Zotonic
Zotonic
zotonic@conference.zotonic.com
Wednesday, 13 March 2013< ^ >
arjan has set the subject to: Zotonic - the Erlang Content Management Framework
Room Configuration

GMT+1
[00:21:39] Arjan joins the room
[00:47:04] Arjan leaves the room
[01:03:01] Maas leaves the room
[01:04:09] Maas joins the room
[01:04:49] Maas leaves the room
[07:30:55] Maas joins the room
[07:52:45] simon.smithies joins the room
[08:13:11] Maas leaves the room
[08:31:53] Arjan joins the room
[08:49:31] Arjan leaves the room
[08:50:25] Arjan joins the room
[09:12:25] Arjan leaves the room
[09:13:21] arcusfelis joins the room
[09:24:52] Arjan joins the room
[09:31:54] arcusfelis leaves the room
[09:41:15] arcusfelis joins the room
[09:56:48] Arjan leaves the room
[09:57:21] Arjan joins the room
[10:15:38] simon.smithies leaves the room
[11:29:16] Andreas Stenius leaves the room
[11:32:58] Andreas Stenius joins the room
[12:14:38] <Arjan> for some reason, on each request, an ACL for {view, false} is done
[12:14:58] <Arjan> the mod_acl_simple_roles does (no longer) know how to handle that
[12:39:43] <Andreas Stenius> hmmm..
[12:40:13] <Andreas Stenius> {view, false} sholudn't reach the ACL module.. it should be resolved in z_acl
[12:40:42] <Andreas Stenius> going to read up on some code to confirm...
[12:44:21] <Andreas Stenius> Marc changed it recetly in this commit: https://github.com/zotonic/zotonic/commit/dafc7cb05f3381afe91d5282ddd9c494f222789c
[12:44:38] <Andreas Stenius> so controller_template calls controller_page:get_id/1
[12:45:01] <Andreas Stenius> and that returns false if the id is unknown or invalid
[12:45:24] <Andreas Stenius> ping Arjan ;)
[12:46:08] <Andreas Stenius> and indeed it doesn't end up in mod_acl_simple_roles (as far as I can tell, it shouldn't, at least)
[12:46:41] <Andreas Stenius> to me, it looks like it's all in order
[12:47:14] <Andreas Stenius> I made a comment on Marc's change, however: https://github.com/zotonic/zotonic/commit/dafc7cb05f3381afe91d5282ddd9c494f222789c#commitcomment-2740842
[12:47:49] <Arjan> ah yes
[12:47:52] <Arjan> I tracked it down
[12:47:53] <Arjan> the problem is,
[12:48:00] <Arjan> that controller_page:get_id/1 returns false when no id is set
[12:48:02] <Arjan> instead of undefined
[12:48:19] <Andreas Stenius> yeah, https://github.com/zotonic/zotonic/blob/master/modules/mod_base/controllers/controller_page.erl#L107
[12:48:33] <Arjan> Is It Safe to change it? :)
[12:48:36] <Andreas Stenius> to me it could just as well be undefined
[12:48:40] <Arjan> agreed
[12:48:41] <Andreas Stenius> not sure
[12:48:58] <Andreas Stenius> I can dig around see if I see anyone making any distinction of it
[12:49:09] <Arjan> its just used in controller_page
[12:49:19] <Arjan> because it was only exported on mar 5
[12:49:21] <Andreas Stenius> but my guess the anwer may have been yes, but not any longer
[12:49:38] <Andreas Stenius> yeah, but the result is passed to z_acl
[12:49:57] <Andreas Stenius> change it and see what happens :p
[12:50:00] <Arjan> k
[12:51:00] <Andreas Stenius> it affects this case https://github.com/zotonic/zotonic/blob/master/modules/mod_base/controllers/controller_page.erl#L38
[12:51:29] <Arjan> yes
[12:51:52] <Arjan> (zotonic001@ixion)2> m_rsc:exists(false, z:c(milledoni)).
false
(zotonic001@ixion)3> m_rsc:exists(undefined, z:c(milledoni)).
false
[12:52:06] <Andreas Stenius> ah, yes, I misread it
[12:52:17] <Andreas Stenius> so it should be fine with undefined in place of a false
[12:52:43] <Arjan> think so too
[12:52:46] <Arjan> i'll change it
[12:53:26] <Andreas Stenius> while you're at it, why not move get_id to controller_template.. ?
[12:53:37] <Andreas Stenius> kind of bugs me that they depend on each other
[12:54:46] <Arjan> marc actually did that the other way around on march 5
[12:54:52] <Arjan> better we check with him?
[12:55:01] <Andreas Stenius> no, he just exported it
[12:55:09] <Andreas Stenius> it was already there
[12:55:40] <Andreas Stenius> and got rid of a lighter get_id version in controller_template in favor of the version in controller_page
[12:58:35] <Andreas Stenius> "This fixes ACL issues where 'false' was passed in on ACL
checks." ... ? this shouldn't have been a issue.
[12:58:54] <Andreas Stenius> nevermind, my bad, again
[13:01:13] <Arjan> :)
[13:01:16] <Arjan> back in an hour
[13:16:24] jeff.5nines joins the room
[15:09:02] arcusfelis leaves the room
[15:34:31] Arjan leaves the room
[15:35:29] arcusfelis joins the room
[15:56:09] Arjan joins the room
[17:21:25] Andreas Stenius leaves the room
[17:25:55] Andreas Stenius joins the room
[17:25:58] jeff.5nines leaves the room
[20:08:12] arcusfelis leaves the room
[20:25:34] Arjan leaves the room
[20:29:28] Arjan joins the room
[20:56:21] Arjan leaves the room
[21:02:24] Arjan joins the room
[21:51:16] <Andreas Stenius> what do you say, isn't time we merged in the new stats stuff soon... ?
[21:51:31] <Andreas Stenius> It's not complete or anything, but what's there seems to run quite nice :)
[21:51:52] <Andreas Stenius> just pushed a few tweaks, supporting multiple series in the line charts...
[21:51:57] <Andreas Stenius> next up: legends
[21:56:49] Arjan leaves the room
[21:58:37] <Andreas Stenius> new stats screenshot: https://github.com/mmzeeman/zotonic/pull/2#issuecomment-14868010 :)
[22:44:53] Andreas Stenius leaves the room
[23:10:37] Arjan joins the room
[23:35:44] Arjan leaves the room
Powered by ejabberd Powered by Erlang Valid XHTML 1.0 Transitional Valid CSS!