<zarvox>
though I guess if you don't mind requiring a root install for spk dev to work today, then you have sufficient permission to set up more UIDs tomorrow
<kentonv>
zarvox: OTOH, if everything is running under vagrant-spk, we can play with UIDs all we want.
<zarvox>
Oh yeah. I worry slightly about how this would interact with vboxsf and p9fs and such, but I guess the fuse layer sits atop all that.
<paulproteus>
There should be a ShareLaWTeX: ftp://ftp.heanet.ie/mirrors/sourceforge/l/project/la/lawtex/lawtex-doc.pdf
<paulproteus>
btw < 3 ftp://
<kentonv>
ftp is the worst
<kentonv>
who designs a protocol where the server connects back to the client, seriously? :P
<jadewang>
hey paulproteus
<kentonv>
paulproteus: meetup.com just prompted me to "chip in" when I RSVP'd
* paulproteus
waves, heading to dinner, will read scrollback later!
gopar has quit [Remote host closed the connection]
gopar has joined #sandstorm
joshbuddy has quit [Quit: joshbuddy]
mnutt__ has joined #sandstorm
* paulproteus
waves to mnutt__
ragesoss has quit [Read error: Connection reset by peer]
ragesoss has joined #sandstorm
mort___ has quit [Quit: Leaving.]
gopar has quit [Quit: Leaving]
jadewang has quit [Remote host closed the connection]
jadewang has joined #sandstorm
ecloud has quit [Quit: No Ping reply in 180 seconds.]
ecloud has joined #sandstorm
jadewang has quit [Remote host closed the connection]
ocdtrekkie has quit [Quit: No Ping reply in 180 seconds.]
ocdtrekkie has joined #sandstorm
jadewang has joined #sandstorm
joshbuddy has joined #sandstorm
jadewang has quit [Remote host closed the connection]
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 256 seconds]
joshbuddy has quit [Quit: joshbuddy]
NwS has joined #sandstorm
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 264 seconds]
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 256 seconds]
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 244 seconds]
todayman_ has joined #sandstorm
todayman has quit [Ping timeout: 250 seconds]
todayman_ is now known as todayman
amyers has joined #sandstorm
mnutt_ has quit [Quit: mnutt_]
mnutt_ has joined #sandstorm
mnutt_ has quit [Quit: mnutt_]
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 256 seconds]
erikoeurch has joined #sandstorm
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 244 seconds]
amyers has quit [Quit: Leaving]
YuviPanda|zzz is now known as YuviPanda
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 248 seconds]
NwS has quit [*.net *.split]
NwS has joined #sandstorm
<paulproteus>
Mornin'
joshbuddy has joined #sandstorm
pouledodue has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
<erikoeurch>
Is it reasonably that a new, empty Paperwork grain uses 25 MB of memory when Gitlab (existing repo with a bunch of commits) uses 4 MB?
<paulproteus>
If the number is from the top-bar, then storage, not RAM.
<paulproteus>
But yeah -- Paperwork iirc uses MySQL so has a way bigger disk footprint.
<paulproteus>
But also 25 MB isn't super-duper much in 2015.
<paulproteus>
I think kentonv might disagree; good for me he's probably not awake.
<paulproteus>
In all seriousness, yeah, 25 MB sure is noticeably more than 4 MB.
<dwrensha>
Maybe someday we'll figure out how to configure/recompile mysql/mariadb to not use tens of megabytes for an empty database
<paulproteus>
MySQLite
<paulproteus>
(I had no point, I just wanted to say)
<erikoeurch>
paulproteus, ah, ok! Well, in any case it seems like quite a lot
<paulproteus>
Agreed.
<erikoeurch>
well, relatively quite a lot, haha
<paulproteus>
It's like 1/4 of a ZIP disk!
<erikoeurch>
I have a few notebook grains and a few other grains and I'm already consuming about 100 MB
<paulproteus>
I think to some extent this is going to be like early Android apps and power usage.
<paulproteus>
Over time, developers and the platform will figure out how to make things more efficient as power users see these stats and begin to care.
<erikoeurch>
or Sandstorm will simply increase quotas :)
<paulproteus>
(-:
<erikoeurch>
seeing as storage is fairly cheap, that will eventually happen, right?
<erikoeurch>
if everything else goes well
<paulproteus>
Presumably.
<paulproteus>
[this is not a promise of future service! or pricing! etc.!]
<paulproteus>
Other hosting providers sure have offered more disk per dollaras time has gone on, anyway.
<erikoeurch>
I realized I shouldn't try to get you to speculate about that :)
<paulproteus>
(-:
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 264 seconds]
jleo has quit [Ping timeout: 250 seconds]
jleo has joined #sandstorm
erikoeurch has quit [Ping timeout: 264 seconds]
jadewang has joined #sandstorm
jadewang has quit [Remote host closed the connection]
<maurer>
dwrensha: Is there a reason that ::std::any::Any is used for the return type of get_descriptor for your ClientHooks? I'm trying to write similar code for the OCaml implementation nad it's unclear to me why those aren't just returning OwnedCapDescriptors
<maurer>
The only time you seem to case out on them being something else is to automatically create a new export from an unexported ClientHook, but ist seems like an Option<OwnedCapDescriptor> encapsulates that
<maurer>
*it
erikoeurch has joined #sandstorm
<dwrensha>
maurer: OwnedCapDescriptor is defined in the rpc crate
<dwrensha>
hm... though it seems like that definition could be moved to the base crate
<maurer>
dwrensha: OK. In that case I can probably get away with a "cap_descriptor option" in my implementation
<maurer>
because everything's going into one ocaml library
jadewang has joined #sandstorm
<maurer>
(This isn't necessarily a problem with capnp-rpc-rust, I'm just trying to make sure I have a full understanding since so far I count 5 different representations of capabilities in the various implementations, so it was totally possible I was missing something important)
<zarvox>
Hmm, I think I should add another X-Sandstorm-Passthrough value: user-agent. Piwik appears to like it for devices detection.
jadewang has quit []
pouledodue has joined #sandstorm
jeffmendoza has quit [Remote host closed the connection]
jeffmendoza has joined #sandstorm
<zarvox>
Looks like I can pass it in-band in the track content instead, thanks navigator.userAgent
jadewang has joined #sandstorm
joshbuddy has joined #sandstorm
<paulproteus>
++ re: in-band
<paulproteus>
This is an amusing micro case study in user tracking.
<zarvox>
What does it let us conclude, if anything?
<paulproteus>
I guess that most of the anti-privacy "attack surface" is available in JS and doesn't need HTTP header help.
<paulproteus>
Shrug, though.
<paulproteus>
Probably actually it lets us conclude nothing.
<paulproteus>
BTW zarvox there's apparently license unclarity around redistributing Ubuntu (!?) so good thing you switched vagrant-spk to be based on Debian.
* zarvox
takes a bow
<zarvox>
perhaps amusingly, I probably wouldn't have done it if the LTS Ubuntu release had included nsenter
<paulproteus>
kentonv + zarvox: I'm thinking about "Security Practices Overview" and "Pull-Request---Code-Review-Procedure" - they arguably don't belong in docs/developing/* since all the stuff there is for third-party app developers.
<paulproteus>
So I could make a new docs section called "Sandstorm itself" with those two documents. But then I'd be tempted to move "App Developer Handbook" into "Sandstorm itself" and now I'm in a pile of doubt.
<paulproteus>
So in the absence of input, I think I'll "just" put those two under "Developing" and act as though nothing's wrong.
<paulproteus>
So here what I'm saying is that if you have input, I'd welcome it. I'm also OK with just doing something and letting people provide feedback and will lean toward that.
jadewang has quit [Remote host closed the connection]
erikoeurch has quit [Quit: Leaving]
<XgF>
paulproteus: I think "developing" should mean "developing sandstorm" and app development should be under "app_development"
<paulproteus>
This is w/r/t folder structure?
<XgF>
yes
<paulproteus>
Unfortunately oops I already sort of picked one. I guess I can change it, but I am shy about breaking any existing links since I have no plan for redirects.
<paulproteus>
I did adjust the left nav to say "Developing apps" though.
<kentonv>
"security practices overview" could arguably go under "using", or otherwise some sort of "sandstorm internals" section
<kentonv>
that is to say, I think a lot of users want to understand Sandstorm's security model
<paulproteus>
I'm +0 on developing/ == apps, and internals/ == things related to Sandstorm itself
<kentonv>
I think "code review procedure" can stay on the wiki. It specifically relates to using the git repo, not Sandstorm.
<kentonv>
Arguably "technical overview" could also be under "using", though honestly I'm not sure I like having that sort of ad-hoc bullet list anywhere in the docs; it should be replaced with something better
<paulproteus>
I agree re: that sort of ad-hoc bullet list [doesn't belong] in the docs
<paulproteus>
I only imported it for completeness.
<paulproteus>
Can only rewrite so many things at once. (-:
<kentonv>
what do you think about putting "how it works"-type things under "using" rather than "developing apps"?
<paulproteus>
I'm at least +0 on the idea.
<paulproteus>
It's at least not actionable information.
<kentonv>
it would also make the "developing apps" section look less intimidating. :)
<paulproteus>
That is in fact a goal of mine, too. (-:
<paulproteus>
Tragically I use the word "Overview" to name the index of each section.
<kentonv>
another thing I think we should do there is group together "API features" (web publishing, SMTP, etc. -- things not every app uses) separately from essentials
<zarvox>
I'm +1 on pull-request/review hygiene living on Github for now.
<paulproteus>
I guess "How it works" can => using/ and keep the "How it works" name.
<kentonv>
Yeah, I would actually title it "how it works" rather than "technical overview". And then over time we should make the page better.
<kentonv>
(some of it is pretty outdated, even...)
<kentonv>
and "security practices overview" could maybe become "Why it's secure" or "How it's secure" or something
<paulproteus>
re: outdated: Yeah kentonv I figure I can entice you to fix it if it's in the repo (-:
<paulproteus>
And if there's like prev/next buttons.
<zarvox>
as a lazyuser, the formatting on https://docs.sandstorm.io/en/latest/guided-tour/ makes it tricky for me to parse that I should skip the long Install Sandstorm section that covers a screen height and a half and go straight to the "Hands-on tour" section - can the <h2>s get more margin-top?
<ocdtrekkie>
I think marketplaces or bazaars are more likely in deserts at oasises than stores.
<paulproteus>
zarvox: secretly I think it's fine as a lazy user to skip the "Install" section!
<paulproteus>
I think what I really want from that page is a page that focuses on the demo, but is usable at a meetup when I tell people who are new to Sandstorm but are mega Linux nerds [used endearingly; self included] who want to run it themselves.
<zarvox>
I think if I'm a lazyuser and I've already clicked three times to get to that page I'm not interested in reading two pages of prose; I want to get to the numbered list ;)
<paulproteus>
I do see what you might mean, looking at it in my browser.
<paulproteus>
Also rofl literally _instance_ since apparently I don't know Markdown.
<paulproteus>
Asheesh's Dark Secret: Really I might never learn Markdown.
<kentonv>
paulproteus: most markdown implementations accept _, but I think there's a space missing between "_instance_" and "of". :)
<paulproteus>
Anyway brb moving "Technical overview" and renaming it to "How It Works".
<zarvox>
I might attempt to find a more easily browseable theme at some point
<paulproteus>
+1
<kentonv>
eventually Nena should probably make a theme. :)
<paulproteus>
Whenever you want to talk themes zarvox I can talk more about it.
<paulproteus>
Trying to focus on themes as little as possible until someone cares a lot or finds a really good reason to care, since there is so much else to care about.
<zarvox>
agreed
<paulproteus>
(but if there's a good reason to care about it, then I'm OK spending some care on it. I switched to this theme in particular for some reason that I've already forgotten but can go figure out when we revisit themes)
<zarvox>
Oh, I think part of my twitchiness is from the sidebar having everything open at all times, rather than collapsed
<kentonv>
paulproteus: I think you switched randomly when I vaguely didn't like the previous theme. :)
<paulproteus>
zarvox: just shrink your screen width (-;
<paulproteus>
bd file bug against mkdocs' readthedocs theme
<paulproteus>
semi sorry to be terse; back in a bit!
dlitz has joined #sandstorm
<zarvox>
paulproteus: I don't want to file a bug against someone else yet since I'm not sure if it's a problem in mkdocs or something out of date or if there's a different way we should be organizing our docs to get that behavior instead
dlitz has quit [Read error: Connection reset by peer]
<paulproteus>
I think it is though a mkdocs rtd theme bug. I haven't double-verified that so feel free to file a Sandstorm bug requesting me to do so.
dlitz has joined #sandstorm
<paulproteus>
Seems very meta but is pretty reasonable IMHO.
<paulproteus>
The main result is I'll Tweet at the nice mkdocs person and he'll be like brb fixed it for you
NwS has quit [Ping timeout: 246 seconds]
dlitz has quit [Read error: Connection reset by peer]
dlitz has joined #sandstorm
dlitz has quit [Read error: Connection reset by peer]
dlitz has joined #sandstorm
decipherstatic has joined #sandstorm
<paulproteus>
Arguably "Using" is now "Using & understanding"
dlitz has quit [Ping timeout: 248 seconds]
dlitz has joined #sandstorm
dlitz has quit [Read error: Connection reset by peer]
<paulproteus>
I wrote that when I was using the old theme!
<zarvox>
I should maybe save my feedback on things until you've been through them once.
<paulproteus>
Oh actually I remember now.
<paulproteus>
The reason I switched theme was, on a non-very-wide screen, the nav goes to the top of the screen in the old theme, and I found that personally rather annoying.
<paulproteus>
I'm trying to go through things about 0.8 times as part of the mass migration.
<paulproteus>
I edit each document in accordance with how much I like the doc and how patient I feel at that moment.
<kentonv>
feel free to make a list of things that I should review/edit
<paulproteus>
++
jadewang has joined #sandstorm
jadewang has quit [Remote host closed the connection]