01:10 <simonft> paulproteus: maybe the links after "Install apps to create" on the homepage should go to the app listing on https://sandstorm.io/apps/ and not the homepage of the app. Especially when the app page is just something like http://libreboard.com/

01:10 <simonft> although I'm not sure if it's possible to link to a specific app on the app listing page.

01:20 <paulproteus> Yeah, I wonder about that. I vaguely like the idea of giving the apps google-juice, but yeah linking to the app-list listing is probably a better idea.

01:21 <paulproteus> I also think you can't at the moment link to them.

01:21 <paulproteus> But that's fixable.

01:23 <simonft> paulproteus: I would even give each of the apps their own page rather than just a list on one page, but I'm assuming that's planned for the future since you'll need it when the number of apps expands.

01:25 <paulproteus> Yeah, indeed.

01:26 <paulproteus> BTW someone needs to give riseup.net a Sandstorm install.

01:26 <paulproteus> IMHO.

01:26 <simonft> paulproteus: why/what for?

01:26 <paulproteus> Then all @riseup.net people could have the ability to create riseup.net Etherpads, etc.

01:27 <paulproteus> And install whatever apps they want on that server, etc.

01:27 <simonft> paulproteus: like https://pad.riseup.net/ ?

01:27 <paulproteus> And they could make personal TODO lists online with Meteor TODOs, which they could share with others if they want.

01:28 <paulproteus> Yeah, like that, but where the cost per server app made available to them is zero, rather than a high-ish sysadmin cost for each app that the nice riseup people install.

01:28 <simonft> paulproteus: From talking to riseup.net admins, I think their biggest issue with running services is not so much the cost of sysadmin time to run the service but instead the time to deal with abuse of the service.

01:28 <simonft> I'm curious if sandstorm would help with that.

01:29 <paulproteus> If it's locked so it only @riseup.net people can create app instances, then it could help, yeah.

01:29 <paulproteus> Centralized auth is built-in with Sandstorm, so that falls out naturally from the platform.

01:30 <simonft> their server offering shell accounts was also locked to @riseup.net accounts, but they still had tons of issues.

01:30 <simonft> though shell accounts are a lot harder to secure.

01:30 <paulproteus> Sandstorm apps are also sandboxed, which presumably also massively helps.

01:30 <paulproteus> also brb getting off a BART train

01:31 <simonft> paulproteus: this actually raises an interesting question of what logging sandstorm does by default.

02:21 <paulproteus> Fascinating, simonft. The answer for now is that it does a bunch of logging and isn't super careful-or-not about having a log retention policy, but that can likely be easily changed.

02:21 <paulproteus> I will start an email thread.

16:52 <paulproteus> phildini: /win 8

16:52 <paulproteus> ... oops.

16:52 <paulproteus> well phildini now you know i'm thinking fondly of you

17:48 <paulproteus> Oops I was prepping for the wrong talk.

18:41 <sunu> paulproteus: Hi! I was just trying out sandstorm. Wondering whether there a way to set a permanent url to an app instance?

18:42 <sunu> Also the packaging tutorial in the README is a 404.

20:29 <phildini> hello lovely people. portland also says hello.

21:17 <zarvox> https://en.wikipedia.org/wiki/SQL-Ledger seems like it could be an interesting app to package for Sandstorm.

21:19 <maurer_> zarvox: http://hledger.org/ might be easier, and is still open source, unlike what you linked

21:20 <zarvox> oooh, that's even better

21:21 <zarvox> and we can probably get a much smaller .spk using Haskell than perl :D

22:43 <paulproteus> sunu: Oh no! re: packaging 404 readme

22:43 <paulproteus> Also hi sunu ! (-:

22:44 <paulproteus> sunu: https://docs.sandstorm.io/en/latest/vagrant-spk/packaging-tutorial/ is the working link.

22:44 <paulproteus> Can I get you to submit a pull request and accept my apologies?

22:47 <paulproteus> ... I see what the problem is.

22:47 <paulproteus> I think I can submit that PR.

22:54 <paulproteus> (Done.)

22:54 <paulproteus> I should have better-tested my changes to README; thanks for finding.