Wednesday, 19 December 2012< ^ >
arjan has set the subject to: Zotonic - the Erlang Content Management Framework
Room Configuration

[00:11:48] Arjan leaves the room
[00:32:27] maas.maarten.zeeman leaves the room
[07:26:23] Arjan joins the room
[08:11:02] Arjan leaves the room
[08:31:27] Maas joins the room
[08:35:13] Arjan joins the room
[08:47:40] Andreas Stenius joins the room
[09:23:57] Maas leaves the room
[09:49:48] <Andreas Stenius> sorting media files on uploaded date seems broken... Arjan, look at the new screenshots you uploaded... the admin dashboard for instance, ends up on page 3, when it ought to be on page 1 (when sorting on Uploaded -)
[09:50:00] <Arjan> oh?
[09:50:19] <Arjan> I thought I fixed that
[09:58:41] <Arjan> lager:set_loglevel(console_backend, debug).
[10:00:41] <Andreas Stenius> nice
[10:02:09] <Andreas Stenius> I think there's quite a bit of more feedback to be given on errors, though. The other day I had a typo in the template. Sure, I got a detailed error in the console, but it would've been nice if it presented something about it on the web page too... so that's on my wish list for the near future...
[10:06:27] <Arjan> yep
[10:38:56] maas.maarten.zeeman joins the room
[11:34:53] <maas.maarten.zeeman> Yep, that would be nice for my colleague too.
[11:38:16] <Arjan> the chrome 500 page is not very usable indeed
[11:55:49] <maas.maarten.zeeman> I love the documentation :-)
[11:56:07] <maas.maarten.zeeman> http://zotonic.com/docs/0.9/manuals/modules/versioning.html?highlight=datamodel
[11:57:01] <Arjan> maas.maarten.zeeman: thanks :)
[11:58:04] <maas.maarten.zeeman> Saves a lot time. Otherwise I would have spent a lot of time digging through zotonic code.
[11:58:11] <Arjan> good that it works for you; pretty sparse docs but usable
[11:59:35] <maas.maarten.zeeman> There is a lot to explain.
[11:59:56] <Arjan> what I often miss is doc on return values of functions
[11:59:58] <Arjan> e.g. specs
[12:00:21] <Arjan> does it return just an id or {ok, Id}, that kind of stuff
[12:00:41] <maas.maarten.zeeman> Yes, specs are nice. Busy with adding them to tubby.
[12:00:51] <Arjan> :)
[12:01:02] <Arjan> cool you picked that name ;)
[12:01:04] <maas.maarten.zeeman> Trying to generate that with dialyzer and typer
[12:01:47] <maas.maarten.zeeman> Not very successful so far.
[12:02:07] <maas.maarten.zeeman> It just generates run(_) -> any() spec.
[12:02:18] <maas.maarten.zeeman> not very helpful.
[12:03:00] <maas.maarten.zeeman> It helps with catching type errors to standard erlang modules though.
[12:05:06] <maas.maarten.zeeman> Planning to release it this evening after adding specs and some minor tweaks.
[12:10:43] <maas.maarten.zeeman> btw do you know if there is a way to do optional compilation based on erlangs release number? -ifdef('R15')?
[12:11:00] <Arjan> I think I seen something like that
[12:11:08] <Arjan> somewhere
[12:12:07] <maas.maarten.zeeman> Met too, it could be that you knew it from the top of your head.
[12:12:15] <Arjan> dont know
[12:15:25] <maas.maarten.zeeman> ack-grepping, can be done later.
[12:53:51] <Arjan> I just made an awesome commit
[12:54:26] <Arjan> mod_development: Re-register notifier hooks after .erl compilation

When a file is recompiled, the observe_ and pid_observe_ functions are
automagically registered in the running sites; no need anymore to
restart your site for the observe functions to work.
[12:54:40] <Arjan> only for linux developers though as it hooks into the auto recompile
[13:01:21] <Andreas Stenius> cool :)
[13:23:00] <maas.maarten.zeeman> I just realized we have another naming conflict. Blocks.... they can be a {% block %} in your template or a block in your resource.
[14:12:55] <Arjan> oops :)
[14:12:59] <Arjan> you are right
[14:13:11] <Arjan> we should rename the resource blocks
[14:13:33] <Arjan> because they are newer and becaose template blocks share their name with Django
[14:21:18] <maas.maarten.zeeman> Nice one... just after redoing mod survey
[14:21:52] <maas.maarten.zeeman> Was looking around to see how it worked. Of course no documentation (yet)
[14:22:01] <maas.maarten.zeeman> :-p
[14:22:25] <maas.maarten.zeeman> Need a custom text block which is sanitized differently.
[14:23:21] <maas.maarten.zeeman> img links should be "repaired"
[14:25:19] <Andreas Stenius> ok, so maybe have rsc blocks => rsc sections... ?
[14:44:53] <maas.maarten.zeeman> Sections hmmm. How do other cms systems call these?
[14:58:58] <Arjan> +1 for sections
[14:59:14] <Arjan> mod_survey is pretty self-explanatory :p
[14:59:26] <Arjan> there is *some* docs actually
[15:14:50] <maas.maarten.zeeman> I would also like to add a notifier call to sanitize props/blocks. Now everything is always handled by z_html:escape_props, no way of escaping that
[15:16:42] <Arjan> you have my blessing :)
[15:16:51] <Arjan> ...if you document it, of course ;)
[15:23:23] <maas.maarten.zeeman> Bla, maybe I'll just use false for escape props ;-)
[15:24:18] <maas.maarten.zeeman> Of course I will document it.
[15:27:29] <Arjan> hehe
[15:27:53] <Arjan> did any of you notice that the p= argument gets lost with the new admin login
[15:28:00] <Arjan> the redirect-after-login feature
[15:29:42] <maas.maarten.zeeman> No, when did that happen.
[15:30:21] <Arjan> on urls like this
[15:30:21] <Arjan> localhost:8000/admin/logon?p=%2Fadmin%2Ffoo%2Fimport
[15:30:26] <Arjan> I already found it
[15:30:36] <Arjan> the page argument is overwritten inside admin_logon.tpl
[15:33:40] <Arjan> fixed
[15:38:56] <maas.maarten.zeeman> A bit scary to fool around with the html escaping
[16:06:44] <maas.maarten.zeeman> Hmm I think I can do what I want with the sanitize_element notification, and then do something extra in rsc_update
[16:07:15] <Arjan> from where does the sanitize_element trigger?
[16:09:31] <maas.maarten.zeeman> inside z_html, that is not so nice
[16:10:21] <maas.maarten.zeeman> It is now used to repair img src elements in z_admin
[16:11:04] <maas.maarten.zeeman> https://github.com/zotonic/z_stdlib/blob/master/src/z_html.erl#L433
[16:11:27] <maas.maarten.zeeman> Downside is that you don't know any other props of the resource here
[16:14:03] <Arjan> indeed, that does seem like too low-level for me
[16:21:08] <maas.maarten.zeeman> Hmmm. maybe escape_props should be adapted a little bit then.
[16:21:37] <maas.maarten.zeeman> It sure contains a lot of internal zotonic logic.
[16:21:51] <maas.maarten.zeeman> Like here: https://github.com/zotonic/z_stdlib/blob/master/src/z_html.erl#L68
[16:26:24] <Marc Worrell> yep, those callbacks should be in the 'options' list.
[16:26:46] <Marc Worrell> This was the quickest way to get the thing separated into its own library, without breaking anything :)
[16:27:02] <maas.maarten.zeeman> :-)
[16:27:24] <Marc Worrell> about the sections/blocks/whatever names
[16:27:38] <Marc Worrell> We have had a discussion back then how to name the things.
[16:27:56] <Marc Worrell> Section was one of the options, I think that was scrapped due to the html <section> tag.
[16:28:04] <maas.maarten.zeeman> That was right indeed.
[16:28:21] <Marc Worrell> can't find the discussion though
[16:35:01] <Arjan> heh
[16:44:34] <Marc Worrell> scary on OS X: ERTS_FP_CHECK_INIT at 0x8a7a9: detected unhandled
[16:46:40] <maas.maarten.zeeman> oops
[16:47:05] <Marc Worrell> some overflow during geo calculations
[16:47:40] <maas.maarten.zeeman> and than crash, bam, gone vm?
[16:49:31] <maas.maarten.zeeman> Eeks.
[16:55:57] <Marc Worrell> nope, just humming along :p
[16:57:08] <maas.maarten.zeeman> Just resets f
[16:57:15] <maas.maarten.zeeman> the fpu and continue...
[16:58:52] <maas.maarten.zeeman> Fail fast only goes so far I guess.
[17:24:27] <maas.maarten.zeeman> I would like to add a #block_escape notification so I can use that to have a custom escape function.
[17:25:03] <maas.maarten.zeeman> The default escaping is fine. I see not much need to customize that part.
[17:28:12] maas.maarten.zeeman leaves the room
[17:33:42] maas.maarten.zeeman joins the room
[17:34:05] <maas.maarten.zeeman> First things first, bike home and eat.
[17:38:34] maas.maarten.zeeman leaves the room
[17:43:17] Arjan leaves the room
[18:11:48] Arjan joins the room
[18:41:25] <Andreas Stenius> picking up the blocks/section topic...
[18:41:36] <Andreas Stenius> typo3 calls them elements (or content elements)
[18:42:46] <Andreas Stenius> http://docs.typo3.org/typo3cms/GettingStartedTutorial/EditAndCreatePagesAndContent/NewPage/Index.html#creating-page-content-elements
[18:43:52] <Andreas Stenius> Arjan: about the p= getting lost, I've noticed it, but was looking at too much other stuff to pay any attention to it at the time. +100 for fixing it ;)
[18:58:59] Arjan leaves the room
[19:01:02] Arjan joins the room
[19:14:05] maas.maarten.zeeman joins the room
[20:11:30] maas.maarten.zeeman leaves the room
[20:46:51] Andreas Stenius leaves the room
[20:49:07] andreas.stenius joins the room
[20:55:49] <andreas.stenius> wow, git surprised me just now (in a positive way)... I was on a dev branch (say, foo), and ran git fetch --all to check for new commits, and found a bunch that I wanted on my master. Usually, I check out master, do a pull, then rebase foo onto master. Now, I didn't want to bother with all that, so I though I'd just update master without checking it out first, running: git pull . origin/master:master well; that went smooth as butter, more so, it also rebased my foo branch for me..! (guess that came from the rebase option set for pull).. so that one pull got me up-to-date without having to switch branches at all :D
[20:56:37] <Arjan> wow
[20:56:56] <Arjan> that is serious git foo
[20:57:01] <Arjan> :)
[20:57:07] andreas.stenius liky :p
[20:57:58] <andreas.stenius> oh, that's actually a fetch and a pull... don't know if I can skip the fetch step for this to work (but I can live with that) ;)
[20:58:16] <Arjan> if youre wondering about my last commit, it prints something like this,
[20:58:17] <Arjan> https://gist.github.com/4339940
[20:58:17] <andreas.stenius> don't know = suspect not
[20:58:30] <Arjan> verry handy for newcomers
[20:58:40] <Arjan> yesterday again somebody thought zotonic listened on port 2525
[20:58:48] <Arjan> (for http)
[20:59:25] <andreas.stenius> ah, that ought to be clear enough.. :)
[20:59:52] <andreas.stenius> 2525 just screams smtp though, doesn't?
[21:00:44] <Arjan> well it was still weird that it told its smtp port but not its http port
[21:00:46] <andreas.stenius> I didn't get your last commit, that came after my "git foo"
[21:00:52] <Arjan> it is a web framework
[21:00:54] <Arjan> :p
[21:00:54] <andreas.stenius> true
[21:00:56] <Arjan> huh
[21:01:19] <andreas.stenius> you said if I was wondering about your commmit... I hadn't seen it yet... :p
[21:01:40] <Arjan> oh
[21:01:45] <Arjan> dont you get the mails?
[21:01:48] <Arjan> :)
[21:01:49] <andreas.stenius> typing like a duck again... :p
[21:01:53] <Arjan> or the tweets :p
[21:01:53] <andreas.stenius> sure
[21:02:02] <andreas.stenius> tweets..? no.
[21:02:12] <Arjan> speaking of tweets...
[21:02:21] <Arjan> why did that stop working...
[21:02:55] <andreas.stenius> did something tweet each commit... ?
[21:03:03] <Arjan> I set that up monday
[21:03:11] <Arjan> https://twitter.com/zotonicdev
[21:03:30] <andreas.stenius> oh, cool
[21:04:13] <Arjan> wow, https://github.com/zotonic/zotonic/settings/service/twitter
[21:04:20] <andreas.stenius> that's cool - makes it way easier to tweet the world when we fix things (and others to tweet when we break them.. :p )
[21:04:30] <Arjan> do you see that error page?
[21:04:36] <andreas.stenius> I'm no admin.... :/
[21:04:44] <Arjan> oh
[21:04:56] <Arjan> its an octocat falling in an abyss with a http 500 sign
[21:05:03] <andreas.stenius> lol
[21:05:22] <andreas.stenius> did you see their page of octocats (and other things..) loads of them..
[21:05:28] <Arjan> this one, https://github.com/500
[21:10:23] <andreas.stenius> nice one :)
[21:12:51] <andreas.stenius> ah, here it is: http://octodex.github.com/
[21:31:51] Maas joins the room
[23:15:08] andreas.stenius leaves the room
Powered by ejabberd Powered by Erlang Valid XHTML 1.0 Transitional Valid CSS!