonasw

    {xep 0159}

  45. Bunneh

    jonasw: Spim-Blocking Control (Standards Track, Deferred, 2006-07-11) See: https://xmpp.org/extensions/xep-0159.html

  46. zinid

    the question is what client will support this?

  47. Ge0rG

    none!

  48. Guus

    zinid: as soon as you build it, pretty much every client :)

  49. zinid

    Guus, I built it for IBR ages ago, still only minority of clients support it

  50. Guus

    I assume that's because IBR is either not used as much, or used in a more 'trusted' setting (of an internal network).

  51. Guus

    if it's easy for you to build, please consider adding it, at least as a configurable option?

  52. Guus

    it'll surely not hurt the spim battle.

  53. Ge0rG

    0159 depends on Privacy Lists, which are deprecated.

  54. Guus

    eternal fame will be your part.

  55. Ge0rG

    I have two dozens of pending subscription requests

  56. Ge0rG

    PARS should also be able to pass through that captcha

  57. vanitasvitae

    The way I imagine it doesnt require client support

  58. Guus

    vanitasvitae: surely, the initiator's client should support it?

  59. Ge0rG

    vanitasvitae: please add a remark regarding PARS

  60. vanitasvitae

    Ge0rG: what is pars?

  61. vanitasvitae

    Guus: nope, it could be implemented purely server side

  62. Guus

    vanitasvitae: at some point, a person needs to verify that there's a person, not a spambot. Where does that happen?

  63. SouL

    vanitasvitae, do you mean solving the captcha in the browser?

  64. vanitasvitae

    > https://github.com/processone/ejabberd/issues/2246

  65. SouL

    jabber.ru used to do that

  66. zinid

    SouL, for joining rooms, yes, that's standard ejabberd's feature

  67. zinid

    but you only need to solve it in browser if you have retarded client

  68. zinid

    normal clients render captcha just fine

  69. SouL

    Yeah, what I tried to say is that a link is provided too heh ^^

  70. zinid

    true

  71. Guus

    oh, right. Yeah, I'd also see the 'open link in browser' as a fallback for clients that are not compliant.

  72. Ge0rG

    vanitasvitae: PARS is https://xmpp.org/extensions/xep-0379.html

  73. zinid

    vanitasvitae, ok, I have assigned the issue to myself, will see what I can do

  74. vanitasvitae

    zinid: nice

  75. Ge0rG

    Dave Cridland: there is still the technical challenge of not hosting the new XEP under xep-0404.html (leading zero)

  76. jonasw

    Ge0rG, I’m sure we can put a redirect from xep-404.html to xep-0404.html :)

  77. Ge0rG

    jonasw: the other way 'round would make more sense.

  78. jonasw

    and also be much more tricky

  79. jonasw

    I don’t want to mess with editor tooling on that level.

  80. Ge0rG

    don't think so

  81. jonasw

    maybe if it’s done at the end of the build stage

  82. Ge0rG

    jonasw: we could make XEP-404 work as a generic error page and serve it as that, then just blacklist 404.

  83. Ge0rG

    jonasw: wouldn't it suffice to create a xep-404.xml and build that?

  84. jonasw

    yeah, and then all tooling breaks because it expects filenames to have xep-%04d.html format.

  85. Ge0rG

    how is the "next autoincrement value to be used" determined?

  86. jonasw

    yeah, and then all tooling breaks because it expects filenames to have xep-%04d.xml format.

  87. jonasw

    max(accepted_xeps.keys()), actually

  88. Ge0rG

    a rewrite rule on the server, that doesnt do 30x

  89. Ge0rG

    hm. oka

  90. Ge0rG

    +y

  91. jonasw

    yeah, that’d work of course

  92. jonasw

    that’s kinda what I meant by redirect :)

  93. jonasw

    Ge0rG, https://github.com/xsf/xeps/blob/master/tools/accept.py#L78 to be specific ;)

  94. jonasw

    but that can easily be overridden once to avoid 404 to be created when 403 is the newest one.

  95. Ge0rG

    jonasw: I'm sure max() will work out properly if there is a xep-404

  96. jonasw

    (FWIW, I would’ve preferred to avoid relying on %04d, but taking all XML or xep-*.xml files in the directory did wrong things because of templates and such)

  97. Ge0rG

    yeah

  98. jonasw

    I noticed that the copyright in the XEP page headers is still "© 1999 – 2017 XMPP Standards Foundation. SEE LEGAL NOTICES."

  99. jonasw

    do we want this to be "1999 – 2018" on all the xeps?

  100. Ge0rG

    jonasw: yes please

  101. Ge0rG

    But you might want to get Board approval for that.

  102. SamWhited

    nah, then it'll sort as 404 not 260 and be way higher than the other xeps (assuming python where leading 0 is octal)

  103. jonasw

    SamWhited, not in python3 anymore ;)

  104. jonasw

    (in python3, 0123 is simply syntax error)

  105. SamWhited

    oh, weird

  106. SamWhited

    nothing to do here then

  107. jonasw

    SamWhited, would you be more in favour of #559 if it were a separate XEP?

  108. SamWhited

    I'm not sure, I need to think over how the change will affect things (assuming that's the muc change)

  109. jonasw

    yah, it’s the MUC change

  110. Ge0rG

    This is not PHP, where "0280" parses as 2.

  111. MattJ

    and is == to TRUE, FALSE, "2", "0" and "sausages"

  112. nyco

    Test C

  113. jonasw

    ohai nyco

  114. nyco

    Test M

  115. intosi

    So it's not vegan, much like the new notes?

  116. nyco

    seems to work for now

  117. ralphm set the topic to

    XSF Board Meeting | Logs: http://logs.xmpp.org/xsf/ | Agenda https://trello.com/b/Dn6IQOu0/board-meetings

  118. ralphm bangs gavel

  119. MattJ

    intosi, I've not tasted either of them

  120. nyco

    these days, I lose messages, got delays

  121. ralphm

    0. Welcome & Agenda

  122. ralphm

    Who do we have today?

  123. nyco

    o/

  124. Guus

    me

  125. MattJ

    nyco, then I suggest you use a different server/client, otherwise it will be hard to participate...

  126. MattJ

    I'm here

  127. Martin

    I'm here

  128. nyco

    it would be nice if XMPP worked fine...

  129. MattJ

    Working fine for me, but anyway...

  130. nyco

    for me

  131. ralphm

    Full house, splendid

  132. nyco

    it seems you don't receive my messages in one-to-one chat

  133. Guus

    (that's a first, I think?)

  134. ralphm

    Any additional topics for the agenda?

  135. MattJ

    None here

  136. jonasw

    ralphm, yeah, whether I shall update the copyright notice on XEPs with this year.

  137. jonasw

    (it currently reads "© 1999 – 2017")

  138. Guus

    I've got nothing that's not already on Trello.

  139. nyco

    please add the card

  140. ralphm

    I don't think copyrights work like that

  141. jonasw

    I have no idea either

  142. jonasw

    but having not the current year there looks weird

  143. jonasw

    cf. https://xmpp.org/extensions/xep-0301.html

  144. ralphm

    ok, putting that in later

  145. Guus

    (not sure if that's up to board?)

  146. jonasw

    Guus, me neither, but it feels boardish

  147. jonasw

    (and Ge0rG from council said so :))

  148. ralphm

    1. Board priorities

  149. ralphm

    I believe we still need to schedule a video conference on this, no?

  150. nyco

    yep

  151. nyco

    that's on me

  152. nyco

    when's best? time day of week what week?

  153. Guus

    let's schedule that via mail?

  154. nyco

    maybe not this week, nor next

  155. nyco

    ok

  156. nyco

    I'll send a poll

  157. ralphm

    great

  158. MattJ

    +1

  159. nyco

    if I had your nogo, that could help

  160. nyco

    if any

  161. Guus

    perhaps combine with summit?

  162. nyco

    good question

  163. ralphm

    Guus: this was indeed suggested, but one of us isn't going to be there physically

  164. nyco

    I feel like those discussions would benefit a lot from members but I fear the crowd will be too large for everyone to be heard

  165. Guus

    ralphm: perhaps one joining in remotely?

  166. ralphm

    indeed

  167. nyco

    that said, this discussion would also benefit from non-members, that is the rest of the world

  168. Guus

    nyco: to clarify, I was not suggesting to discuss this during the summit itself, but in the evening or something.

  169. Guus

    wiht just board members.

  170. ralphm

    Let's continue this on list then

  171. nyco

    that's two more hours of work, I guess people wanna have beer

  172. Guus

    on list. :)

  173. ralphm

    2. Bus Factor Bank Account

  174. ralphm

    Did you discuss this last week?

  175. nyco

    yep

  176. Guus

    I still would like to know what the bank procedures are on this. Haven't talked to Peter about this in a couple of weeks though.

  177. ralphm

    So we need to track him down on that, ok.

  178. Guus

    if bank procedures are adequate, I'd prefer not to add individuals.

  179. ralphm nods

  180. ralphm

    3. Copyright statements on XEPs

  181. Dave Cridland

    My suggestion on copyrights would be to claim the present year always.

  182. Guus

    if all our current XEPs list an end date in 2017, I'd simply update that to 2018.

  183. ralphm

    agreed

  184. Guus

    that said: I'm not sure how these are supposed to work.

  185. ralphm

    I think that's usual practice. Whether it holds up in court is another mattre.

  186. ralphm

    matter

  187. Dave Cridland

    ralphm, It only matters after what, 75 years? I think we're safe.

  188. ralphm

    Dave Cridland: :-D

  189. Dave Cridland

    ralphm, Might want to make a note to revisit after we're all dead.

  190. Guus

    I'd say: make it 2018.

  191. ralphm

    noted

  192. ralphm

    4. AOB

  193. Dave Cridland

    Summit status?

  194. nyco

    Summit?

  195. Guus

    You're leaving out the items for discussion?

  196. nyco

    I need to book the dinner

  197. nyco

    same place as usual? auberge bretonne?

  198. nyco

    if no opposition, I'm on it

  199. Guus

    works for me.

  200. nyco

    consider it done

  201. nyco

    what else?

  202. Dave Cridland

    I suggested to Guus that we might want to specifically seek out people to invite. The Google open source office people might be around, for example.

  203. Guus

    financials-wise, I have questions

  204. Guus

    Ah, I thought you ment to invite our GSoC students.

  205. Guus

    but yeah, I'd be in favour of both.

  206. MattJ

    Likewise

  207. Dave Cridland

    There's probably other folks we might invite, if we think about it. Maybe even the Matrix folks. If only so we can stick them with the bill.

  208. ralphm

    hah

  209. nyco

    good idea, they may already have a plan, they got a company, a community

  210. nyco

    but yeah we can merge

  211. ralphm

    anything else here?

  212. nyco

    any RTC

  213. nyco

    nope

  214. Guus

    Shall I extend the invitation to Google?

  215. ralphm

    Guus: please do

  216. Guus

    + students?

  217. nyco

    +1

  218. nyco

    we gotta keep the number acceptable

  219. Guus

    students are two persons.

  220. nyco

    so that I book without a too large error margin

  221. nyco

    ok,Google? Matrix? How many?

  222. SaltyBones

    Oh, are people going to fosdem?

  223. Guus

    (well, three, but I'm guessing one is not coming)

  224. Guus

    I'll extend just to Google, and ask for attendance numbers.

  225. MattJ

    SaltyBones, yes, people are going to FOSDEM

  226. Dave Cridland

    SaltyBones, Lots of people, and for two days before as well.

  227. SaltyBones

    Can I join some of your meetings?

  228. ralphm

    https://wiki.xmpp.org/web/Summit_22

  229. ralphm

    SaltyBones: have a look there

  230. Dave Cridland

    SaltyBones, Please do.

  231. Guus

    I'd like to have us address the financials for a bit

  232. ralphm

    Guus: shoot

  233. Martin

    Agreed Guus

  234. Guus

    a) apart from a van, we have no sponsors

  235. Guus

    b) see Matt's mail from earlier today

  236. Guus

    XSF will cover the diner, I assume. Will it also cover lunch at Cisco?

  237. ralphm

    If there's no sponsors, I believe we should, yes.

  238. ralphm

    Of course I'd prefer companies stepping up for sponsorship.

  239. Guus

    We could have individuals pay for themselves, but I'd prefer not to.

  240. nyco

    I'll try mine

  241. Guus

    Does the XSF also cover transportation to/from the dinner?

  242. MattJ

    Generally, yes

  243. Dave Cridland

    Guus, Traditionally, yes.

  244. Guus

    (and, obviously, I'd prefer sponsorship, but if we cant find any...)

  245. Guus

    ok, thanks

  246. Guus

    I'll try to figure out how this works with Lunch at Cisco, offline.

  247. Guus

    nyco, you were going to arrange for dinner transportation, correct?

  248. ralphm

    It is easy really

  249. ralphm

    We pay at the register

  250. ralphm

    dinner transportation is generally a few cabs

  251. ralphm

    of course they'd have to be called in time

  252. Kev

    And generally what happens is that individuals (often Isode) end up paying for the cabs.

  253. Kev

    Which might not work so well this year.

  254. Guus

    we don't have an XSF credit/debit card I think, so someone will have to front?

  255. nyco

    Guus yes

  256. nyco

    Uber? taxi? coach?

  257. Guus

    I have no preference for a type of transportation.

  258. Dave Cridland

    We should arrange something ahead of time, after last year.

  259. nyco

    running will do, then

  260. nyco

    tell me your preferences

  261. ralphm

    Anyway, in terms of sponsorship for the Summit, we could write an e-mail asking for sponsorship, like we did last time

  262. nyco

    email written to ESL

  263. MattJ

    For what it's worth, I don't think it's ever been difficult for us to find sponsors - all we need to do is ask

  264. ralphm

    For the more general topic, maybe we should discuss this next week in person

  265. ralphm

    MattJ: right

  266. ralphm

    I'll write that e-mail

  267. ralphm

    anything else?

  268. Guus

    (not from me)

  269. nyco

    https://trello.com/c/vFYiyf9I/300-summit-sponsoring

  270. Guus

    please look at the commitment list? :)

  271. Guus

    there has been an Item For Discussion on Trello that has gone undiscussed: the surveying thing

  272. Guus

    can we either discuss or archive that?

  273. Guus

    https://trello.com/c/wIlPrAyC/258-consider-surveying-membership-to-work-out-what-the-issues-are-and-what-the-board-can-do-to-help

  274. ralphm

    Indeed, it's been there for a few weeks, and I thought it would be part of the general priorities discussion

  275. ralphm

    Does that make sense?

  276. nyco

    https://trello.com/c/8XwcLKzf/301-summit-dinner

  277. Guus

    might be a chicken/egg type of deal, but, sure.

  278. nyco

    a lot of effort, for what result?

  279. nyco

    if someone commits to this, thx, I will provide some support

  280. ralphm

    All right.

  281. nyco

    ok

  282. ralphm

    5. Date of Next

  283. nyco

    +1w, +1

  284. nyco

    ;-)

  285. Guus

    that's the summit

  286. ralphm

    I'm not sure if doing a meeting at the summit is useful

  287. ralphm

    with our regular agenda, that is

  288. Guus

    I'd at the very least would want to be attending the summit at that hour

  289. ralphm

    So I propose we do our next normal meeting in 2 weeks

  290. MattJ

    I'm inclined to agree

  291. Guus

    perhaps we could do the prio thingy in the sidelines of the summit, but for the regular meeting, +2w would be my preference.

  292. Martin

    +2W sounds good to me

  293. ralphm

    good

  294. ralphm

    6. Close

  295. ralphm

    Thanks all!

  296. nyco

    merci !

  297. ralphm bangs gavel

  298. jonasw

    \o/

  299. Guus

    thank you

  300. ralphm set the topic to

    XSF Discussion | Logs: http://logs.xmpp.org/xsf/ | Agenda https://trello.com/b/Dn6IQOu0/board-meetings

  301. MattJ

    Thanks

  302. ralphm

    And see most of you next time, of course!

  303. nyco

    ;-)

  304. ralphm

    next week that is

  305. nyco

    weather not super optimistic

  306. Ge0rG

    Sigh. It's been almost a month and I haven't yet managed to reach stpeter. Maybe I need to switch to email

  307. intosi

    Ge0rG: usually the best option.

  308. jonasw

    Ge0rG, or comment on random old issues on the xmpp.org repository :D

  309. jonasw

    which, probably, is swithcing to email

  310. Ge0rG

    jonasw: TBH I just didn't want to implement Data Forms in yaxim for preauth-IBR :<

  311. jonasw

    Ge0rG, you don’t need to fully implement them, do you?

  312. jonasw

    like with ad-hoc

  313. jonasw

    more worrisome is that IBR doesn’t seem to define form field names for username and password

  314. SouL

    jonasw: if a XEP that has been deferred gets an implementation, will become again draft at least? Do you know?

  315. jonasw

    SouL, no

  316. jonasw

    it won’t

  317. SouL

    Oh

  318. jonasw

    the state transition is Experimental -> Deferred, not Draft -> Deferred

  319. jonasw

    to get a XEP from Deferred to Experimental, all it needs is an update by the author

  320. jonasw

    (non-editoral update)

  321. SouL

    What if it does not need any update?

  322. jonasw

    then the author can ask council to move it to Draft

  323. jonasw

    for this it needs to pass through Last Call

  324. SouL

    Ah ok!

  325. jonasw

    I don’t think that Draft needs to have implementations technically, but council will like it if it does

  326. SouL

    I will and then ask again..

  327. SouL

    😞

  328. jonasw

    but advancing a XEP to Draft without implementations is likely to be a bad move; without implementations, there’s no deployment experience and there will likely be issues which should be fixed before Draft

  329. SouL

    Sorry jonasw, I said Draft but I wanted do say Experimental

  330. SouL

    So Experimental -> Deferred -> Experimental

  331. SouL

    My bad for causing confusion

  332. SouL

    I guess the case would be the same, vote again, like a ProtoXEP?

  333. jonasw

    SouL, I gave you some feedback :)

  334. jonasw

    (on list)

  335. jonasw

    no, Deferred -> Experimental is purely procedural (just like Experimental -> Deferred) and does not need a vote.

  336. jonasw

    if the author publishes an update on a Deferred XEP, it automatically becomes Experimental again, no vote needed.

  337. SouL

    jonasw: I see you sent some feedback, thank you, when it was submitted, there were not many comments on it. I will take a look, thanks for clearing things up too

  338. zinid

    is processing hints supposed to be deferred?

  339. jonasw

    zinid, it was experimental before, it hasn’t been updated in 12 months, it gets deferred

  340. jonasw

    there was some discussion around deprecating it, but I don’t think there has been a vote

  341. jonasw

    SouL, you’re welcome! I didn’t see it when it went Experimental

  342. zinid

    jonasw, I know that it is deferred due to inactivity, I just wonder what council think about it

  343. jonasw

    SouL, so this automated status change to Deferred is a good thing; it reminds the community that the XEP exists and something should be done about it.

  344. jonasw

    zinid, we’ll see

  345. zinid

    yes, deferring due to inactivity is a good idea

  346. goffi

    hi there https://news.ycombinator.com/item?id=16234057

  347. goffi

    it's a link to my blog to present new web framework feature

  348. Ge0rG

    goffi: what's the title? when deep-linking, HN will downgrade the votes

  349. Ge0rG

    ah, "Show HN: Decentralized web framework based on XMPP (Salut à Toi)" in https://news.ycombinator.com/newest

  350. jonasw

    goffi, feature request for https://www.goffi.org/b/96207aea-9bd8-4333-a346-63638c041ef7/build-

  351. jonasw

    click to expand should not be necessary when viewing a single post

  352. SamWhited

    It might just be me, but I didn't actually understand what the thing was that post was talking about (is it just web auth using your XMPP account?). Maybe link to the source or a tutorial or somewhere that provides more information?

  353. SamWhited

    It sounded intriguing though!

  354. moparisthebest

    has anyone done anything like uh, imap over xmpp

  355. jonasw

    moparisthebest, I think Salut á Toi has something along those lines

  356. Zash

    in what sense?

  357. jonasw

    more like XMPP over imap though

  358. moparisthebest

    well, my phone currently keeps 2 TCP connections open/idle, XMPP and IMAP IDLE

  359. moparisthebest

    it would be handy if the xmpp app just passed new messages to the email app, and I only needed 1 connection :)

  360. Zash

    Tunnel them both over ssh

  361. jonasw

    moparisthebest, why do you care?

  362. Zash

    Didn't GTalk have some protocol for notifying about new messages?

  363. moparisthebest

    doesn't really solve the problem, imap idle has to refresh the connection every so often etc

  364. moparisthebest

    jonasw, yea to be honest it's basically not a problem, just wondered if it'd been done before, I'm always surprised by random XEPs

  365. goffi

    jonasw: yes I know, it's not stable release yet, it will be done then

  366. jonasw

    goffi, maybe a bad move to announce it on HN then?

  367. jonasw

    but that’s just me

  368. goffi

    SamWhited: planing to write a tutorial too, but it takes time (and I have a paid job not related to XMPP)

  369. goffi

    moparisthebest: yes, we have SMTP and IMAP, not as a gateway but client side (to use Thunderbird to read your XMPP messages)

  370. moparisthebest

    ah the opposite, still interesting goffi

  371. goffi

    note that it was before than Thunderbird actually read XMPP ;)

  372. goffi

    jonasw: well Show HN is exactly for that if I understood correctly

  373. goffi

    stuff advanced enough to try, but not necessarly finished and production ready

  374. goffi

    any way when I post to HN I don't expect it to reach first page :)

  375. jonasw

    goffi, I don’t visit HN, so I have no idea :)

  376. moparisthebest

    it might be interesting to just replace sending and recieving in an email client with doing that over xmpp (which at server side would just shove it off to a real imap/smtp server)

  377. jonasw

    moparisthebest, no, it would not

  378. jonasw

    you would always incur the factor 4/3 for base64-ing things

  379. jonasw

    unless you map mime to xmpp with SHIM, whihc probably has its own set of interesting issues

  380. moparisthebest

    it's all text based, why not shove it through un-encoded?

  381. jonasw

    I /think/ that you can do 8-bit things in SMTP

  382. goffi

    oh actually it's first in "show" section (https://news.ycombinator.com/show)

  383. jonasw

    moparisthebest, also, SMTP and IMAP are octet-based

  384. jonasw

    not text-based

  385. jonasw

    (like XML is)

  386. goffi

    moparisthebest: there is an email gateway somewhere in my TODO list

  387. jonasw

    at least on the layer the application sees, XMPP is text-based (it has a well-defined encoding everywhere). in email, there’s no such thing (see the hacks you need to do to get unicode into a subject line)

  388. moparisthebest

    jonasw, I mean you can send emails via SMTP with telnet, text based enough

  389. jonasw

    of course, all of this could be solved in one way or another (e.g. by fully mapping MIME into XML somehow)

  390. Link Mauve

    SMTP at least is not octet-based, it’s ASCII-based.

  391. jonasw

    moparisthebest, telnet is byte-based too

  392. Link Mauve

    You can’t use more than 7 bits per octet.

  393. jonasw

    you can, with the 8BIT extension; but that’s not my point, really

  394. moparisthebest

    it reads lines one by one etc etc

  395. Link Mauve

    Hence the encoding monsters.

  396. moparisthebest

    I think it'd be fairly straightforward, not sure

  397. jonasw

    moparisthebest, that’s not text-based ;)

  398. moparisthebest

    version #1 would just replace imap idle for push notifications

  399. moparisthebest

    over xmpp

  400. goffi

    and commented by edhelas from Movim. But we didn't made a video to show how it's cool and never seen before :D

  401. Dave Cridland

    Link Mauve, Actually, IMAP is 8-bit clean, and ESMTP normally is with the extension. Both can also transfer binary (instead of, say, UTF-8) with (rare) extensions.

  402. Dave Cridland

    There's also a bunch of extensions to "Always UTF-8", roughly, as part of EAI, as I recall.

  403. Link Mauve

    I know pretty much nothing about IMAP, this was about plain SMTP.

  404. Link Mauve

    Thanks. :)

  405. Link Mauve

    I’ll read about them.