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
<isd> Reminder: office hours in 1 hour.
<JacobWeisz[m]> Can't make it today, but you approved all of my accomplishments this week.
<JacobWeisz[m]> I have nothing else to contribute.
<abliss> is there a way to force sandstorm to 'empty the trash'? e.g. i seem to have several versions of the wekan app in /opt/sandstorm/var/sandstorm/apps, despite having uninstalled it and having no grains of it.
<abliss> (and each is 600MB+)
<JacobWeisz[m]> It's supposed to purge those at some point...
<isd> Do you have wekan grains in the grain trash? How long has it been?
wings has quit [Ping timeout: 256 seconds]
cstrahan has quit [Ping timeout: 256 seconds]
cstrahan has joined #sandstorm
Mitar has quit [Ping timeout: 264 seconds]
Mitar has joined #sandstorm
<JacobWeisz[m]> Ian, any idea where is the best place to both delete a directory in a app's repo and symlink to to /var? For upstreaming purposes, it should stay in Git but I definitely want the Sandstorm package to make it go to writable storage.
<JacobWeisz[m]> It'd have to be in setup.sh, wouldn't it?
<JacobWeisz[m]> I'm trying to package at ocdtrekkie/kanboard and the data directory needs to be writable. I don't think I need anything presently in it though.
<abliss> i'd do it in build.sh i think
<abliss> `rm -rf data/; ln -sf /var data`
<JacobWeisz[m]> I'll try it. Not that this is an important packaging project. Because I forgot which app I was supposed to play with packaging when I started packaging this one... But I might as well see if I can finish it...
<JacobWeisz[m]> So many kanban boards out there...
<JacobWeisz[m]> Gah, this is a "can't do on a Windows host" issue.
<JacobWeisz[m]> When you try to make a symbolic link in the folder shared with the host machine, you get a protocol error.
xet7 has quit [Quit: Leaving]
<isd> Yeah, last I checked windows filesystems didn't support symlinks? Not sure if that's still true.
<TimMc> There are NTFS junction... somethings.
<isd> ...we should change the static publishing failure page to something that explains the sandstorm host matching thing more clearly
<JacobWeisz[m]> Probably
<isd> I feel like 99% of the time when people see that it's because they're trying to do a basic sandstorm box setup and don't have something quite right, but the message is all about how to set up static publishing.
<JacobWeisz[m]> Agreed
<isd> So it's probably the wrong error message to be displaying.
<isd> It should also use sandstorm CSS to give the user a cue that they have actually gotten as far as connecting to sandstorm.
* isd starts working on a patch
<JacobWeisz[m]> I wanna express how weird it feels to me that the HTML in this PR is in a c++ file.
<isd> Yeah, but to do anything else it would need to figure out where to route the request. And it can't just be a static file because it includes some dynamic information.
<isd> And we don't have a templating language for C++ already.
<isd> But that was actually really irritating when testing this, since it meant I actually had to rebuild the full sandstorm bundle on each change.
<JacobWeisz[m]> It's not so much a complaint as an i,i comment.
<isd> I figured. I kinda felt the same way.
<JacobWeisz[m]> My only question based on reading is on the PR. A screenshot of the end result might be nice though.
<isd> Made the change and uploaded a screenshot.
frigginglorious has joined #sandstorm
frigginglorious1 has joined #sandstorm
frigginglorious has quit [Ping timeout: 256 seconds]
frigginglorious1 is now known as frigginglorious
<abliss> made some tweaks to my ttyd package and uploaded a new spk to github. new demo grain here: https://sandstorm.rapidraven.com/shared/fUMYLS4uozrabgQshl-gNRnaaCrHVBka1_y8BDTnR3v
<isd> Nice. A lot of stuff does seem to work now
<isd> History, scrolling less...
<isd> Job control would be great but this is not half bad as-is.
xet7 has joined #sandstorm
<isd> I wish our build system didn't rebuild meteor if I've only made C++ changes
<isd> (not including node-capnp)