#Loqitantek: petermolnar left you a message 1 day, 15 hours ago: PHP 5.2 is without namespaces, anonym functions and it's brutally slow compared to 7.0, so yes, it's quite bad.
#petermolnarI'm getting desperately tired of WordPress, and if I change, I'll make my content _actually_ human readable, which leaves me with plain text with rather strict syntax for anything that requires parsing
#tantekpetermolnar I've been gradually implementing and selfdogfooding the features of tantek.com/w/Markdown
#tantekso far very few surprises, but definitely a few minor iterations
#petermolnarby the way, I want to keep the idea of fenced code blocks
#tantekthe trick with more memorable linking syntax (with all the video / image / URL / linktext ) possibilities is to prefer "richer" / "more primary" content first
#petermolnarthe ```language in my opinion is useful
#tantekpetermolnar I've never seen ``` used for that before markdown
#tantekit's a total made-up hack with zero basis in research AFAIK
#tantekdo you have something like a wiki yourself?
#tantek(obv my wiki is on a silo still, but better than nothing)
#aaronpkLinode is doing some server maintenance and needs to reboot all my servers
#aaronpkone (that isn't doing much) is scheduled for tomorrow 5am PDT, the rest (including the indieweb server) are 4am on Friday.
#aaronpkI have to decide whether to use this opportunity to move to their new infrastructure which will require similar downtime. Probably a good move otherwise I have to do this againz
#bearmy opinion is to never decide to test moving to new infra because of a quickly arriving infra reboot
#aaronpkIf I can finish my work release on the early side I can maybe do the reboots tonight
#aaronpkit should be a seamless move, Linode is good about that
#bearoh - letting them move your server image... that's different
#aaronpkits just moving to their KVM hosts from the XEN one
#bearsorry, I thought you were going to test some new provisioning script or somesuch
#bearI'm +1 for getting them moved ahead of the force reboots
#aaronpkI have too many apps on the server to try any sort of automatic provisioning stuff. Also that is a lot of extra work and I don't think it provides much benefit overall
#bearI keep saying that it doesn't have to be *you* that does the work
#aaronpkYeah but I do have to know how it works still
#bearand the benefit is that this documents for others how to do something similar
#bearsure, I get that you will want to know how it works - i'm not talking about something complicated
#aaronpkI have a text file I keep notes in right now
#bearmore like a series of bash scripts that sets up the server for the apps and then some that install/start the apps
#aaronpkI did write a pretty thorough automation script for the Geoloqi servers a while ago
#aaronpkbut it involved some config file templates and tricky bash scripts to move them around and replace pieces
#bearall I've ever tried to do, and will still try to do, is remove the *chore* of infra maintenance from you so you have brain cycles for the other stuff
#bearbecause we (the IndieWeb community) are very very close to having infrastructure that is required to have high uptimes
#aaronpksorry, I don't mean to sound cranky or ungrateful
#aaronpki think i might be missing an nginx plugin too, crap.
#bearthat is (to be honest) much better than I normally see from a dev
#bearmaking that into some simple form of automation would be very easy
#aaronpki realized a couple years ago that every time i have to figure out what i installed i was wasting time, and it'd be easier if i just wrote it down
#bearpulling known versions of your key services is also a great thing
#aaronpkyou'll notice sometimes i link to stackoverflow or other places where the process has changed over the years, so i know where to check if something is updated
#aaronpki also realized that if i get in the habit of updating nginx/php/ruby regularly, i get better at it, and doing smaller changes more frequently means fewer things break
#aaronpkso at the beginning of the year i decided i would try upgrading everything on the system every 3 months
#bearthat was what I was crankily referring to above :)
#bearchange management done incrementally is always better
#aaronpkone of my servers that's being rebooted tomorrow morning is about 7 years old, and i've been trying to move off of it for about 4 years. it's too old to upgrade now. it's so old that the openssl bugs found weren't even in its openssl yet.
#aaronpkthis is why i'm trying to get better about regular updates :)
#bear7 yrs old - I yell at my team when a server is more than a month old
#aaronpki finally shut down one of my other servers a couple months ago after finally moving everything off of it. but that was about 2 years later than I wanted to shut it down
#aaronpkthis is definitely a bad habit of mine. the "oh i'll totally move all these apps to this new server let me go set up the new server now" and then not realizing some dependency or something that makes me leave the old server online for 2 years more than it should have been
#aaronpkso i'm actually down to just one old server now. my target is having 3 linodes.
#miklbthis is the full error message in logs EPROTO 140499128305536:error:14077438:SSL routines:SSL23_GET_SERVER_HELLO:tlsv1 alert internal error:../deps/openssl/openssl/ssl/s23_clnt.c:769