Chris Koeritz [Fri, 10 Nov 2017 16:03:17 +0000 (11:03 -0500)]
back to the chaos factory
this is the scrambled up version still in progress, and it's totally appropriate to the development branch, given that no one else is seeing that branch yet.
Fred T. Hamster [Fri, 10 Nov 2017 14:08:03 +0000 (09:08 -0500)]
back to square one
rebuilding the git pull process from scratch, since the simplistic approach we had enjoyed so well for so long totally doesn't work with branches, and our new fangled cobbled together version from across the internet isn't right either. need to build reliable pieces one at a time.
Chris Koeritz [Fri, 10 Nov 2017 00:59:49 +0000 (19:59 -0500)]
so not quite grokked
there really is a chain going all the way back to master. bummer. however, i think you can get all of the pending updates from upstream with a pull --all. trying that now.
Chris Koeritz [Fri, 10 Nov 2017 00:51:47 +0000 (19:51 -0500)]
even more careful push and pull
this may working now; we are comparing the branch to the parent, and were definitely seeing those as different at one point. but now with a healthy development branch, the parent and branch are the same. so the rule should be, hopefully, (1) when parent is different than self, you pull from your parent but push to yourself and (2) when parent and self are same, just push and pull. if that's not the right grokking of it, then i will need to rethink some things.
Chris Koeritz [Thu, 9 Nov 2017 22:19:40 +0000 (17:19 -0500)]
swerved off into new branch
now have the basics of standing up the whole app and site done. just want to make sure i don't clobber anything, and i wanted a development branch anyway.
Chris Koeritz [Thu, 9 Nov 2017 19:57:45 +0000 (14:57 -0500)]
switched sense of install
reached tipping point where we are putting more installs in /usr/local/fred than in $HOME. so it makes sense to fix the root installer's feisty meow apex variable.
Chris Koeritz [Thu, 9 Nov 2017 16:26:25 +0000 (11:26 -0500)]
tastier
still debugging, but also added a check on whether the local repo is clean or not. if it's clean, we don't try to commit, since that alone throws an error.
Chris Koeritz [Thu, 9 Nov 2017 16:14:26 +0000 (11:14 -0500)]
added careful error checking to rev control
these had some checks on actions, but not on all. i want anything that fails during an update or a push to cause the whole machinery to stop right there, because we cannot afford to be missing merge screw-ups.
Fred T. Hamster [Thu, 9 Nov 2017 13:47:03 +0000 (08:47 -0500)]
opensim script improvements
added helpful info in backup_opensim and got rid of silly defaults.
added new script to whack all the running opensim instances. especially useful right now since my sims are hanging after finishing shutdown.
Fred T. Hamster [Thu, 9 Nov 2017 05:35:59 +0000 (00:35 -0500)]
fixed sim lister
stopped being fooled by multiple matches on things from the example region file. then realized i don't need to even scan that, since i should only take in ini files and that's an example file.
Chris Koeritz [Thu, 9 Nov 2017 03:39:14 +0000 (22:39 -0500)]
tasty reborn terminal locking script
thought this wasn't handy, but i've had several times recently where it would have been. newly cleaned it up and added helpful text feedback, plus now it will show the nechung oracle when locked.
Fred T. Hamster [Mon, 6 Nov 2017 15:03:07 +0000 (10:03 -0500)]
hoping the screen debacle is done
it looks like they unbroke -L for 'screen' tool, such that it no longer requires a log file parameter (which had broken just everything). and now the log file parm is still available, but you jam it together with the -L flag.