<jdenz>
Nothing major and I'm just playing around, but I noticed that the WordPress version in the Sandstorm app is not the most recent. It's 4.3 but there's a 4.3.1 update available. When I check for an update in the WordPress admin in Sandstorm, it doesn't show the update. Is this done on purpose? If so, how does one update an instance? Or ... is this yet another peculiarity of my nginx setup?
home has quit [Quit: Leaving]
<jdenz>
I guess, ultimately, keeping WordPress updated is less important when using Sandstorm.
home has joined #sandstorm
<dwrensha>
jdenz: The app isn't showing the update because it doesn't know about it, because it doesn't have a capability to make external network requests
<jdenz>
Ok, thanks.
<dwrensha>
but even if it could make such requests, it couldn't update, because the WordPress PHP code is in read-only storage
<dwrensha>
so the Sandstorm package would need to be updated for you to get the latest version
home has quit [Read error: Connection reset by peer]
NOTevil has quit [Remote host closed the connection]
<jadewang>
you know what would be totally awesome, @dwrensha?
<jadewang>
skitch could upload to a Sandstorm app
<jadewang>
the auto-download from evernote is pretty annoying, I agree
NOTevil has joined #sandstorm
NOTevil has quit [Quit: Leaving]
jacksingleton has joined #sandstorm
<kentonv>
actual output from a GPG command: gpg: Ohhhh jeeee: ... this is a bug (../../g10/gpg.c:2010:main)
<paulproteus>
kentonv: I am literally lol-ing.
<kentonv>
I have no idea how to tell what I did wrong...
jacksingleton has quit [Ping timeout: 246 seconds]
<kentonv>
paulproteus: I think I read the error in your voice.
mnutt_ has quit [Quit: mnutt_]
gopar has joined #sandstorm
jdenz has left #sandstorm [#sandstorm]
<kentonv>
paulproteus: I guess I'm glad because it convinced me not to use whatever gpg was laying around the server in Blackrock, and instead use the one bundled with Sandstorm, which seems to have fixed the problem (though possibly only because setting up to use the Sandstorm one means entering the self-container)
<paulproteus>
1. Fascinating.
<paulproteus>
2. Consider using gpg2.
<paulproteus>
I'm supposed to tell you to apt install gpg2 but I can't remember why.
<paulproteus>
(Other than the fact that gpg (1) is on life-support and gpg2 is actively maintained.)
<kentonv>
what is the advantage of gpg2?
<paulproteus>
gpg (1.x) I mean not like man 1 gpg
<paulproteus>
"I can't remember why"
<kentonv>
If it's "it has a sane programmatic interface", awesome
<kentonv>
switching should be orthogonal to what I'm doing now in any case
<paulproteus>
I don't know but from man 1 gpg2:
<paulproteus>
Keep in mind that, if this program is used over a network (telnet), it is *very* easy to spy out your
<paulproteus>
passphrase!
<paulproteus>
so hmm.
<paulproteus>
I see no particular significant difference.
<ocdtrekkie>
This is a neat open source web app thing: http://aprt.us/
<paulproteus>
Wow neat.
<paulproteus>
...
<paulproteus>
I literally know these people.
<kentonv>
then they must know about sandstorm!
<paulproteus>
I hung out with Joshua Horowitz and Paula Te yesterday and saw the lunar eclipse.
<paulproteus>
I was going to tell Paula about Sandstorm but got distracted.
<paulproteus>
I'm sure Joshua knows about it.
mnutt_ has joined #sandstorm
<paulproteus>
ocdtrekkie: You should email Paula & Bret & cc: me!
<kentonv>
oh Vi Hart, she's Andy's friend (and internet-famous)
<ocdtrekkie>
But you already know them! :P
<paulproteus>
But they won't believe anyone but me cares about Sandstorm unless you do!
<ocdtrekkie>
Would it be easier to create a Github issue suggesting Sandstorm? They don't have emails posted.
<dwrensha>
it looks like your 'a lifetime isn't unifying with your 'r lifetime there
<maurer>
Hm, OK, I'll try to run around annotating lifetimes
<maurer>
dwrensha: Shouldn't 'a and 'r be independent since each is bound by for?
<maurer>
Switching to "pub fn convert_type(type_reader : <holmes::h_type::Owned as OwnedStruct>::Reader)
<maurer>
seepub fn convert_type(type_reader : <holmes::h_type::Owned as OwnedStruct>::Reader)
<maurer>
ms to work
<maurer>
(sorry for doublepaste)
<maurer>
(e.g. holmes::h_type::Reader does not work, <holmes::h_type::OwnedStruct>::Reader does)
<maurer>
The trouble I'm running into here is that if my capnprototype is T, I want struct_list::Reader<T::Owned> and F : Fn(T::Reader) -> U
<maurer>
and I want them to be the same, but you can't parameterize on modules
<maurer>
which means that I end up needing to specify that the type of the input of the function is the associated reader type of the owned type that is in reader
<dwrensha>
I think I need more context. You're writing a function that takes two parameters, one of type struct_list::Reader<T::Owned> and one of type F where F: Fn(T::Reader) -> U ?
<maurer>
dwrensha: It's a map+collect function, so it
<maurer>
*it's basically turning struct_list<T::Owned> into Vec<U> using T::Reader -> U
<maurer>
I appear to be able to get away with this if I declare all the parameter functions with a funny signature
<dwrensha>
ok. I think that gives me enough information to play with such an example on my end...
<maurer>
I think I can make it work with the current version, but it requires me to do <t::Owned as OwnedStruct>::Reader as the parameter to my parameter functions
<dwrensha>
ah, and you want to change `T: FromStructReader<'a>` to `T: capnp::traits::OwnedStruct<...>`
<maurer>
Basically, since the parameter to struct_list is no longer a reader but an owned
<maurer>
Using "<holmes::val::Owned as OwnedStruct>::Reader" where I have "holmes::val::Reader" seems to work, but I don't know yet since I need to change everything in the file
<maurer>
(though that's also _super_ verbose)
<dwrensha>
ok, I got it to compile with a `where F: Fn(<T as ::capnp::traits::OwnedStruct<'a>>::Reader) -> U`, which I guess is the verbosity you're pointing out
<dwrensha>
you need the "as" part in order to specify the constraint on the 'a lifetime
<dwrensha>
`<T as Owned<'a>>::Reader` is how we fake `&'a T`
<dwrensha>
creating synthetic versions of Rust's borrowed pointers (as we've done using associated types) does kind of make you appreciate the stuff that's built in to Rust's borrowed pointers
<dwrensha>
I imagine it could be nicer with higher-kinded types, but those are in the distant future.
<dwrensha>
maurer: I think the full verbosity is only needed in generic cases. If you're dealing with concrete types, you shouldn't need to worry about the fancy associated types stuff.
gopar has quit [Remote host closed the connection]
<maurer>
dwrensha: OK. Thanks for thinking about it
simonv3 has quit [Quit: Connection closed for inactivity]
jacksingleton has quit [Ping timeout: 264 seconds]
larjona has quit [Quit: Konversation terminated!]
jacksingleton has joined #sandstorm
fonfon has quit [Ping timeout: 265 seconds]
mnutt__ has quit [Quit: mnutt__]
jeffmendoza has quit [Ping timeout: 265 seconds]
jeffmendoza has joined #sandstorm
jeffmendoza has quit [Ping timeout: 265 seconds]
gopar has joined #sandstorm
Guest59200 has quit [Ping timeout: 246 seconds]
paulproteus has quit [Ping timeout: 246 seconds]
dvn has quit [Ping timeout: 246 seconds]
iangreenleaf has quit [Ping timeout: 246 seconds]
preilly has joined #sandstorm
iangreenleaf has joined #sandstorm
preilly is now known as Guest34201
paulproteus has joined #sandstorm
dvn has joined #sandstorm
gopar has quit [Remote host closed the connection]