kentonv changed the topic of #sandstorm to: Welcome to #sandstorm: home of all things sandstorm.io. Say hi! | Have a question but no one is here? Try asking in the discussion group: https://groups.google.com/group/sandstorm-dev | Public logs at https://botbot.me/freenode/sandstorm/
nicoo has quit [Ping timeout: 255 seconds]
nicoo has joined #sandstorm
namark has quit [Quit: Leaving.]
samba_ has quit [Quit: WeeChat 2.1]
jemc has quit [Ping timeout: 256 seconds]
jemc has joined #sandstorm
<ocdtrekkie> I am not sure I have ever seen something so dystopian as what I watched today.
<ocdtrekkie> For the first time, I am actually literally terrified of where we are headed.
jemc has quit [Ping timeout: 240 seconds]
xet7 has joined #sandstorm
<ccx> ?
nwf has quit [Ping timeout: 260 seconds]
_whitelogger has joined #sandstorm
namark has joined #sandstorm
jemc has joined #sandstorm
zopsi has quit [Ping timeout: 255 seconds]
nicoo has quit [Remote host closed the connection]
nicoo has joined #sandstorm
samba_ has joined #sandstorm
jemc has quit [Read error: Connection reset by peer]
<TimMc> ocdtrekkie: spill
jemc has joined #sandstorm
samba_ has quit [Ping timeout: 260 seconds]
zopsi has joined #sandstorm
ocdtr_web has joined #sandstorm
samba_ has joined #sandstorm
<ocdtr_web> The Google I/O crud. Smart Compose and particularly Duplex. An insane point where they introduced that their AI is going to call people pretending to be human five minutes after bragging about how few voice recordings they now needed to fully emulate someone's voice.
<ocdtr_web> I am actually unsure how they operated their supposedly real demo without violating the law, given that recording a phone call requires two-party consent in California. Presumably they either staged it or went to a friendlier legal jurisdiction to do it.
<ocdtr_web> Perhaps sought consent in advance. But nonetheless, it's hard to see how that service could even be legal in much of the United States, considering you would arguably have to record the call to be sending the audio to Google's AI for it to process.
<jemc> perhaps trying to stretch the legal definition of what it means to "record" a call
<jemc> something about streaming vs retention
<ocdtr_web> They also state nothing about their plans on retention, but generally Google improves their models via new recorded data.
<ocdtr_web> It is hard to see how they intend to operate this product without recording the calls outright.
<ocdtr_web> I wonder if they might "stream only" calls involving states with two party consent, and record any data between people in one party consent, since they can surely fit "permission to record all your calls" into their Terms of Service.
<ocdtr_web> Or arguably, Google can consent to record a call between Google's AI and another caller, in a one party consent state.
Telesight has joined #sandstorm
<ocdtr_web> But between that and Smart Compose, which pushes Gmail users into letting the AI write their emails for them, it seems like we're increasingly moving into a space where not using Google products isn't enough: I need to cut myself off from people using Google products.
_whitelogger has joined #sandstorm
LostInTheSauce has joined #sandstorm
<LostInTheSauce> Hello, I have a question about DNS entries for sandstorm install
<ocdtr_web> Fire away
<ocdtr_web> (Though I cannot guarantee I know the answer)
<LostInTheSauce> My site keeps failing the self-test
<LostInTheSauce> I have a wildcar dns entry but I just can't get it to work
<LostInTheSauce> I see the errors in the javascript console
<LostInTheSauce> and I can do a lookup of the self-test url, but for whatever reason it keeps failing
<LostInTheSauce> Sandstorm WILDCARD_HOST self-test failed. Details: Error: network
<LostInTheSauce> at XMLHttpRequest.m.onreadystatechange (f1aa7768e91637c9cc03f30953cf15e7042e71d8.js?meteor_js_resource=true:25)
<LostInTheSauce> those are the javascript console errors
<ocdtr_web> Does your DNS provider support sub-sub-domains?
<LostInTheSauce> this is locally hosted
<LostInTheSauce> i have a bind dns server
<LostInTheSauce> i can do lookups and get responses for sub domains
<LostInTheSauce> sub sub domains that is
<ocdtr_web> Ah, so this DNS entry isn't served out to the web. Hmmm.
<ocdtr_web> What's your wildcard host entry in your config look like?
<LostInTheSauce> i have tried many variants
<LostInTheSauce> but currenlty it is:
<LostInTheSauce> hold on, guess my last attempt to fix it broke it
<LostInTheSauce> but i have tried *.sub. A ip
<LostInTheSauce> i have tried CNAMEs
<LostInTheSauce> multi variations work with dig,nslookup,host etc
<LostInTheSauce> just never makes sandstorm happy
<ocdtr_web> WILDCARD_HOST=*.proasis.provatek.com:6080 would be my assumption based on the URL in the error message above.
<LostInTheSauce> yeah thats what the config for sandstorm looks like
<LostInTheSauce> i thought you meant my DNS server
<ocdtr_web> It's highly likely I won't understand your DNS server config. This stuff is fairly far outside my area of expertise.
<ocdtr_web> Are you using anything in front of this with regards to Nginx or a cert?
<LostInTheSauce> no
<ocdtr_web> That rules out all of the ideas from old issues I'm reading.
<ocdtr_web> If you open a browser window and go to http://foo.proasis.provatek.com:6080 what do you get?
<LostInTheSauce> site can't be reached
<LostInTheSauce> I think i know the problem
<ocdtr_web> If the wildcard was working, that should've said something like "no such grain for id: foo", in theory.
<ocdtr_web> Let me know if you figure something out.
LostInTheSauce has quit [Quit: http://www.kiwiirc.com/ - A hand crafted IRC client]
jemc has quit [Ping timeout: 246 seconds]
_whitelogger has joined #sandstorm
drebs has joined #sandstorm
<drebs> hello,! I'm haviing trouble debugging email sending errors in sandstorm, maybe some of you might bring some light!? :)
<drebs> first, i'm a bit confused on where to find logs for my sandstorm installation
<drebs> should they be in /opt/sandstorm/var/log/sandstorm.log ?
<drebs> when i test email sending in sandstorm admin page, the error shows up there but not in the logs
<drebs> oops, now it showed. Good first result of rubber ducking ;)
<drebs> well, not sure, actually, the error when sending email is different from errors i see in the .log file
<drebs> OK, so this is the error I see when testing email sending:
<drebs> Error while trying to send test email: {"code":"ECONNECTION","errno":"EBUSY","syscall":"getaddrinfo","hostname":"mail.gandi.net","host":"mail.gandi.net","port":465,"command":"CONN"} [other-email-sending-error]
<drebs> it used to work correctly until some days ago, and testing with swaks from command line in the same server using the same account works fine.
<drebs> Can someone give me a clue on what might be wrong here?
jemc has quit [Ping timeout: 256 seconds]
jemc has joined #sandstorm
Telesight has quit [Remote host closed the connection]
<TimMc> drebs: That's pretty weird. Doesn't ring a bell...
isd has joined #sandstorm
natea has joined #sandstorm
drebs has quit [Read error: Connection reset by peer]
natea has quit [Quit: natea]
samba_ has quit [Ping timeout: 250 seconds]
samba_ has joined #sandstorm
harish has joined #sandstorm
harish has quit [Remote host closed the connection]
xet7 has quit [Remote host closed the connection]
<ocdtrekkie> Someone else had more or less the same message on the 27th, though they had ECONNREFUSED and drebs got EBUSY