#social 2018-04-20

2018-04-20 UTC
#
dansup
Thanks for the opinions on javascript!
timbl joined the channel
#
ajordan
dansup: can't tell if you're being sarcastic? I assume no?
#
dansup
Heh, nope
#
dansup
I was just wondering what people thought about it, I know mastodon uses react and quitter is a js heavy UI.
#
Loqi
[Kit Darko] @dansup A full page reload just for liking something though? 🤮
#
dansup
I'll be using js but it wont be required, the js experience will be a lot smoother though
JanKusanagi joined the channel
#
ajordan
dansup: excellent haha, just checking
#
ajordan
I mean FWIW pump.io requires JS too, but that bothers me
#
ajordan
honestly I want to throw away all the frontend code we have currently and try for a simpler model
#
ajordan
and also I'd like to do a better job of supporting people without JS on
#
ajordan
Legacy(tm)
#
dansup
Nice, I think a fresh UI could help both pump and gnu social a lot!
#
dansup
I started a gnu social clone using the laravel framework, https://mastodon.social/@dansup/99807919226835952
#
Loqi
[dansup] Sneak peek of the basic bootstrap UI.
evanp joined the channel
#
dansup
hey evanp, big fan of your work on statusnet and pump! I've recently been working on new patches for GNU social (including ActivityPub support), and tinkering with stuff like https://git.gnu.io/gnu/gnu-social/issues/308
#
dansup
I ran a statusnet instance from 2012 to 2015 on the hyperboria network (wasnt federated since it required cjdns access)
evanp joined the channel
#
ajordan
ahh so to be clear dansup
#
ajordan
that refactoring I was talking about would be different than a UI redesign :-)
#
ajordan
though I'd like to improve the web UI too haha
#
dansup
ajordan: Oh I see, I might be able to help with that. I'm familiar with bootstrap, bulma, semantic and foundation. https://dansup.github.io/bulma-templates/
#
ajordan
ah nice
#
ajordan
I mean so am I, it's just a matter of time
#
ajordan
I'm bad about reviewing PRs too unfortunately
#
ajordan
there's some low-hanging fruit like e.g. getting us off Bootstrap 2 lol
#
dansup
Yeah I hear ya, I've been trying to add bcrypt/argon2I support to GNU social and performance/security takes precedence over UI
#
ajordan
yeah for sure
#
ajordan
I have so many ideas to improve security
#
ajordan
just never enough time...
evanp and cdchapman joined the channel
#
dansup
ajordan: does pump support ActivityPub or does it just support ActivityStreams 1.0?
#
ajordan
dansup: AP support is WIP
#
ajordan
actively being worked on
#
dansup
ah nice
cdchapman joined the channel
#
ajordan
I expect AS2 support will ship in our next beta
#
ajordan
that's blocked on me doing a review unfortunately
#
dansup
if you support image uploads, I will be able to federate your content to the instagram clone
#
ajordan
we do!
#
ajordan
(fair warning I'm going to bed literally any minute now)
#
dansup
goodnight!
#
Loqi
sleep tight!
#
cwebber2
ajordan: :D :D
#
cwebber2
go go AP in pump.io!
vasilakisfil, xmpp-social, mib_lfcapy, githree, fr33domlover, Zakia, gf, hadleybeeman and downey joined the channel
#
Loqi
[Gargron] In that case maybe you don't want polls, maybe you want to use Strawpoll... Either that or ActivityPub needs a way to display answer counts without listing individual answers. Tbh, imagine a typical poll on Twitter, it can get thousands of replies. T...
#
cwebber2
Gargron: *reading*
#
cwebber2
Gargron: I gave a suggestion
#
Loqi
Ok, I'll tell them that when I see them next
#
Loqi
[cwebber] Maybe each answer that shows up in the poll could itself be an AS2 collection which can give its totalItems, but which can (optionally, if non-anonymous) also expose its items (which are the actual response objects) in case anyone wants to "audit" it...
#
Gargron
is evan back in activitypub development?
#
ajordan
Gargron: evan's been hacking on AP in pump.io
#
cwebber2
!tell pfrazee is this true? does Dat use nodejs' representation of json as its "canoicalization" format? https://news.ycombinator.com/item?id=16879954 Is there any plans to change that?
#
Loqi
Ok, I'll tell them that when I see them next
#
Loqi
[Vendan] (I'm the author of the repo in question) Note that this repo has been sidelined, as I have fundamental issues with the protocol SSB is built on. Unless there's changes to how the messages are signed and verified, I'm not planning on putting any seri...
#
cwebber2
oh wait
#
cwebber2
SSB not dat
#
cwebber2
nm then :D
#
cwebber2
I keep mixing up SSB and dat
#
cwebber2
!tell pfrazee once again I mixed up SSB and dat apparently. Never mind and carry on!
#
Loqi
Ok, I'll tell them that when I see them next
evanp joined the channel
#
cwebber2
hey hey evanp
#
cwebber2
I left you messages with Loqi ;)
#
cwebber2
I think you need to speak to get them! :)
#
ajordan
ahhh SSB
#
ajordan
it *does* have a lot of interesting ideas
#
Loqi
[zeveb] I know that I sound like a broken record, but this is exactly the issue which canonical S-expressions were designed for, and which SPKI wrestled with & solved twenty years ago. The SPKI version of a message would look something like (I've removed...
#
cwebber2
yeah yeah canonical s-exps for life
#
cwebber2
notes they co-authored a protocol that does not use canonical s-exps but shhh
evanp joined the channel
#
ajordan
cwebber2: heathen!
#
ajordan
has exactly zero opinion on canonical s-exps
#
cwebber2
canonical s-exps are great and it makes me sad that SPKI/SDSI didn't take off (well, I'm sad about that for lots of reasons)
#
ajordan
searches
#
ajordan
aw man
#
ajordan
that looks really nice :/
#
cwebber2
yeah instead we got X.509 and SSL Certificate Authorities
#
ajordan
no kidding
#
ajordan
at least we have Let's Encrypt I guess?
#
cwebber2
Let's Encrypt is a useful tarp (built an maintained by wonderful people) thrown over the current dystopia of certificate authorities
#
cwebber2
if you haven't considered it before, consider that SSL CAs are only as strong as the *weakest* CA you have
#
cwebber2
any CA, if compromised or unethical, can lie
#
cwebber2
and in fact, some CAs will lie if you pay them to
#
cwebber2
I forget which but some of the CAs allow corporate environments to purchase a "service" where they will allow you to MITM sites on a corporate network
#
ajordan
cwebber2: I'm aware of all the problems with PKI (moxie's talk on Convergence is really good for that and also I taught a class on this lol)
#
ajordan
and I totally agree
#
michiel
cwebber2: would SPKI have solved that issue?
#
ajordan
right now the situtation is still pretty bad but I think Certificate Transparency will actually get us to a decent place
#
cwebber2
michiel: yes, SPKI/SDSI as I understand it allowed there to not be one "central authortiy" but allowed one to set up a trust network
#
cwebber2
michiel: it had a petnames system using path based names IIRC
#
cwebber2
however most of my knowledge comes from only skimming the specs, but reading and falling in love with the canonical s-exps one ;)
#
cwebber2
and talking with ocap folks
#
cwebber2
and also Christopher Allen
#
cwebber2
who is the person behind tls/ssl
#
cwebber2
and also *did not want* a certificate authority system, and wanted to use SPKI/SDSI
#
cwebber2
well, the main person behind, I should say
#
ajordan
oh interesting
#
ajordan
according to someone moxie talked to they just "threw [PKI] in at the end"
#
cwebber2
btw if you read the petnames thing (unfinished) that I was working on with Mark Miller and etc for Rebooting Web of Trust that's a good idea of how a different naming and security model could work: https://github.com/cwebber/rebooting-the-web-of-trust-spring2018/blob/petnames/draft-documents/making-dids-invisible-with-petnames.md
#
ajordan
I wonder if he talked to Christopher Allen or someone else
#
cwebber2
sadly unfinished
#
cwebber2
yes Christopher Allen told me something like "we tried to warn Netscape about all the problems that would happen if they went with a CA model but they were like, this is easy to ship, so let's roll it out"
#
cwebber2
the petnames stuff will probably be more important as we're starting to see more experiments with activitypub over tor onion services
#
ajordan
Kip Hickman (sp?) apparently
#
cwebber2
and webfinger sure as heck won't work for that :P
#
cwebber2
especially not as onion v2 services roll out with a zillion character long onion names
#
cwebber2
petnames reduce phishing risks as well
#
ajordan
lol I've had a vanity onion v2 generator running for... let's see
#
ajordan
two months apparently
#
ajordan
I should just give up right?
#
Loqi
[AJ Jordan] I'm now generating (vanity) Tor onion service keys for strugee.net and nodecompat.com. Onion services coming Real Soon Now™!
#
cwebber2
vanity .onions are funny but also kind of dangerous
#
ajordan
dangerous how?
#
cwebber2
especially the facebook one worries me
#
cwebber2
they increase phishability
#
cwebber2
if you get used to the idea that something like facebookcorewwwi.onion is really facebook
#
ajordan
ohhh yeah that's interesting
#
cwebber2
then facebookcorewwii.onion may be an easy way to trick people
#
cwebber2
and this will be worse in tor onion services v2
#
cwebber2
where they're much larger
#
cwebber2
even harder to distinguish
#
cwebber2
if you only look at the first N characters, you might say "seems legit?" and go through
#
cwebber2
and then bam, phished
#
cwebber2
petnames are the right answer
#
ajordan
well that depends on how you keep track of onions
#
ajordan
I have a text file where I record known onions
#
cwebber2
ajordan: that's an informal petnames system :)
#
cwebber2
now let's make a formal one
#
ajordan
you could argue that v2 will improve things because it forces more people into that kind of setup
#
ajordan
ah true
#
cwebber2
bookmarks are also petnames
#
cwebber2
now, if you add edge names
#
cwebber2
where I can recommend you
#
cwebber2
er, recommend to you
#
cwebber2
some names
#
cwebber2
cwebber => rhiaro
#
cwebber2
here's how you find Amy
#
cwebber2
now you start to establish a system where, like real world connections between people
#
cwebber2
we can introduce folks to each other
#
cwebber2
and also!
#
cwebber2
DNS doesn't need to appear
#
cwebber2
it just becomes an equal participant
#
cwebber2
if you have a dns petname, then
#
cwebber2
dns => dustycloud.org
#
cwebber2
er, doesn't need to disappear :)
#
cwebber2
same thing with namecoin or even your local university staff directory
#
rhiaro
considers herself found
#
cwebber2
MIT Staff => Gerald Sussman
#
cwebber2
hi rhiaro :)
#
ajordan
hm interesting
#
ajordan
also side note because of this conversation I was like I wonder what rhiaro is up to? and so I went to http://amy.gy and well
#
ajordan
poop emoji :P
#
rhiaro
That's because online rhiaro is authoritative not amyg
#
rhiaro
rhiaro.co.uk should tell you what I'm up to more or less
#
ajordan
lol yeah I went there
#
rhiaro
is on a bus to Prague
#
ajordan
I couldn't remember the URL and wanted to be lazy so I took a guess :D
#
ajordan
rhiaro: nice, how come?
#
rhiaro
Where to start
#
ajordan
hahaha
#
Loqi
hehe
#
rhiaro
Convenient connection to Lyon for webconf is the simplest explanation
#
rhiaro
Anyone here going to webconf next week? I know Sandro is
#
ajordan
ahh nice
evanp left the channel
#
aaronpk
I haven't been paying enough attention to conferences ahead of time to actually go to much this year. Hopefully next year I can do more since I can plan ahead now
cdchapman, jankusanagi_, fr33domlover, JanKusanagi and bwn joined the channel
cdchapman joined the channel