Monday, 17 December 2012< ^ >
arjan has set the subject to: Zotonic - the Erlang Content Management Framework
Room Configuration

[00:52:31] Maas leaves the room
[08:17:34] maas.maarten.zeeman joins the room
[08:35:05] maas.maarten.zeeman leaves the room
[08:42:04] maas.maarten.zeeman joins the room
[08:44:28] andreas.stenius joins the room
[08:59:30] Arjan joins the room
[09:27:57] <Arjan> http://blog.pamelafox.org/2012/12/a-tale-of-two-bootstraps-lessons.html
[09:28:27] <Arjan> that's why we're still on 2.0 and not on latest...
[10:04:47] <andreas.stenius> yikes!
[10:07:19] <andreas.stenius> just realizing that I'm committing to a master that is due to be a release... perhaps we should introduce something called feature-freeze, and branch of an upcoming release and only apply bugfixes and other minor corrections on that branch until release day, keeping new features on master... what do you think?
[10:07:40] <Arjan> plus with bootstrap there is no way (that I know of) to look at older documentation versions :p
[10:07:43] <Arjan> andreas.stenius: that is a good idea
[10:07:56] <Arjan> I can actually just make the release branch
[10:08:12] <Arjan> right now
[10:08:22] <Arjan> since today is release day anyway :)
[10:08:26] <andreas.stenius> yeah, we don't need to freeze, ... exactly :)
[10:08:51] <andreas.stenius> but I think my last commit is nice to have in 0.9
[10:08:56] <Arjan> k
[10:09:00] <Arjan> you were just on time then :p
[10:09:07] <andreas.stenius> :p
[10:10:04] <Arjan> here we go: git checkout -b release-0.9.x master
[10:10:12] <andreas.stenius> had a few weaknesses in indenting closing %} that was not on the same line as the opening {%, yet not alone on the last line...
[10:10:20] <Arjan> :)
[10:10:22] <andreas.stenius> yay! :D
[10:11:08] <Arjan> and pushed
[10:11:23] <Arjan> https://github.com/zotonic/zotonic/tree/release-0.9.x
[10:11:32] <andreas.stenius> got it :)
[10:12:00] <Arjan> how do you feel about the NEWS file?
[10:12:12] <Arjan> it kind of overlaps with doc/dev/releasenotes/
[10:12:14] <andreas.stenius> getting my trackz proj up-to-date, using my ny rbac module ;)
[10:12:29] <andreas.stenius> hmm...
[10:12:47] <Arjan> maybe we should only put unreleased features in there
[10:13:00] <Arjan> and have a reference to doc/dev/releasenotes to see older releases
[10:13:15] <Arjan> so that it basically becomes a log of new features
[10:13:33] andreas.stenius is thinking...
[10:13:49] <Arjan> and once we do the release, we move the news file to doc/dev/releasenotes/version.rst, and create a new (empty) one which just says "hey look in doc/..."
[10:14:05] <andreas.stenius> how about keeping the last release note in NEWS... ?
[10:14:28] <andreas.stenius> uh
[10:14:36] <Arjan> yes that is also nice
[10:14:39] <andreas.stenius> and on master, that is, the upcoming one...
[10:14:49] <Arjan> then it just duplicates one set of release notes right?
[10:14:56] <Arjan> instead of everything
[10:15:09] <andreas.stenius> yeah, and that is a copy of NEWS > rel_x.x.x
[10:15:17] <Arjan> k
[10:15:40] <Arjan> the zotonic_install script should also be removed from the root of the repo I think
[10:15:45] <Arjan> and updated
[10:15:47] <andreas.stenius> so we can update NEWS as we go, and on release day we have the notes ready :)
[10:15:53] <Arjan> yep inded
[10:16:38] <andreas.stenius> there's a slight diff betweern NEWS and rel_x.x.x. though
[10:16:52] <andreas.stenius> I like the git log in the rel notes
[10:17:06] <andreas.stenius> but maybe they don't need to be in the NEWS file (keeping it a bit more compact)
[10:17:22] <andreas.stenius> so a cp then git log >> rel_x.x.x.
[10:17:28] <Arjan> yeah
[10:17:33] <Arjan> git shortlog, actually
[10:17:39] <andreas.stenius> right
[10:18:04] <andreas.stenius> I've never actually used the command in question, but I think it does a good job :p
[10:27:10] <Arjan> https://twitter.com/ZotonicDev
[10:27:21] <Arjan> will publish zotonic commit messages
[10:28:12] <Marc Worrell> :-)
[10:28:35] <Marc Worrell> Zal het even noemen op @zotonic
[10:31:09] <Arjan> heeft nog niets getweet tot nu toe ;)
[10:31:16] <Arjan> wel handig ja
[10:33:04] <andreas.stenius> > 'll Just call on @ zotonic
> has never tweeted so far ;-)
> handy yes
[10:33:16] <andreas.stenius> (thanks to google translate.. :p )
[10:34:03] <Arjan> :p
[10:35:14] <Marc Worrell> :p
[10:35:20] <Marc Worrell> Dutch is taking of the World!
[10:35:22] <Marc Worrell> Yeah!
[10:35:43] <Marc Worrell> Everybody better start exercising theuir gGggggGgggGg
[10:36:44] <maas.maarten.zeeman> You ain't much if you ain't Dutch they told me when I was i Michigan
[10:36:54] <maas.maarten.zeeman> :-)
[10:37:31] <andreas.stenius> hehe :p
[10:38:09] <maas.maarten.zeeman> A lot of dutch people moved to that location in the past.
[10:38:45] <Marc Worrell> Michigan? Guess it is flat as a pancake then? With lots black&white cows
[10:39:08] <maas.maarten.zeeman> There was even a dutch village, with bright orange plastic roof tiles.
[10:40:00] <Arjan> 2 years ago when I was in the states I passed through a town called "Nederland"
[10:40:06] <Arjan> nothing dutch about it though :p
[10:40:13] <Arjan> Nederland, Colorado
[10:40:30] <andreas.stenius> isn't there a sort filter!?
[10:40:31] <maas.maarten.zeeman> holland, michigan
[10:41:22] <andreas.stenius> I have `a for p in id.o.pred` and want to sort them on title...
[10:41:36] <Arjan> not possible like that
[10:41:37] <andreas.stenius> *a `for...
[10:41:49] <Arjan> .o. always sorts on edge order
[10:42:09] <Arjan> you have to use m.search[{query hasobject=... sort=`pivot_title`}] for that
[10:42:11] <andreas.stenius> yeah, but I get a list of id's...
[10:42:15] <andreas.stenius> can't I sort them?
[10:42:30] <andreas.stenius> that's awkard...
[10:42:31] <Marc Worrell> guess you need to add it :)
[10:42:37] <andreas.stenius> :p
[10:42:41] <Arjan> btw
[10:42:47] <Arjan> how are we gonna call the next release?
[10:42:49] <Arjan> 1.0? or 0.10?
[10:42:55] <Arjan> or something completely different?
[10:42:56] <Marc Worrell> would be nice to have something like: mylist | sort : `title`
[10:43:03] <andreas.stenius> I'm feeling I use stuff in a way that you don't... I keep bumping into thinigs... ;)
[10:43:22] <Marc Worrell> andres: that is good :)
[10:43:33] <andreas.stenius> yeah, and a lot of work :p
[10:43:54] <Marc Worrell> We can call it 1.0 - but do we want elastic in that one?
[10:43:55] <andreas.stenius> I like 0.10... until we have 1.0 ready
[10:44:01] <Marc Worrell> good
[10:44:04] <Marc Worrell> 0.10 then
[10:44:17] <Marc Worrell> we can always decide not to release and skip to 1.0
[10:44:23] <andreas.stenius> yep
[10:44:55] <andreas.stenius> but I also like to not add "big" features to a x.y.Z release...
[10:45:11] <andreas.stenius> those ought to be minor fixes and bug corrections...
[10:45:34] <andreas.stenius> that's my way, at least :)
[10:45:40] <Marc Worrell> like changing the datamodel upside down :p — really something for 0.9.1 :p
[10:45:44] <andreas.stenius> bah, I meant to write view, not way
[10:45:53] <andreas.stenius> lol
[10:46:14] <Marc Worrell> mixing extra modules in should be ok though
[10:46:21] <Marc Worrell> moving them out should be a bigger release
[10:46:39] <maas.maarten.zeeman> :-)
[10:46:47] <andreas.stenius> +1 :)
[10:47:58] <maas.maarten.zeeman> sigh, what a crap browser is ff on the mac.
[10:50:04] <maas.maarten.zeeman> btw, chrome will block if you error in a websocket. still bleeding edge.
[10:54:08] <maas.maarten.zeeman> sorry, had to get that off my chest. :-) This poor imac has only 1 gig of ram.
[10:54:21] <Marc Worrell> that is a bit low...
[10:57:07] <maas.maarten.zeeman> It can be upgraded to 4. It is normally only used for browsing and mailing, but nowadays you can't do even that with 1 gig.
[10:59:38] <maas.maarten.zeeman> Annoying that chrome has a global js variable named 'event'...
[11:06:37] <maas.maarten.zeeman> Yesterday I moved mod_reg (process registry) into an erlang app. https://github.com/mmzeeman/reggy. I want to do the same with a module mod_pool which now manages pools with handlers e.g. for websockets, sessions.
[11:13:12] <maas.maarten.zeeman> no, weird, it is not global, magically available in the scope of an event handler.
[11:17:21] <Marc Worrell> strange, really a way to introduce errors in scripts...
[11:18:47] <maas.maarten.zeeman> indeed, especially refactoring handlers.
[11:19:11] <maas.maarten.zeeman> separate out a small part and forget to rename something frome event.data to data.
[11:19:55] <maas.maarten.zeeman> works on chrome, not on ff and ie (as it should)
[11:20:49] <Marc Worrell> they really need to make these things more strict - I dislike this kind of automagic stuff in language implementations
[11:21:19] <Marc Worrell> javascript is weird, this makes it even weirder
[11:21:30] <maas.maarten.zeeman> IMHO the most funny js object is the location object.
[11:21:39] <Marc Worrell> (or would it be because of a side effect of their javascript-replacement dart?)
[11:21:54] <maas.maarten.zeeman> hmm, i don't know.
[11:22:09] <Marc Worrell> I am not sure if you can call the location object a object… it behaves strange
[11:22:19] <maas.maarten.zeeman> don't think so because safari does the same.
[11:23:27] <maas.maarten.zeeman> In documentation they say that the href property is a default property. If you don't mention it, then that is used.
[11:23:28] <Marc Worrell> i think that there is some compiler/interpreter going on with that "object"
[11:23:43] <Marc Worrell> +magic
[11:23:46] <maas.maarten.zeeman> location = "foo" is the same as location.href = "foo"
[11:24:59] <maas.maarten.zeeman> yeah. I tried to wrap the location object to have a way of catching navigates.
[11:25:27] <maas.maarten.zeeman> there is not a navigate event :-)
[14:01:00] <maas.maarten.zeeman> is tobbe a good name for a pool implementation?
[14:12:35] <Arjan> heh
[14:34:19] <andreas.stenius> lol- well, it's a name alright :p
[14:35:44] <Arjan> or tubby
[14:35:46] <Arjan> :p
[14:35:51] <Arjan> bit more international
[14:47:57] <andreas.stenius> ok, I've pushed the new sort filter! ;)
[14:48:06] <andreas.stenius> see docs :p
[14:48:41] andreas.stenius dashes out the door to pick up the kids from day care...
[14:51:14] <Arjan> did you backport?
[15:01:45] <maas.maarten.zeeman> tubby sounds good.
[15:09:00] <maas.maarten.zeeman> also checked reggy code with dialyzer. you have to wait a bit for the generation of the plt, but then very nice tool. also typer.
[15:27:04] <andreas.stenius> @Arjan: no, but I can do that, just not right now (later today, pbly)...
[15:28:18] andreas.stenius leaves the room
[15:30:02] <Arjan> https://gist.github.com/4318674
[15:30:10] <Arjan> release notes
[15:30:15] <Arjan> anything else I forgot?
[15:30:23] <Arjan> I will go through the git log
[15:30:38] Andreas Stenius joins the room
[15:31:05] <Arjan> Andreas Stenius: ^--
[15:31:45] <Andreas Stenius> I'm throwing a quick glance at it, but I have a phone call I have to make, so I'll get back after that...
[15:32:43] <Arjan> sure
[15:32:51] <Andreas Stenius> looks good :) be back in a while...
[15:36:07] <Arjan> the git shortlog is maybe not so usable this time
[15:36:08] <Arjan> https://gist.github.com/4318723
[15:36:10] <Arjan> too many things
[15:36:11] <Arjan> :p
[15:42:44] <Arjan> Zotonic's core routines which serve a greater purpose, like date and
string manipulation and HTML sanitization, been moved out of the main
repository into a new subproject called z_stdlib.
[15:43:20] <maas.maarten.zeeman> Yes, that was a nice move
[15:44:12] <Arjan> :)
[15:44:15] <Arjan> worth mentioning
[15:46:25] <maas.maarten.zeeman> The new blocks structure also deserves a mention
[15:47:01] <maas.maarten.zeeman> or was that new in 0.8?
[15:48:30] <Arjan> no 0.9
[15:48:32] <Arjan> you're right
[15:48:53] <maas.maarten.zeeman> you almost forget after having used it for some time
[15:50:17] <Arjan> yup
[15:50:22] <Arjan> this release is pretty long overdue :p
[15:50:45] <Arjan> While editing resources it is now possible for editors to add
different "blocks" of content to the page. Blocks can be custom
defined and have their own templates.
[15:51:12] <maas.maarten.zeeman> I'm using the zotonic_status site more and more. Nice
[15:51:35] <Arjan> yeah we really should add more stuff there
[15:51:44] <Arjan> like real-time stats
[15:51:52] <maas.maarten.zeeman> was just typing that.
[15:51:54] <Arjan> that is already built but needs porting
[15:52:10] <maas.maarten.zeeman> also feedback for stop restart flush would be nice
[15:52:17] <Arjan> http://zotonic.com/media/inline/id/928
[15:52:18] <Arjan> true
[15:52:47] <maas.maarten.zeeman> yeah those. is that hard to add?
[15:53:06] <Arjan> not too hard I think
[15:53:16] <Arjan> but it adds hooks deep down the core
[15:53:20] <Arjan> counting queries etc
[15:53:25] <Arjan> and mostly in webzmachine actually
[15:54:18] <Arjan> ok updated release notes, https://gist.github.com/4318674
[15:59:52] <maas.maarten.zeeman> nice
[16:01:19] <Arjan> I tweaked the 1st core feature
[16:01:25] <Arjan> The ability was added to aid and promote "mobile first" development
using automatic user agent classification. Amongst others, the
template selection mechanism is now based on the detected user agent
[16:06:18] <Andreas Stenius> yep, this is a rather big release
[16:06:31] <Andreas Stenius> wow, jabber.org takes it's time delivering messages...
[16:07:45] <Arjan> heheh
[16:11:58] <Andreas Stenius> added has: https://gist.github.com/4318983/revisions
[16:13:08] Andreas Stenius is rebooting into Fedora...
[16:13:24] Andreas Stenius leaves the room
[16:14:56] andreas.stenius joins the room
[16:17:52] <Arjan> andreas.stenius: thanks
[16:18:01] <Arjan> documenting survey handlers while marc fixes his last things
[16:18:02] <Arjan> :p
[16:18:57] <andreas.stenius> ok
[16:19:09] <andreas.stenius> I "backported" my sort filter..
[16:19:53] <andreas.stenius> it was as easy as: git co -b release-0.9.x origin/release-0.9.x git merge master^1 (the ^1 to skip my news file update).
[16:20:16] <andreas.stenius> So, can add that to the list of new filters...
[16:21:33] <andreas.stenius> about the docs, are docs/latest/ going to follow master, or the latest release?
[16:23:20] <Arjan> I'd say, master
[16:23:31] <andreas.stenius> the shortlog is rather lengthy indeed, but isn't nice to list the ppl that have contributed there any how... ?
[16:23:53] <Arjan> I can add a list of contributors to the release notes?
[16:24:00] <Arjan> but everybody's name is in the main contributors file anyway
[16:24:26] <andreas.stenius> yeah, that might do it. Just a list so you can see who've been involved with this release...
[16:24:39] <andreas.stenius> It was quite a few ppl this time, which is nice :)
[16:25:06] <Arjan> k
[16:26:53] <Arjan> Ahmed Al-Saadi
Alain O'Dea
Andreas Stenius
Arjan Scherpenisse
Artur Wilniewczyc
Barco You
François Cardinaux
Grzegorz Junka
Hans-Christian Espérer
Joost Faber
Kunthar Daman
Maas-Maarten Zeeman
Marc Worrell
Motiejus Jakštys
Piotr Meyer
Tom Bradley
[16:28:49] <Arjan> updated gist
[16:29:53] <Arjan> ok
[16:32:01] <Arjan> ok pushed
[16:32:11] <Arjan> still not the release notes
[16:32:17] <Arjan> waiting on marc
[16:32:20] Arjan taps his fingers
[16:32:32] <Arjan> Marc Worrell: how is it hanging?
[16:32:41] <andreas.stenius> I'll add a note about the sort filter...
[16:33:02] <Arjan> k
[16:33:11] <Arjan> i would have actually preferred the name rsc_sort
[16:33:13] <Arjan> but ok :p
[16:34:58] <andreas.stenius> https://gist.github.com/4319128/revisions
[16:35:02] <andreas.stenius> ok, we can rename it
[16:35:07] <andreas.stenius> I'm not religious :p
[16:35:43] <Arjan> it might be confusing otherwise
[16:35:47] <andreas.stenius> as it is, rsc_sort makes sense, but it could be made more generic...
[16:35:52] <Arjan> e.g. {{ [1,3,2]|sort }}
[16:36:00] <andreas.stenius> yeah, exactly..
[16:36:03] <Arjan> should just output [1,2,3]
[16:36:04] <Arjan> k
[16:36:26] <andreas.stenius> hmmm.... that should be fairly easy...
[16:36:45] <andreas.stenius> your waiting for marc any way, aren't you... ? :p
[16:38:11] <Arjan> yep
[16:38:13] <Arjan> go ahead
[17:11:27] <Arjan> okay guys
[17:11:34] <Arjan> doing final checks and then doing the release!!
[17:16:06] <Arjan> nice country list
[17:17:17] <andreas.stenius> I'm having a bug I'm fixing in the tpl mode... will be a few minutes... (not sure how many minutes you need...)
[17:19:02] <Arjan> nooooooo
[17:19:03] <Arjan> :p
[17:19:12] <Arjan> I'm merging NEWS file conflicts :p
[17:19:29] <andreas.stenius> oh..
[17:19:40] <Arjan> no worries
[17:23:03] <andreas.stenius> ok, pushing it...
[17:23:25] <Arjan> k
[17:25:49] <andreas.stenius> done.
[17:26:13] <andreas.stenius> tweaking the tpl mode is tricky... fixing one part, breaks another.. :p
[17:26:16] <andreas.stenius> ok, dinner time
[17:26:23] andreas.stenius goes away...
[17:26:31] <Arjan> enjoy!
[17:26:39] <Arjan> when you're done I've baked the 0.9 :p
[17:26:41] <maas.maarten.zeeman> eet ze
[17:37:26] <Arjan> wtf
[17:37:28] <Arjan> ixion ~/tmp/zotonic/deps/mimetypes> ./rebar compile
==> mimetypes (compile)
ERROR: git describe --always --tags `git log -n 1 --pretty=format:%h .` failed with error: 128 and output:
fatal: Not a git repository (or any parent up to mount point /home)
Stopping at filesystem boundary (GIT_DISCOVERY_ACROSS_FILESYSTEM not set).
fatal: Not a git repository (or any parent up to mount point /home)
Stopping at filesystem boundary (GIT_DISCOVERY_ACROSS_FILESYSTEM not set).
[17:37:40] Maas joins the room
[17:37:53] <Arjan> anyone?
[17:37:57] <Arjan> rebar compile uses git?
[17:38:51] <Maas> deps maybe?
[17:39:08] <Arjan> it is mimetypes
[17:39:53] <Arjan> yep even when you download mimetypes from github as zip
[17:39:54] <Arjan> it fails
[17:40:00] <Maas> new deps inside mimetypes
[17:40:15] <Arjan> no
[17:40:37] <Arjan> ah
[17:40:45] <Arjan> mimetypes app.src has the following:
[17:40:47] <Arjan> {vsn, git}
[17:40:56] <Arjan> so rebar tries to do get the git version
[17:41:00] <Maas> oh yes
[17:41:10] <Arjan> but there is no .git dir when you download it as .zip
[17:41:14] <Maas> nasty
[17:41:33] <Maas> same happens when you download zotonic as zip from zotonic
[17:42:16] <Arjan> zucht
[17:42:25] <Maas> troost
[17:43:22] <Arjan> only mimetypes is borked
[17:43:29] <Arjan> all other submodules do it right
[17:43:55] <Maas> so via git submodules is better then rebar deps?
[17:45:28] <Arjan> I want people to be able to install zotonic without needing git
[17:46:00] <Arjan> from a single zip file
[18:08:12] Arjan leaves the room
[18:48:33] maas.maarten.zeeman leaves the room
[18:58:09] Arjan joins the room
[19:01:18] Arjan leaves the room
[19:35:23] Arjan joins the room
[21:22:56] Maas leaves the room
[21:25:26] Maas joins the room
[21:51:57] <Arjan> Renewed the screenshots on http://zotonic.com/gallery
[21:52:04] <Arjan> long overdue..!
[22:11:54] <andreas.stenius> I was thinking of those a few days ago...
[22:12:07] <andreas.stenius> looks really nice now! :)
[22:12:27] <andreas.stenius> cudos to the release engineer, too ;)
[22:14:44] <Arjan> tx
[22:14:52] <Arjan> its a lot of work
[22:17:06] <Arjan> also did the debian packagfe
[22:18:14] <andreas.stenius> nice
[22:18:26] <Arjan> now I'll pour myself a double whisky and crash on the couch :p
[22:18:38] <andreas.stenius> you've earned! :)
[22:19:10] <Arjan> cu tomorrow!
[22:19:14] <andreas.stenius> but when you get back , we ought to look at creating a 0.9 branch for the docs so we don't have the 0.10-dev title on each page of the 0.9 docs... that looks a bit funny...
[22:19:22] <Arjan> you're right
[22:19:23] <andreas.stenius> cya :)
[22:19:26] <Arjan> cu
[22:19:44] Arjan leaves the room
[22:38:33] andreas.stenius leaves the room
[23:48:20] Maas leaves the room
[23:49:53] maas.maarten.zeeman joins the room
[23:51:34] maas.maarten.zeeman leaves the room
Powered by ejabberd Powered by Erlang Valid XHTML 1.0 Transitional Valid CSS!