<_habnabit>
dwrensha, so it's intentional that when pipelining you're just given a Reader even when it might be a type you've declared locally in your own vat, right? is it possible to reflect the actual type out of it?
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 246 seconds]
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 240 seconds]
n8a has quit [Ping timeout: 245 seconds]
n8a has joined #sandstorm
afuentes has quit [Ping timeout: 258 seconds]
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 258 seconds]
<dwrensha>
_habnabit: The Pipeline should be properly typed.
<dwrensha>
_habnabit: what do you mean by "a type you'e declared locally in your own vat"?
<_habnabit>
dwrensha, i mean that i, the server, declared a type that implements an interface, and so if the client is pipelining, i want to be able to get the instance of my type back out when it's passed in to an rpc method that takes an instance of that interface
<dwrensha>
so something like `interface foo { bar @0 () -> Baz; useBaz @1 Baz -> (); }`
<dwrensha>
the server has implementations of Foo and Baz
<_habnabit>
right
<dwrensha>
and if the client passes a Baz back in to `useBaz()`, the server ought to be able to use the concrete type
<_habnabit>
yes, exactly
<_habnabit>
but iff it's the right Baz
<_habnabit>
e.g. in a >2 party system, not someone else's remote Baz
<DanC_>
is there really so little demand to use spk-dev on Ubuntu 16.04?
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 246 seconds]
isd1 has joined #sandstorm
isd1 has quit [Read error: Connection reset by peer]
isd has quit [Remote host closed the connection]
jemc has joined #sandstorm
jemc has quit [Ping timeout: 248 seconds]
robbintt has joined #sandstorm
<robbintt>
o/ I'm having an issue in the noisebridge sandcats instance.
jadewang has joined #sandstorm
<robbintt>
I've tried sharing a grain inside both gogs and gitlab -- both grains did not autocomplete valid users and said "Unknown user" for a user that exists.
<robbintt>
Also, the `email invite` feature didn't deploy an email.
<robbintt>
I have admin privileges, the 'original admin' for the account was able to add and did see autocomplete. Tested on chrome, firefox, safari.
<robbintt>
clarification: the 'original admin' for the 'sandcats instance' was able to add and did see autocomplete. he has a different account
<dwrensha>
robbintt: autocomplete will only work for users that are among your "contacts"
jadewang has quit [Ping timeout: 245 seconds]
<robbintt>
dwrensha: good to know, what is "contacts"? I did a quick google and checked all my panels
<dwrensha>
someone gets added to your contacts when you share a grain to them via a secret link
<dwrensha>
"contacts" is kind of an implicit concept right now
<dwrensha>
so if you haven't shared with someone yet, you need to share via link
<dwrensha>
and then after that they'll show up in autocomplete
<robbintt>
I see. so I have to share a grain with the person with a secret link, then i can share a grain with them via the autocomplete panel?
<dwrensha>
yeah
<robbintt>
ok np
<robbintt>
do you know offhand where this contacts concept is managed? i would really like to add a 'manually add contact' feature or something.
<robbintt>
doing an n-way handshake with my organization won't work.
jemc has joined #sandstorm
<robbintt>
anyways, no insult intended. i just want to improve the feature.