14:37 <abliss> yep, that seems to be the issue. i wonder if something inside ekam had cached the location of the meteor paths.

14:38 <abliss> oops no i spoke too soon, node_modules is still messed up :(

14:42 <abliss> seems like ekam didn't do all the rebuilding it should when i restarted continuous? i had to touch the node-capnp.cc file again

14:43 <abliss> then 'make bundle' didn't redo what it should...

14:44 <abliss> since it doesn't seem to have any dependency in the Makefile on the contents of node_modules/. so, touching make-bundle.sh and trying again...

14:50 <abliss> Well, now it gets far enough to boot, but the main page just shows "iron:router

14:50 <abliss> No route definitions found."

14:51 <kentonv> check JS console

14:59 <abliss> thanks... `There are multiple templates named 'changelog'. Each template needs a unique name.`

15:00 <kentonv> does 'git status' show some files that git isn't tracking in your source tree? Perhaps one called 'changelog.html'?

15:00 <kentonv> it was probably generated by an older version of the makefile and now the file has moved to a new location but you have to delete the old one otherwise meteor will find it...

15:01 <abliss> yep! trashing it. thanks.

15:02 <abliss> sadly `make update` again didn't figure out what it needs to rebuild. i'm gonna try touching make-bundle.sh again

15:05 <abliss> (sure, now seems like a good time to be recompiling bignum.o again... ;)

15:19 <abliss> no luck... trashing copies in tmp/ and shell/client/imorts... yay that worked!

15:22 <abliss> yahoo, i've got letsencrypt sandcats certs!

17:12 <isd> hooray!

17:13 <isd> Yeah I think I hit the issue with the changelog at some point when that got moved too.

17:13 <JacobWeisz[m]> I remember that.

17:27 <JacobWeisz[m]> So, like, for #1588, it says I can open the branch in GitHub Desktop, which then opens a pr/1588 branch, which isn't published to GitHub. So I think if I then hit Publish I should have the contents of the #1588 PR in a branch within the sandstorm repo?

17:28 <JacobWeisz[m]> Most of why I hate git is I swear it's just hard to be 100% sure you aren't about to do something dumb.

17:29 <JacobWeisz[m]> And I think this PR predates GitHub allowing people to allow maintainers to edit the PR itself.

17:32 <isd> I've never used GitHub desktop (or any of the github specific tooling for that matter)

17:33 <isd> ...usually the UX warts are on the Git side, rather than GitHub, but this sounds like the reverse.

17:35 <JacobWeisz[m]> I mean, I think I understand? I am annoyed that the web interface doesn't have a command for this.

17:37 <isd> Wow, that is a giant list of grammar fixes.

17:38 <isd> (Honestly with this kind of thing I generally find it's a better use of time to fix these things myself that to go back and forth with a non-native speaker trying to get their grammar to be perfect...)

17:48 <JacobWeisz[m]> Yeah, and it's a 2016 PR. So I'm just gonna take it over.

17:50 <JacobWeisz[m]> IMHO I've seen plenty of PRs before stalled on trivial fixes and I'd agree it's ideal to just take it and commit your fixes on top.

17:55 <isd> I used to work at a university where we had lots of student interns, most of them non-native speakers. I very quickly learned not to sweat the small stuff.

17:57 <JacobWeisz[m]> I mean I think some people might be upset if you take over their PR, but if the commit they did is there, and they haven't followed up in like a week. This four year old PR hasn't done any favors for anyone.

18:01 <isd> Yeah. Often though I'll either merge and add a grammar fix commit or send a PR with a pile of fixes into their PR; it's just not worth the trouble to describe a pile of trivial changes in English and then force the submitter to replicate them.

19:51 <JacobWeisz[m]> Okay, #3317 is off and running, I think? GitHub Desktop actually has some nice UI buttons when you poke around it. It's got a rebase on master button I employed here.

22:34 <JacobWeisz[m]> #3318 is another commit extracted from the bowels of history.