rustyrazorblade has quit [Ping timeout: 272 seconds]
jacksingleton has joined #sandstorm
eternaleye has quit [Quit: Konversation terminated!]
gemlog has joined #sandstorm
<gemlog>
@paulproteus oauth just worked behind nat after the update!!! thx :-)
<gemlog>
google oauth
<paulproteus>
gemlog: Yay!
<gemlog>
:-)
<paulproteus>
I believe it was zarvox's code that put it all together.
<gemlog>
Then thank you very much to zarvox too!
<gemlog>
I have been able to play anyhow. I enabled passwordless logins and invited myself from the emergency admin panel. But no I've finally gained admin.
<gemlog>
wno
<gemlog>
now
<gemlog>
I'm too excited to type properly. Sorry. ;-)
<gemlog>
k. Just wanted to let you guys know it worked for me and my nat instance. So now both work, PLUS we now have https. So cool.
<gemlog>
g'nite
gemlog has left #sandstorm ["Konversation terminated!"]
<paulproteus>
/usr/bin/yes
<zarvox>
paulproteus: gemlog: I don't think my patchset got merged yet, so I'm not sure how I could have done that. :P
<paulproteus>
Oh! Huh.
<paulproteus>
Maybe the update was the restart, and that's that.
<paulproteus>
i,i just do an update every day
qube23 has joined #sandstorm
<qube23>
hi!
<paulproteus>
Hi qube23 1
<paulproteus>
!
<qube23>
I'm probably just using it wrong, but still: is it possible to lose write access to my own grain? I have a WeKan board here that does not let me edit it anymore. oO
<paulproteus>
qube23: If you're the grain's owner, you do have all permissions.
<paulproteus>
So it might be a WeKan bug.
<paulproteus>
If it is, then maybe we can get your help filing the bug.
<paulproteus>
You can tell if it's your grain based on if you see the "Download backup" button.
<paulproteus>
If you do, it's yours.
<qube23>
Button is there, so it is my grain.
* paulproteus
frowns.
<paulproteus>
That's sad then!
<paulproteus>
Sounds like an app bug.
<qube23>
Better than a Sandstorm bug, isn't it? ;)
<paulproteus>
Maybe you can share write access to it with mquandalle and he can figure out what's going on.
<paulproteus>
I'm introducing Sandstorm bugs right now, so I definitely feel that. : P
<qube23>
The data is sensitive, so I can't really share access to that grain. I'll try to reproduce the issue with another one though.
<paulproteus>
b
<paulproteus>
b == thumbs up
<paulproteus>
Feel free to try to repro on oasis.sandstorm.io
<paulproteus>
I also wonder - if you download a backup & create a new grain, does the bug still affect you?
<qube23>
Yes, still there when I restore from the zip :-)
<paulproteus>
Sweet (-:
heliostatic has joined #sandstorm
heliostatic has quit [Quit: Be back later ...]
indiebio has quit [Ping timeout: 246 seconds]
indiebio has joined #sandstorm
<ocdtrekkie>
I am having a bad time getting the ownCloud Client compiled for testing.
jacksingleton has quit [Ping timeout: 252 seconds]
heliostatic has joined #sandstorm
<paulproteus>
ocdtrekkie: Aw.
<ocdtrekkie>
cmake is hard.
heliostatic has quit [Ping timeout: 264 seconds]
<ocdtrekkie>
Could NOT find SQLite3 (missing: SQLITE3_LIBRARIES SQLITE3_INCLUDE_DIRS)
<ocdtrekkie>
(Required is at least version "3.8.0")
<paulproteus>
I mean, then try the things that you're already trying.
<ocdtrekkie>
I did that first.
<paulproteus>
o
<ocdtrekkie>
It installed a lot of things.
<ocdtrekkie>
But apparently not enough things.
<paulproteus>
... you are tempting me to help you compile things.
<paulproteus>
I just love compiling things.
<ocdtrekkie>
lol
<ocdtrekkie>
I thought of something, lemme try it.
<paulproteus>
I'm going to go home first (at a burger place, apparently, at the moment) so back online in ~30 min.
<paulproteus>
So if you need help after that I'm on board for a little bit, at least.
<ocdtrekkie>
What I thought didn't help.
<ocdtrekkie>
I should be around quite a while yet.
<ocdtrekkie>
Enjoy the trip home.
<paulproteus>
b
<paulproteus>
sandcats alert; looking into it
<paulproteus>
fixed automatically apparently
groxx has quit [Quit: Connection closed for inactivity]
<kentonv>
yes, fixed by automatic kenton activation
<kentonv>
unfortunately, we are apparently no closer to figuring out why the event loop stalls
bb010g has joined #sandstorm
<kentonv>
on the bright side it's kind of amazing to get an update notification for like 25 apps at once, click "apply" while on the app grid, and watch all the icons change.
<zarvox>
and if it still complains, see which version you have installed with "apt-cache policy libsqlite3-dev"
<zarvox>
On Linux distros, a frequent pattern is to separate libraries into packages with runtime binaries (on Debian, the package is usually named libwhatever) and packages with compilation headers/include paths (on Debian, usually libwhatever-dev)
<indiebio>
hi larjona, so on storm, if I want to create the titanpad equivalent... I press new ... and then what?
* indiebio
goes and reads the email first ...
<zarvox>
the big plus button, install Etherpad, and then Etherpad will appear in your app list, and you can make a new instance
<indiebio>
ah thanks. so storm takes existing things and hosts them?
<zarvox>
Sandstorm is a platform for running applications, with an emphasis on ease-of-use and security-by-default.
<indiebio>
\o/ thanks zarvox
<zarvox>
:)
<zarvox>
glad to help! feel free to ask more questions, and I'll answer them either now, or in the morning when I read the scrollback
<indiebio>
so far so good, very user-friendly
<indiebio>
what is dogfooding?
<indiebio>
I get the gist, but why is it called that?
jacksingleton has quit [Ping timeout: 265 seconds]
<zarvox>
I think it's about subjecting yourself to the quality of your product. The term might have originated from Kal Kan Pet Food, where the president was said to eat a can of dog food at shareholders' meetings.
<larjona>
indiebio etherpad is already installed; only need new -> etherpad document
<zarvox>
larjona: apps are installed (or at least, displayed) per-user, not globally on the instance
<indiebio>
I see :)
<indiebio>
link shortener ... deb.li says it's there. do I need to install it and run from terminal, or is there a place to put my long URL in on the site? (Also, do I ask this in this channel, in #debian-welcome or ...? does it matter?
<zarvox>
under the hood, Sandstorm will deduplicate if multiple people have the same app package installed, but otherwise unrelated users could spam each other's UI, which is undesirable
<zarvox>
In team settings, people will probably want to be able to "preinstall" a group of apps, but we've not built that sort of thing yet
drako22 has joined #sandstorm
<drako22>
Good morning. I have a quick question about capnp; after a List in a capnp message has been initialized with a fixed size, can the list be extended? (append)
<drako22>
I am unable to find something like append() in the template class in any.h
<zarvox>
No, lists must be fixed size due to arena allocation. Consider using a std::vector to build up your list, and then once it's complete, copy it into the capnp List?
bb010g has quit [Quit: Connection closed for inactivity]
<drako22>
zarvox; another stupid question. How do I allocate new objects to put into the list (::Builder, I guess)?
<drako22>
With the MallocMessageBuilder?
<drako22>
I guess it has to do with the Orphan class?
<zarvox>
MallocMessageBuilder and orphans will work. I've honestly not used the C++ capnproto interface directly much
<zarvox>
If you know the size up front, you can just builder.initThatField(n)
<drako22>
right, that is the problem. I don;'t know yet.
<drako22>
Found out
<drako22>
getOrphanage()
<drako22>
build them there, then adopt them
<zarvox>
great! sorry I'm not more help; I've mostly used the node-capnp bindings, so I don't know all the C++ implementation details offhand
<zarvox>
It's late here; I'm going to head to bed. I (and others!) will read the backlog when we return, so feel free to leave questions here, or via email :)
paulproteus has quit [Read error: Connection reset by peer]
paulproteus has joined #sandstorm
isd has quit [Quit: Leaving.]
drako22 has joined #sandstorm
<drako22>
Good night guys.
<drako22>
I will repeat the problem; I have to build a dynamic size list of elements T, in which its size I cannot deduce it before traversing the data structure (and traversing it twice does not work either). I was trying to use a vector of Orphans, but apparently they get 0ed once the auto references go out from stack
<drako22>
I had a std::vector<Orphan<T>>, and I was pushing back with std::move/kj::mv.
<drako22>
Any idea on how to solve this issue?
<paulproteus>
Hey drako22
<paulproteus>
I don't personally know but I know kentonv is out in a meeting so hopefully he'll have some advice for you soon. If not, consider sending a mail to the Cap'n Proto email list.
<drako22>
ok, thanks
<drako22>
do you know if he will be back in say 2h ?
acous has quit [Changing host]
acous has joined #sandstorm
<neynah>
It's kind of strange that draw.io gives me an option to name my file when I create a new diagram-yet this isn't reflected in the grain title. :((((
<paulproteus>
File a hug report
<neynah>
*hug
<XgF>
paulproteus: Can a grain even update its' own title?
<kentonv>
drako22: vector<Orphan<T>> works, but you do eventually have to "adopt" each orphan into the message tree later on. So once you know the overall size of the list, allocate the list (getting a list builder) and then list.adopt(i, kj::mv(vec[i])) for each element.