#indieweb 2017-08-22

2017-08-22 UTC
#
amz3` If my understanding is correct, the notion of time is loose in a distributed network
#
amz3` if say, some one claim posted something @ t0
#
amz3` if say, some one claim posted something @ t1
#
amz3` someelse can as easily claim he posted something @ t0 that is
#
amz3` before
#
amz3` so... time is built in a distributed social network based on pure trust of the participant to behave correctly and properly use timestamps
#
Loqi [superfeedr] "? Mastodon is big in Japan. The reason why is… uncomfortable by Ethan Zuckerman" by Chris Aldrich on 2017-08-20 http://boffosocko.com/2017/08/19/mastodon-is-big-in-japan-the-reason-why-is-uncomfortable-by-ethan-zuckerman/
#
amz3` you can verify the time something was posted, except if you keep track of the whole distributed system
#
amz3` at some point
#
amz3` the only thing that you can keep track of, i think, is *precedence*
#
KartikPrabhu amz3`: what is the use-case?
#
aaronpk i suppose so? that seems like overthinking it
#
KartikPrabhu i simply trust the timestamp of the replies I get because I don't see the downside
#
KartikPrabhu what is reply?
#
Loqi A reply (or comment) is a kind of post that is a text (typically, though photos are possible too) response to some other post, that makes little or no sense without reading or at least knowing the context of the source post https://indieweb.org/reply
#
amz3` Even precedence, can not be trusted, you can reply to a post that doesn't exist yet...
#
amz3` maybe i am overthinking it
#
KartikPrabhu amz3`: I don't see a use case for this level of "trust"
#
amz3` my point, is that there is not interest in keeping track of time in distributed social network
#
GWG Evening
#
KartikPrabhu amz3`: maybe, why should there be unless there is a use-case for it
#
aaronpk the one thing i find interesting about this idea is that when you want to "prove" to someone that you posted something on a certain date you can really only do so by pointing to a copy of it somewhere where you are unable to modify the creation date, e.g. a twitter copy or web.archive.org
#
aaronpk like if i said "i totally said this thing 5 years ago" and sent you a link to a post on my site, you have no reason to believe it
#
KartikPrabhu aaronpk: you can modify a twitter copy by deleting it and writing anothre tweet with a different time
#
KartikPrabhu also again what is the use-case for such a "proof"?
#
aaronpk KartikPrabhu: well the trick on twitter is to set an account to private, write multiple tweets with different outcomes, then delete all but the desired tweet and make the account public
#
KartikPrabhu aaronpk: yup also that
#
aaronpk i can't remember off-hand but i have definitely seen examples of people referencing old tweets to make a point about when something was claimed
#
KartikPrabhu also I write posts with a futured publish date which counts as a draft and people with the URL can still reply to it even though it isn't published yet!
#
amz3` KartikPrabhu: the use case, is in science for instance, to tell whoever made the discovery first
#
KartikPrabhu amz3`: even current publishing does not guarantee that
#
amz3` agree
#
KartikPrabhu unless you trust the publisher who can use any date they please
#
amz3` it's might explain so many paper don't have dates
#
amz3` *why so many *papers
#
KartikPrabhu you can't even guarantee that a printed book was actually published on the date printed in it
#
amz3` yes
#
amz3` what matters is what comes next, who links to your published asset and the trust that is around it
#
amz3` if time is dubbed significant
#
KartikPrabhu amz3`: I still don't see why it "matters"
#
amz3` I am trying to build a "static" social network on top of git.
#
aaronpk can't you fake timestamps in git too?
#
amz3` I am building a file based database that is stored in git
#
amz3` the file based database is happend only
#
amz3` s/happend/append/
#
amz3` yes you can
#
amz3` fake timestamps in git too
#
amz3` but I was wondering how I could create an UX with time that can be fake
#
aaronpk yes https://www.google.com/search?q=github+commit+graph+art&source=lnms&tbm=isch
#
amz3` considering time can be fake
#
amz3` and basicaly, since time can not be trusted, I can not rely on time to create "list of posts"
#
KartikPrabhu amz3`: when someone hits publish you automatically use the current time
#
KartikPrabhu don't let user select the published time
#
amz3` the whole database is shared among participant, i don't control what they put in the database, only that they can't modify old entries
#
aaronpk there are two timestamps to consider when building a /reader ... the time the post self-reports as the published time, and the time the reader found it. if you want to show the user a list of things they haven't seen yet, use the time the reader time. if you want to show a list of things in supposedly chronological order, show the published time
Exodist joined the channel
#
amz3` aaronpk: that's interesting
#
amz3` aaronpk: I was thinking about: “the reader time”
#
amz3` keeping track of what's new/fresh in the social network and focus on that
#
amz3` I mean use that information to display things
#
amz3` tx
#
amz3` I think, I won't include publication time, it will be too confusing
#
amz3` since it can not be trusted
#
amz3` I made good progress on this project, I started it like... 2 or 3 years ago
#
amz3` tantek seemed interested but then I got side tracked
#
amz3` maybe this time I can make it happen
#
amz3` the idea behind that, is to have some kind of collective work that is never in the hand of a single person
#
amz3` well silos... you know the story
#
amz3` it was delight to discuss with you
#
amz3` tx
wolftune and [kevinmarks] joined the channel
#
[kevinmarks] This problem of public timestamping is solved by Bitcoin, but at a very high cost
#
aaronpk no, that system just has different tradeoffs
#
aaronpk for example you *can't* backdate things, which sometimes is a thing i want to do
[manton] joined the channel
#
[manton] In Micro.blog I use the published date reported by the blog and sort on that. I had planned to do more checking to ensure the dates aren't wildly off, but in practice it hasn't been a problem yet. (Checking that they aren't in the future would solve most problems.)
#
aaronpk yeah in practice people don't actually try to mess with timestamps for this stuff
#
[manton] This is also related to post editing, which I think has more potential for abuse.
#
aaronpk i think the most likely error is around timezones. like forgetting to include a timezone offset and using only the local time.
#
[manton] Yep. There are several places that timezones can be off in a platform like this.
wolftune joined the channel
#
Loqi [superfeedr] "? Pivot time: searching for an Open Web blogging model | AltPlatform" by Chris Aldrich on 2017-08-18 http://boffosocko.com/2017/08/18/pivot-time-searching-for-an-open-web-blogging-model-altplatform/
[cleverdevil] joined the channel
#
Loqi [superfeedr] "Packing day three: Oban 14" by Chris Aldrich on 2017-08-21 http://boffosocko.com/2017/08/21/packing-day-three-oban-14/
mblaney and KartikPrabhu joined the channel
#
mblaney there's another way to possibly solve the timestamp issue, you could ask a 3rd party to store the timestamp and link to the post you want verified.
#
mblaney if others trust the 3rd party, great. if not, add more 3rd party sites to do the verification.
#
mblaney would work for the whole content of the post too, just store a hash and verify that hasn't changed.
#
aaronpk That sounds like an interesting service
#
mblaney yeah I put some thought into it for my distributed payment system thing.... will get to it one day :-P
NrG1z3r joined the channel
#
sknebel good morning indieweb!
#
sknebel reading logs, there is a bunch of existing stuff for timestamping
#
sknebel e.g. signed executables generally have to have not only a a signature, but only a signed timestamp proving *when* they were signed (which really just is the person creating the binary going to a trusted authority and saying "here is a signature I'm attaching to my file, could you please sign a statement saying when you've seen it", which is then attached to the file as well)
#
sknebel certificate transparency logs have interesting properties in that regard as well (they don't explicitly contain timestamps, but they are logs where you can detect changes to the history)
loicm, jihaisse and [kevinmarks] joined the channel
#
[kevinmarks] You could use archive.org to do this. Or hash archive or https://metrognomo.com
cweiske and barpthewire joined the channel
#
cweiske oh. nobody wants linkback support in stackoverflow: https://meta.stackoverflow.com/q/355408/282601
#
cweiske +5/-16
[pfefferle] and DanC joined the channel
#
sknebel huh
#
sknebel also interesting the suggestion in the opposite direction and the assumption that it would "have to be" an annoying notification
friedcell and schmarty joined the channel
#
elie cweiske: I backed you up
#
petermolnar !tell rhiaro is it possible to use aleph as a search engine for our sites? If yes, had anyone done it so far?
#
Loqi Ok, I'll tell them that when I see them next
#
KartikPrabhu what is aleph?
#
Zegnat Good morning IndieWeb
#
KartikPrabhu what is aleph?
#
Loqi It looks like we don't have a page for "aleph" yet. Would you like to create it?
#
KartikPrabhu hmm
#
Zegnat cweiske, not a lot of feedback on why they do not like linkback though :/
#
rhiaro petermolnar: aleph indexes documents like pdfs, emails, spreadsheets, and provides a frontend for sorting documents and extracted entities (people, companies, land) into collections, and cross-referencing them against other collections to see if there are overlaps. You could download a static copy of your website and have an instance of aleph ingest it, but.. I don't know why you would.. that's nto what it's for
#
Loqi rhiaro: tantek left you a message on 2017-03-30 at 1:23am UTC: can you make it to any of the dates here? add your prefs! https://indieweb.org/Planning#Summit
#
Loqi rhiaro: petermolnar left you a message 5 minutes ago: is it possible to use aleph as a search engine for our sites? If yes, had anyone done it so far?
#
petermolnar I was thinking of feeding a lot of extras beside the text to it, exif data in json, csv from sensors, etc, but that's a completely different use case; thanks :)
#
rhiaro when we want to ingest already structured data sources we have to map them from the existing structure to aleph's ES data model. It's arbitrary, you might as well make your own ES backed thing that's more customised to your use case
#
petermolnar indeed
GWG and jeremycherfas joined the channel
#
jeremycherfas Good morning IndieWeb
KartikPrabhu joined the channel
#
jeremycherfas I saw someone else reading this https://static.land yesterday and thought it might be of interest here, but not sure where in the wiki to put it.
#
jeremycherfas Somewhere here https://indieweb.org/web_hosting#Static_domain_hosting
#
sknebel access to the hosted version seems limited, so not sure how well it fits
#
jeremycherfas OK then. I’ll leave it for now.
#
Zegnat This was an interesting read re the growth of Mastodon: http://www.ethanzuckerman.com/blog/2017/08/18/mastodon-is-big-in-japan-the-reason-why-is-uncomfortable/
j4y_funabashi and [kevinmarks] joined the channel
#
j4y_funabashi morning all
#
[kevinmarks] I'm a bit grumpy that ethan and Co wrote about decentralized web without covering indieweb http://dci.mit.edu/decentralizedweb
#
Zegnat [kevinmarks], maybe because of a bias towards self-contained platforms like Mastodon? IndieWeb doesn’t really show well as a solid network. We’re just people with websites that talk to each other.
#
Zegnat It is interesting SOLID is mentioned on the list there (I haven’t read the whole thing). Are there actual websites implementing any of the SOLID stuff?
j12t joined the channel
#
petermolnar decentralized == fragmented variety of random sofwware that can talk to eachother; this is what's missing from 99% of decentralization efforts
#
Zegnat Hard to keep things apart. I refer to https://colinwalker.blog/2017/04/11/federated-decentralised-distributed/ for that
#
Loqi [Colin Walker] Federated, decentralised or distributed
sl007 joined the channel
#
sl007 hi!
#
Zegnat Hiya sl007
#
Zegnat Also, morning j4y_funabashi, almost missed you there!
[colinwalker] joined the channel
#
[colinwalker] @Zegnat - glad you found that useful ?
#
[colinwalker] And I've just spotted a typo ?
#
Zegnat I am not completely sure how useful though, [colinwalker], I still doubt at times what word should be applied to what idea ;) But that’s probably just the nature of it
#
petermolnar "Back to the Future: The Decentralized Web" - new on AWS!
#
petermolnar /s
#
[colinwalker] Yeah, it's a bit vague and I don't think always applied properly.
#
sknebel hm, I have to re-read that again after some more food and coffee
davidmead joined the channel
#
[colinwalker] A **lot** of coffee!
#
Zegnat That should be a warning on top of some posts “Lots of caffeine required”
#
sknebel wrote my initial ideas down, will re-read later and figure out if they still conflict with your post or not ;)
MylesBraithwaite and yar joined the channel
#
[colinwalker] Cool
jihaisse_, Zegnat, Kaja, sknebel_ and camerongray joined the channel
#
j4y_funabashi hey Zegnat, hows things?
micro_ joined the channel
#
Zegnat Pretty good, getting some work rolling. All fine over there j4y_funabashi?
#
j4y_funabashi yeah good, making progress on my new micropub client / reader
nikivi, strk[m], M-hotzeplotz, schmarty, plindner and barpthewire joined the channel
#
j4y_funabashi have got the happy path for login, posting notes + photos and displaying a feed of posts from your micropub server working :)
#
j4y_funabashi I had to modify my micropub server to return a list of posts on the q=source endpoint if no url is provided
#
sl007 Zegnat - it is somehow scary that nobody put their name to the Guest List of https://indieweb.org/2017/Dortmund ? CAMPFIRE - Festival for journalism
#
Zegnat the guestlist where, on the wiki?
#
sl007 yeah
#
Zegnat j4y_funabashi, it is going to be fun to see clients that will actually display old posts as well!
#
sl007 I mean 5000 people will attend the festival but on the other side there are 38 other workshops and how many will attend ...
#
sknebel sl007: people outside the community very likely won't edit the wiki
#
Zegnat sl007, I think too few of the active IndieWeb wiki/IRC users can make it :( And journalists actually there are less likely to login to the wiki to add their names
#
sl007 And the first point is what makes me :(
#
sl007 sknebel - looking at you !
#
sl007 ;)
#
Loqi *sniff*
#
sl007 also - if anybody comes to your mind living "west coast"
#
Zegnat Of the US? San Fran is west right, so tantek and everyone who attends the SF HWC come to mind
#
sknebel of germany I assume ;)
#
sknebel sl007: did you have joschi retweet an announcement or something? quite a few people that were at other camps probably don't really check the wiki or anything, but might see it there
#
j4y_funabashi Zegnat: yeah definitely
[kevinmarks] joined the channel
#
Zegnat Oooh, Germany’s West Coast. The Netherlands then ;)
#
sl007 Zegnat - sorry for the joke … Yes, there are a few thing the Dutch did better like, well, building coasts ;)
#
cweiske they don't have anything else :)
#
sl007 OK, fun fact - meanwhile received a mail "I checked in via sched, do i have to add me to the guest list" - no idea where I can see who "checked in via sched" (which is the festivals system)
#
Zegnat Do you have any numbers on how many people have checked in via sched at all? Maybe just add the number to the wiki?
#
sl007 Zegnat - yeah, just got it - the number is 5 (five)
#
sl007 Gimme 5
#
sl007 I'll add them to the Guest List now - oh wait the next mail is saying I should become a bitcoin millionaire ;)
#
sl007 ups, now it is 6
#
sl007 They opened their workshop registrations on Saturday …
j12t, loicm, gigitux and KartikPrabhu joined the channel
#
schmarty good morning, indieweb!
snarfed, j12t and wolftune joined the channel
#
aaronpk petermolnar++ for this haha https://chat.indieweb.org/2017-08-22/1503396925667000
#
Loqi petermolnar has 29 karma in this channel (46 overall)
#
Loqi [petermolnar] decentralized == fragmented variety of random sofwware that can talk to eachother; this is what's missing from 99% of decentralization efforts
#
Loqi ahahaha
#
@fletcher0xFF @johnbiggs Invite @manton2 to talk about https://micro.blog/ and the #indieweb movement. (twitter.com/_/status/900002627164844032)
[miklb] joined the channel
#
schmarty aaronpk++ for the Loqi triple-action combo
#
Loqi aaronpk has 524 karma in this channel (1414 overall)
#
aaronpk lol
snarfed joined the channel
#
snarfed https://atomenabled.org/ is still half broken and spammed, and emails to mailto:rfc822%3Bwebmaster@atomenabled.org bounce. sad.
#
aaronpk who maintains the site?
#
aaronpk i guess this is what whois was supposed to tell us
#
Zegnat Could it be Sam Ruby? The page links to the Atom wiki, which he hosts
[kevinmarks], wolftune and eli_oat joined the channel
#
[kevinmarks] likely Sam yes
#
@kevinmarks @samruby @timbray Who runs http://atomenabled.org these days? It is being spammed. #indieweb (twitter.com/_/status/900031906896543744)
[miklb], snarfed, gigitux and [manton] joined the channel
#
[manton] Do any of the other HWC cities use meetup.com or mailing lists to help coordinate event times, or is mostly just on the wiki and individual blog posts / tweets? Someone asked if there was a list for Austin, so want to make sure what we're doing is consistent with other cities.
#
sknebel [manton]: we use meetup.com, because that's the place to be for tech meetups in Berlin (sadly, not a fan of it)
#
sknebel general consensus from a HWC organizers session was that the wiki is primarily community-internal documentation and has basically no outreach value for HWC (even if people stumble over it, they then also have to look at HWCs carefully enough to notice there is a local one, ...), to use whatever works locally and that by far the most effective thing is directly contacting/inviting people via e-mail, twitter or in person
#
[manton] @sknebel Got it, thank you. I like starting with a mailing list for anyone interested, but meetup.com is popular here too and probably worth looking at. (I don't use it much myself either.)
#
sknebel meetup is annoyingly expensive, but got at least a few people
#
[miklb] need indieweb friendly sponsors?
#
[miklb] not that I have an option for that, but wondering if something worth adding to /donate page or the like
#
sknebel we'll have to evaluate if we need it that badly once our subscription runs out
#
sknebel really not a fan of spending money on them, even if it isn't mine
#
[miklb] fair
j12t, gigitux, jmelesky and sketchess joined the channel
#
sketchess Good morning Loqi, good morning Indieweb.
gRegorLove and sl007 joined the channel
#
gRegorLove Good morning, #indieweb
#
sketchess Welcome gRegorLove. :)
j12t and sl007 joined the channel
#
@sl007 @MarcusLindemann Weil Du ja auch Multiplikator bist: Ein charmanter Workshop ist https://indieweb.org/2017/Dortmund ? beim CAM… https://twitter.com/i/web/status/900059740092473344 (twitter.com/_/status/900059740092473344)
raretrack joined the channel
#
@sl007 @LiquidNewsroom Ein charmanter workshop https://indieweb.org/2017/Dortmund ? beim CAMPFIRE Festival für Journalismus ! (twitter.com/_/status/900060916875747329)
j12t, j12t_, snarfed, wolftune, j12t__, KartikPrabhu and [jeremycherfas] joined the channel
#
[jeremycherfas] !tell manton No sign of an update the the micro blog app beta.
#
Loqi Ok, I'll tell them that when I see them next
KartikPrabhu, snarfed, sl007 and [manton] joined the channel
#
[manton] jeremycherfas Actually the beta isn't being updated anymore! You can delete it and switch to the App Store version. The App Store version has a bunch of improvements since the beta.
#
Loqi [manton]: [jeremycherfas] left you a message 34 minutes ago: No sign of an update the the micro blog app beta.
#
[manton] (If I had planned things a little better, I would've made the beta automatically notify you about the new version. Sorry about that.)
j12t and j12t_ joined the channel
#
[jeremycherfas] manton. Oh. I had no idea. Will do that.
#
[manton] Great, let me know if you have any problems.
#
GWG Afternoon
#
ben_thatmustbeme waves to manton
j12t joined the channel
#
[manton] ben_thatmustbeme Hello!
#
ben_thatmustbeme how goes everything with micro.blog
#
[manton] Going well, thanks. Still a lot to do, but we're making improvements all the time, getting closer to a public launch.
#
ben_thatmustbeme exciting
#
ben_thatmustbeme btw, jf2 now has a merged jsonfeed/jf2 profile
#
ben_thatmustbeme i actually got a mime type for it
#
ben_thatmustbeme woah, later than i realized
#
ben_thatmustbeme gotta go catch a train
#
snarfed ben_thatmustbeme++
#
Loqi ben_thatmustbeme has 100 karma in this channel (254 overall)
[miklb], snarfed, jmelesky and davidmead joined the channel
#
@schmarty Here's an interview with @macgenie from the most recent This Week in the #IndieWeb. ? Full Episode:… https://twitter.com/i/web/status/900117568849158144 (twitter.com/_/status/900117568849158144)
snarfed, eli_oat and leg joined the channel