February 28, 2019 
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | >
February 28, 2019 [00:08:14] *** alanc <alanc!~alanc@129.157.69.40> has quit IRC (Remote host closed the connection)[00:08:41] *** alanc <alanc!~alanc@129.157.69.40> has joined #oi-dev[00:08:41] *** ChanServ sets mode: +o alanc[01:26:58] *** igork <igork!~igork@91.204.56.74> has quit IRC (Ping timeout: 250 seconds)[01:48:19] *** igork <igork!~igork@91.204.56.74> has joined #oi-dev[02:34:32] *** v_a_b <v_a_b!~volker@p57A27EBF.dip0.t-ipconnect.de> has quit IRC (Ping timeout: 255 seconds)[07:57:48] *** v_a_b <v_a_b!~volker@p57A27E18.dip0.t-ipconnect.de> has joined #oi-dev[07:58:01] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has quit IRC (Quit: tsoome)[07:58:38] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has joined #oi-dev[08:02:47] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has quit IRC (Client Quit)[08:05:18] <leoric> mnowak_: we have an issue with python 3.5 and depot. Both sysrepo and depot use 32-bit apache. But we don't have 32-bit Python 3.5.[08:05:45] <leoric> So, we a) either switch depot/sysrepo to use 64-bit Apache[08:05:55] <leoric> b) either make apache 64-bit only[08:06:13] <leoric> B) seems to be a bit saner, but will likely break users[08:08:02] <leoric> our apache currently uses env vars to extract modules path[08:08:40] <leoric> it's either /usr/apache2/2.4/libexec or /usr/apache2/2.4/libexec/64[08:09:32] <leoric> we can't go Oracle way and just remove 32-bit parts[08:10:46] <leoric> need to have some compatibility here... Perhaps, keeping symbolic links 64->. and modules-32.load, modules-64.load => modules.load will help....[08:11:28] <leoric> and this means we can't just go dynamic workers as Oracle did, we can't switch them based on configuration (it's likely to be not updated)[08:11:40] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has joined #oi-dev[08:12:41] <leoric> or can... If we create new modules.load so that modules-*.load will not be preserved and move logic there[08:13:38] <leoric> but definitely apache switch to 64-bit-only in more-or-less non-breaking way will likely be not very elegant[08:13:53] <leoric> mnowak_:[08:13:57] <leoric> mno-hime:[08:16:46] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has quit IRC (Ping timeout: 246 seconds)[08:17:28] <leoric> or just screw users with preserve=renameold[08:17:44] <leoric> will have to check the effect[08:23:10] <leoric> I wish we could just document and break things at snapshot... But this means we should have /stable branch. And we have no people to support even /hipster[08:44:39] <mno-hime> leoric, generally, I am for breaking things, documenting way out of the breakage, and making them right for the future[08:45:25] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has joined #oi-dev[09:28:07] *** jimklimov <jimklimov!~jimklimov@broadband-90-154-72-196.ip.moscow.rt.ru> has joined #oi-dev[09:49:19] *** andy_js <andy_js!~andy@94.6.62.238> has joined #oi-dev[10:27:59] *** ptribble <ptribble!~ptribble@cpc92716-cmbg20-2-0-cust138.5-4.cable.virginm.net> has joined #oi-dev[13:12:50] *** wiedi <wiedi!~wiedi@91.64.150.166> has quit IRC (Ping timeout: 272 seconds)[13:18:54] *** wiedi <wiedi!~wiedi@ip5b4096a6.dynamic.kabel-deutschland.de> has joined #oi-dev[14:40:55] <leoric> should we renameold or renamenew given that old httpd.conf will never work?[14:56:27] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has quit IRC (Ping timeout: 240 seconds)[15:36:54] <jimklimov> leoric: upgrades of apache?[15:37:15] <leoric> yes[15:37:28] <jimklimov> I'd say the principle of least surprise (from my recent experience) is to fail the service with incompatible existing config, so admin notices the need to fix it[15:37:40] <jimklimov> rather than seem to provide a blank service with no contents :)[15:38:47] <jimklimov> at least, these failures helped me address problems quickly as I was doing the server upgrades, (and prepare for such problems for next similar cases), rather than by end-user complaints that something is wrong with a site[15:59:06] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has joined #oi-dev[17:16:21] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has quit IRC (Read error: Connection reset by peer)[17:16:46] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has joined #oi-dev[18:10:52] *** ptribble <ptribble!~ptribble@cpc92716-cmbg20-2-0-cust138.5-4.cable.virginm.net> has quit IRC (Quit: Leaving)[20:14:21] *** trisk <trisk!~trisk@2601:196:4700:3f0b:225:31ff:fe02:7a11> has joined #oi-dev[21:00:45] *** trisk <trisk!~trisk@2601:196:4700:3f0b:225:31ff:fe02:7a11> has quit IRC (Ping timeout: 252 seconds)[21:19:42] *** trisk <trisk!~trisk@2601:196:4700:3f0b:225:31ff:fe02:7a11> has joined #oi-dev[21:33:34] *** ptribble <ptribble!~ptribble@cpc92716-cmbg20-2-0-cust138.5-4.cable.virginm.net> has joined #oi-dev[23:12:59] *** ptribble <ptribble!~ptribble@cpc92716-cmbg20-2-0-cust138.5-4.cable.virginm.net> has quit IRC (Quit: Leaving)