jadewang has joined #sandstorm
natea has quit [Quit: natea]
isd has joined #sandstorm
jadewang has quit [Remote host closed the connection]
paroneayea has quit [Read error: Connection reset by peer]
paroneayea has joined #sandstorm
isd has quit [Quit: Leaving.]
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 265 seconds]
natea has joined #sandstorm
gopar has quit [Remote host closed the connection]
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 258 seconds]
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 250 seconds]
mquandalle has quit [Quit: Connection closed for inactivity]
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 252 seconds]
gopar has joined #sandstorm
soulshake has left #sandstorm [#sandstorm]
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 272 seconds]
gopar has quit [Quit: Leaving]
jadewang has joined #sandstorm
bb010g has quit [Quit: Connection closed for inactivity]
jadewang has quit [Ping timeout: 256 seconds]
kxra has quit [Ping timeout: 272 seconds]
kxra has joined #sandstorm
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 272 seconds]
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 256 seconds]
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 264 seconds]
paroneayea has quit [Read error: Connection reset by peer]
paroneayea has joined #sandstorm
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 265 seconds]
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 255 seconds]
decipherstatic has quit [Remote host closed the connection]
mquandalle has joined #sandstorm
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 244 seconds]
natea has quit [Quit: natea]
jadewang has joined #sandstorm
natea has joined #sandstorm
jadewang has quit [Ping timeout: 265 seconds]
paroneayea has quit [Read error: Connection reset by peer]
paroneayea has joined #sandstorm
jadewang has joined #sandstorm
natea has quit [Quit: natea]
natea has joined #sandstorm
jadewang has quit [Ping timeout: 246 seconds]
natea has quit [Quit: natea]
natea has joined #sandstorm
[d__d] has quit [Remote host closed the connection]
[d__d] has joined #sandstorm
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 264 seconds]
natea has quit [Quit: natea]
natea has joined #sandstorm
natea has quit [Ping timeout: 246 seconds]
natea has joined #sandstorm
natea has quit [Quit: natea]
gopar has joined #sandstorm
natea has joined #sandstorm
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 265 seconds]
natea has quit [Quit: natea]
<paulproteus> Wahoo! Heading back to the office. Or, "office", as Kenton says.
<mcpherrin> is the "office" somebody's living room?
<paulproteus> XgF: Yes please do make "GitWeb Pages" for Sandstorm.
<paulproteus> I would migrate a bunch of sites to Sandstorm if that existed; not even kidding.
<paulproteus> Also I think it'd be hilarious if you actually call it "GitWeb Pages".
<ocdtrekkie> kentonv: I'm trying to use a newer meteor-spk, and I'm having an issue. I had to install "scons", and scons says there is no SConstruct file found when I try to make meteor-spk.
<paulproteus> Maybe someone can make a "Auto-sync my personal server with GitHub" Sandstorm app sometime when there are scheduled tasks and/or when Sandstorm is compatible with GitHub Web Hooks, and then people can use Sandstorm for static hosting of sites via git repositories mirrored all over without using the actual GitHub Pages functionality.
<paulproteus> i,i ocdtrekkie bitrot is a powerful force
<paulproteus> Clearly meteor-spk should grow a test suite.
<ocdtrekkie> Someone did Yet Another Slack Clone (this should be a term. YASC) in Meteor.
<ocdtrekkie> And I figured I'd try to run it in Sandstorm for the fun of it, because meteor-spk is really easy. But less so when meteor-spk does not work.
<paulproteus> ocdtrekkie: I'm surprised that you'd need Scons.
<paulproteus> I might check this out right now.
<ocdtrekkie> The meteor-spk makefile uses it, I guess.
<ocdtrekkie> I was going to use the older meteor-spk I have on my machine, but that did not work with the app. Interestingly enough, the app works when I run "meteor", but it gets a runtime exception under Sandstorm.
<ocdtrekkie> Oh, scons is because I tried building it from source for some reason.
<ocdtrekkie> I am going to try the binary distribution link like I did before.
<paulproteus> Presumably we should make the README clearer so you don't make that mistake in the future.
<ocdtrekkie> Eh, the readme is somewhat clear.
<ocdtrekkie> I just read the wrong section.
<ocdtrekkie> (That being said, doing the from source thing didn't work for me, so maybe there's missing instructions there.)
jadewang has joined #sandstorm
<kentonv> ocdtrekkie: is it possible that when you checked out the meteor-spk git repo you did not use --recursive?
<kentonv> that would cause mongo to not be checked out, and mongo is the part that is built with scons
<ocdtrekkie> Ah, that is likely.
<zarvox> jparyani: any idea what's up with jenkins? the last three pretty-trivial-looking pull request tests builds all failed.
<ocdtrekkie> So paulproteus, my meteor-spk issues were PEBKAC issues.
<ocdtrekkie> :)
<dwrensha> zarvox: looks related to the meteor upgrade
<zarvox> ahhh, do we not fully clean the repo on each rebuild? I guess that would make meteor stuff slow.
<zarvox> I should head to the "office" myself!
<dwrensha> the tests start from a clean slate, but it looks like we're invoking `meteor build` in an obsolete way
<kentonv> it's weird that the main sandstorm build is green but the pr builds fail?
<kentonv> lol, I bet there's a space in the directory name
<kentonv> "Sandstorm PRs"
<kentonv> yep
<dwrensha> what
<kentonv> ok, my fault
<kentonv> I tweaked the makefile to avoid using ".." from the "shell" directory so that I could have that directory be a symlink
<kentonv> so it now records the original directory in a variable, then uses "$OLD/shell-build" instead of "../shell-build"
<kentonv> but I forgot to quote it
<kentonv> so if $OLD has a space, it doesn't work
<dwrensha> :\
<kentonv> testing fix..
<paulproteus> Yeah kentonv I figured out that ocdtrekkie didn't git clone --recursive already, and was working on https://github.com/sandstorm-io/meteor-spk/pull/5 rather than discussing it. I figure if an error message can be improved with minimal code cost, it's a good idea to do so.
<paulproteus> So anyway please enjoy that pull request! (-:
<dwrensha> kentonv: so why didn't Jenkins complete shortly after your commits?
<dwrensha> *complain
<kentonv> dwrensha: because it only has a space in the name when testing PRs
<paulproteus> I'm hopeful someone will use this lack of shell quoting in the Makefile to cause arbitrary code execution based on directory name.
<dwrensha> oh right
<paulproteus> I guess that's hard since the prefix is chosen by us, in this case.
<paulproteus> Oh well. Next time!
<kentonv> also, paulproteus, if we're testing their PR, they can already just write code in the PR.
<paulproteus> I mean, yeah, that's the other problem with my plan. (-:
<paulproteus> I realize. (-:
<ocdtrekkie> \o/ paulproteus!
itscassa|away is now known as itscassa
gopar has quit [Quit: Leaving]
natea has joined #sandstorm
<ocdtrekkie> Okay, now that that port is properly done, next project is EtherDraw.
paroneayea has quit [Read error: Connection reset by peer]
paroneayea has joined #sandstorm
<dwrensha> ocdtrekkie: how do I make new cards?
<dwrensha> ah, found it
<ocdtrekkie> I'm not thrillingly excited with Scrumblr's UI. But it basically works, and someone wanted the port, and now it's at least decently supporting the applicable Sandstorm feature.
<ocdtrekkie> Re: the pen moving. I hadn't noticed that before.
<ocdtrekkie> I'm sure the upstream dev felt it was a worthwhile endeavor.
gopar has joined #sandstorm
bb010g has joined #sandstorm
<ocdtrekkie> I actually know nothing of the merits of emacs or vi, but sometimes I can't help myself.
<ocdtrekkie> gedit does it, this "make a backup file" thing seems to be a trend on Linux. Do you guys find it useful? I tend to save a lot, so it rarely has the revision I need when I royally screwed up somewhere anyways.
<dwrensha> Now I'm confused. I don't like the backup feature so I have (setq backup-inhibited t) in my ~/.emacs
<dwrensha> But somehow that doesn't stop these backups from getting created?
<kentonv> yeah I always turn off the "make backups" option on editors.
<XgF> Doesn't Vim have the even crazier (but more useful) swap files?
<dwrensha> apparently it's technically the "autosave" feature that's biting me here.
<kentonv> dwrensha: I think I missed a message somewhere. What problem are you having?
<dwrensha> and the fact that I think I've done all I can do to turn off autosave and backups
<dwrensha> and I still see ".#<filename>" files
itscassa is now known as itscassa|away
itscassa|away is now known as itscassa
natea has quit [Quit: natea]
<dwrensha> oooh those files are actually locks
<dwrensha> they are symlinks that uniquely identify the emacs instance that holds the dirty copy of the buffer
<ocdtrekkie> Ah!
<ocdtrekkie> That's what I assumed the "backup" files were when I first noticed them, until I realized they didn't disappear when I closed the program.
gopar has quit [Quit: Leaving]
XgF has quit [Remote host closed the connection]
XgF has joined #sandstorm
<kentonv> ... but the kernel already implements locking as a kernel feature.
<kentonv> in fact it implements at least three different APIs to two underlying kinds of locks.
<XgF> kentonv: And all of them are broken
<XgF> Like fcntl locks unlock if you close() any FD pointing to the file, even if it's a completely independent FD from you open()ing the same file twice, because reasons
<kentonv> fcntl locks are definitely stupid
<kentonv> flock seems ok
<XgF> Until it encounters NFS, when it does weird things...
<kentonv> oh right, nfs
<kentonv> sigh
<XgF> But in this case I suspect the "lock file" is probably a symlink to the Emacs control pipe or something
<dwrensha> XgF: it links to "david@dwrensha.ws.26053:1429367181". user, host, pid, timestamp
<XgF> And what type of file is that it links to?
<XgF> Or is that the contents of the file?
<dwrensha> that's the name of the file
<kentonv> i.e. it's a symlink that doesn't point to a real inode at all?
<dwrensha> right
<kentonv> symlinks: the best way to store up to 40 bytes directly in an inode, avoiding allocating any blocks.
<kentonv> you can also use xattrs which on ext4 will get you some 100 bytes but it's FS-dependent.
<XgF> On ZFS xattrs involve creating a whole new directory
<XgF> xattrs space is literally a whole new directory structure associated with every file
<XgF> (and yes, 1 goto 1, you can have xattrs on files in xattrspace)
<kentonv> ewwww
<XgF> Not sure how you get into there on Linux or FreeBSD. On Solaris its' a flag to openat
<paulproteus> XgF: Hey when are you going to make GitWeb Pages? : D
<paulproteus> I might sound pleading, but anyway I do actually want to know if you're interested because if so I very much want to test it out after you do make it.
<paulproteus> And then I want to use it forever and it'll make my life so great.
* XgF grabs laptop since desktop isn't yet ready to do Sandstorm work on
<paulproteus> XgF: Er I didn't mean you shoul do it Right Now! (Though if you want to, OK!)
gopar has joined #sandstorm
<XgF> paulproteus: My one question, is there a way to get the static hosting URL without me having to do RPC?
<kentonv> nope
<kentonv> but you can probably reuse whatever hack dwrensha did in wordpress
<kentonv> probably involving a small program that he execs?
<zarvox> doesn't that one wget --mirror the whole site or something?
<paulproteus> +1 and we should document those little things, IMHO
<XgF> Aah yeah, he has a little C++ program
<paulproteus> zarvox: There's also the part where the WordPress UI shows a URL
<paulproteus> (Leaving aside what contents live at that URL)
<kentonv> I suppose we should have a general "sandstorm API tool" program that wraps all the API calls
<kentonv> for people who don't have Cap'n Proto in their chosen language yet
<paulproteus> I'd be pretty +1 on that.
<XgF> Now, to figure out how I get from an X-Sandstorm-Session-Id header to something useful :-)
<XgF> (in this context)
<paulproteus> But I see it in src/sandstorm/sandstorm-http-bridge.c++ so OK!
<kentonv> session-id is needed when you want to make Cap'n Proto calls, in order to associate them with the particular user who made the request.
<kentonv> more specifically it's needed in order to ask sandstorm-http-bridge to provide the SessionContext capability for that user
<kentonv> so probably dwrensha's little app takes it as a parameter?
<XgF> Yeah
<XgF> I guess fcgiwrap should turn X-Session-ID into a HTTP_X_SESSION_ID env var
bb010g has quit [Quit: Connection closed for inactivity]
mort___ has joined #sandstorm
<XgF> I'm doing set PATH=$PATH:/usr/sbin in a shell script, so obviously I can't decide what operating system this is
<paulproteus> : D
<XgF> Evidently dwrensha used a non-Ubuntu distro for gitweb
<zarvox> kentonv posted this as a comment on a pull request, but I found it interesting enough that I want to repost it here: http://www.skyhunter.com/marcs/petnames/IntroPetNames.html
<zarvox> Oh yeah, maybe I should figure out what it'd take to have vagrant-spk use a Debian Jessie image.
<zarvox> Debian is the mother distribution. The Universal OS.
<paulproteus> Plus then zarvox you can give a talk about vagrant-spk at Debconf.
<paulproteus> Although I guess you could either way, really.
<zarvox> Yay! Except then I can't go to CCCamp.
itscassa is now known as itscassa|away
<XgF> "remote: HTTP_X_SANDSTORM_SESSION_ID=1" well thats a surprisingl ybroing number
<paulproteus> XgF: Literally LOL-ing. (-:
<XgF> Currently in the "I need to build capnproto damnit" yak shaving phase
<XgF> Maybe this should just be an update for the existing gitweb package - publish your "public_html" branch or something
<paulproteus> FWIW I would rather it "just" publish the branch called master.
<paulproteus> I say "just" in scare quotes because it's not super reasonable to say anything is "just" anything else.
<XgF> git push origin master:public_html
<paulproteus> That's more work than I was hoping for, but OK I suppose!
<paulproteus> I should clarify and say it's totally OK.
<XgF> Hmm, "remote: kj/common.c++:44: failed: Supposedly-unreachable branch executed." Useful error message
<XgF> OK, for no reason I understand there was a KJ_UNREACHABLE() at the end of getClientId.c++, and I have no idea why this has never blown up before?
paroneayea has quit [Read error: Connection reset by peer]