Zotonic
Zotonic
zotonic@conference.zotonic.com
Tuesday, 24 September 2013< ^ >
Arjan has set the subject to: Zotonic - The Erlang Web Framework & CMS
Room Configuration

GMT+2
[00:10:33] Steffen Hanikel leaves the room
[01:34:05] cillian.deroiste leaves the room
[02:01:21] Mawuli Adzaku leaves the room
[02:01:56] Mawuli Adzaku joins the room
[04:31:46] Mawuli Adzaku leaves the room
[04:33:38] Mawuli Adzaku joins the room
[04:57:44] Mawuli Adzaku leaves the room
[04:58:21] Mawuli Adzaku joins the room
[06:34:52] Mawuli Adzaku leaves the room
[06:37:00] Mawuli Adzaku joins the room
[08:15:39] Mawuli Adzaku leaves the room
[08:21:53] Maas joins the room
[08:23:04] Maas leaves the room
[08:35:42] arjan joins the room
[09:26:11] maas.maarten.zeeman joins the room
[09:38:13] <z-bot> [blaaa] I don;t know why, had not changed anything but today after starting my test-zotonic the log files have appeared in /var/log/. great!
[09:40:36] <Marc Worrell> yeah! :-)
[09:40:51] <Marc Worrell> we need to make that logging better - any patches are welcome! :)
[09:41:26] <Marc Worrell> we also saw that the current (Basho) logger consumes quite a lot of time on busy servers.
[09:55:32] <z-bot> [blaaa] I have no idea if it would fit your goals (and no idea about erlang and what is already available for it) but would logging through a syslog protocol be possible?
[09:56:36] <z-bot> [blaaa] I'm sorry I won't be able to contribute any code...
[09:58:08] <Marc Worrell> good ideas are already a contribution :)
[09:58:58] <arjan> lager most likely already has a syslog backend
[09:59:17] <arjan> https://github.com/basho/lager_syslog
[09:59:31] <arjan> lager is the logging library we use
[10:01:09] <Marc Worrell> We should check for a better default logging
[10:01:23] <Marc Worrell> something with log rotation and more efficient
[10:01:42] <Marc Worrell> would love to see crashes in the log as well - maybe the watchdog process can do the logging?
[10:02:44] <maas.maarten.zeeman> Maybe this: https://github.com/ShoreTel-Inc/erld/wiki
[10:04:18] <maas.maarten.zeeman> Takes care of a couple of things which are a bit awkward in pure erlang.
[10:05:25] Kaos leaves the room
[10:21:17] Kaos joins the room
[10:29:38] <z-bot> [blaaa] The zotonic start script works fine for me now, but I had to increase the startup timeout for the systems .service calling it to a long time... A simple zotonic start takes a little over three minutes on my low-end device... can that somehow be cut down a little?
[10:30:13] <z-bot> [blaaa] Performance after startup is ery good for a blog of a few pages. The low end device is a Beaglebone Black by the way
[10:30:20] <z-bot> [blaaa] o it's very low end
[10:33:18] <arjan> the issue is dat it does a make on startpu
[10:33:28] <arjan> which is a bit weird I agree
[10:35:03] Kaos leaves the room
[10:35:44] <maas.maarten.zeeman> With the erld wrapper you can let erlang behave as a normal daemon. This takes the log rotation configuration out of the responsibility of erlang too. We have logrotate to manage that.
[10:36:15] Kaos joins the room
[10:40:18] <Marc Worrell> I think we should remove the make on startup - and just stop with a warning when some beam files are not found in ebin
[10:41:09] <Marc Worrell> @maas - how is the situation for our windows friends? any system level logrotation there as well?
[10:41:56] <maas.maarten.zeeman> ow yes, erld is unix specific.
[10:42:26] <maas.maarten.zeeman> but our scripts are unix specific too I think
[10:43:43] <Marc Worrell> there is also some .bat - never used it though
[10:47:18] <z-bot> [blaaa] Yes, I had noticed the make on startup. I was looking if I could somehow create some faster check if there are missing beam files and then do a targeted build of them. The way the make works now is just executing all Makefiles in the subdirectories right?
[10:48:28] <Marc Worrell> maybe just "rebar compile" - the current make also goes around updating the deps
[10:49:16] <Kaos> the .bat files are a pain, and doesn't work at the moment..
[10:49:41] <Marc Worrell> windows is a pain, and doesn't work at the moment
[10:49:45] <Marc Worrell> :p
[10:49:50] <Kaos> amen :p
[10:50:04] <Kaos> hasn't worked since v3.11...
[10:50:24] <Kaos> actually, I think the first version I came across was 3.10.. :p
[10:50:36] <Marc Worrell> BTW I think that Google has started displaying notices for IE8 (and older) - might mean that we can also start dishing IE8
[10:50:41] <maas.maarten.zeeman> LOL. Colleague was just explaining to client that IE works "differently" :-)
[10:51:07] <Marc Worrell> yep - very differently - just like good old Mosaic
[10:51:10] <Kaos> LOL... as in "nothing wrong with being different", eh?
[10:51:12] <z-bot> [blaaa] I thought the recent IEs were all right
[10:51:25] <Marc Worrell> IE10 is quite ok
[10:51:32] <Kaos> IE8 isn't that recent..
[10:51:35] <Marc Worrell> don't know about IE11
[10:51:44] <z-bot> [blaaa] No IE8 is old
[10:51:56] <maas.maarten.zeeman> IE10 in "compatibility" mode?
[10:52:16] <Marc Worrell> A friend just had to deal with a big organisation that just "upgraded to IE8"
[10:52:32] <Marc Worrell> incredible - incompetent IT departments
[10:52:54] <maas.maarten.zeeman> Meaning all IE7 quirks turned on
[10:53:23] <Marc Worrell> Guess so - and a kazillion ActiveX components - so that they can't move forward (I guess)
[10:53:26] <maas.maarten.zeeman> In organisations that is quite advanced that they have IE8 :-)
[10:53:47] <maas.maarten.zeeman> Generated website from in-house applications with activex indeed.
[10:53:48] <Kaos> those darn organisations.. :p
[10:54:00] <Marc Worrell> just bomb them
[10:54:17] <z-bot> [blaaa] I believe some organizations are really locked in, can;t really move away from legacy stuff. There are even companies still maintaining old mainframes...
[10:54:23] <Marc Worrell> at least Google is pushing forward with forcing upgrades
[10:54:35] <maas.maarten.zeeman> Mainframes are cool :-)
[10:54:38] <Marc Worrell> nothing wrong with mainframes :p
[10:55:04] <Marc Worrell> and especially nothing wrong with card readers and line printers. Fine pieces of robust technology.
[10:55:11] <maas.maarten.zeeman> VM's done in the right way.
[10:55:12] <Kaos> which may be a counterproductive in that way that managers won't let them in at all... oh well.. maybe we should bomb IE rather then the organisations..
[10:55:35] <Marc Worrell> MS is bombing itself these days
[10:55:42] <maas.maarten.zeeman> LOL
[10:56:07] <Marc Worrell> And our mobile life just got easier with the suicide of BlackBerry
[10:56:51] <maas.maarten.zeeman> Wondering what apple is up to with that "desktop class" A7 chip in a phone.
[10:56:52] <Kaos> why can I pipe the output to a function (in bash) but not redirect the output of a function to another (i.e. this is ok: func1 | func2, but not func2 < func1 )... aargh@bash.
[10:57:25] <z-bot> [blaaa] Most phones are desktop class now, if you look at it from 10 years ago
[10:58:51] Mawuli Adzaku joins the room
[10:59:27] <maas.maarten.zeeman> 64 bit cpu, just add screen mouse and keyboard. (bluetooth mouses are supported in ios7)
[11:00:51] <Marc Worrell> yes, that A7 is really great
[11:01:11] <maas.maarten.zeeman> free iwork apps for all new devices... too, and nice web apps for that as well. They are upto something.
[11:01:25] <Kaos> blarg, wrap it <( func1 ) should work, but then func1 would be run in a subshel, so it can't modify the environment... gah.
[11:01:35] <Marc Worrell> They will use it in their tablets, and I think a game/media console is not far away
[11:02:03] <Marc Worrell> @kaos nice all those () <>
[11:02:49] <Kaos> yeah.. bash gives me headache.. spawns a new shell by just looking at it..
[11:20:02] Steffen Hanikel joins the room
[11:38:10] Mawuli Adzaku leaves the room
[11:38:38] Mawuli Adzaku joins the room
[11:42:15] Mawuli Adzaku leaves the room
[11:42:55] Mawuli Adzaku joins the room
[11:44:55] Mawuli Adzaku leaves the room
[17:03:39] arjan leaves the room
[17:23:39] maas.maarten.zeeman leaves the room
[17:34:32] arjan joins the room
[17:36:40] arjan leaves the room
[17:44:44] arjan joins the room
[17:47:39] arjan leaves the room
[18:39:33] Kaos leaves the room
[18:41:23] Andreas Stenius joins the room
[19:10:39] arjan joins the room
[19:31:52] arjan leaves the room
[20:22:25] arjan joins the room
[20:26:22] cillian.deroiste joins the room
[20:40:34] Maas joins the room
[21:02:43] Maas leaves the room
[21:13:19] Maas joins the room
[21:25:30] Maas leaves the room
[21:26:38] Maas joins the room
[21:40:37] Maas leaves the room
[21:53:12] arjan leaves the room
[23:32:26] Steffen Hanikel leaves the room
Powered by ejabberd Powered by Erlang Valid XHTML 1.0 Transitional Valid CSS!