Zotonic
Zotonic
zotonic@conference.zotonic.com
Saturday, 6 October 2012< ^ >
arjan has set the subject to: Zotonic - the Erlang Content Management Framework
Room Configuration

GMT+2
[00:07:18] Maas leaves the room
[07:25:24] arjan joins the room
[08:34:21] Maas joins the room
[08:35:50] <Maas> Looking at https://github.com/ShoreTel-Inc/erld/wiki
[08:37:35] Ilya Rezvov joins the room
[08:45:59] Ilya Rezvov leaves the room
[09:10:26] Ilya Rezvov joins the room
[09:38:53] <arjan> why do we still have the {host, sitename} option in the config?
[09:39:24] <arjan> since the host is always equal to the basename of the site dir
[09:40:02] maas.maarten.zeeman joins the room
[09:42:45] <maas.maarten.zeeman> hmm. yes indeed
[09:43:20] <arjan> sites manager now checks if the host option given is equal to the basename of the dir
[09:43:25] <arjan> if not it issues a warning
[09:43:43] <arjan> but all that can just be removed, I think
[09:44:21] <arjan> one config option less :)
[09:44:40] <maas.maarten.zeeman> Yes. Btw. what happens if you run multiple sites with the same basename?
[09:44:53] <maas.maarten.zeeman> uhh, hostname
[09:45:19] <maas.maarten.zeeman> ehh, no {host, ...} option. It is not even a hostname.
[09:45:26] <arjan> indeed
[09:45:34] <arjan> I'll remove it from the skeleton
[09:45:47] <arjan> 2 sites with the same name is not possible anyway
[10:05:08] andreas.stenius joins the room
[10:05:45] <andreas.stenius> Marc Worrell: you just had to steel back the MVP badge, did you!? :p
[10:05:58] <andreas.stenius> nice work :)
[10:08:04] <andreas.stenius> arjan: yeah, the host option was needed until about a year ago, when I updated the config stuff, it was made optional.
[10:09:01] andreas.stenius leaves the room
[10:40:37] maas.maarten.zeeman leaves the room
[11:33:01] Maas_ joins the room
[13:27:53] Maas_ leaves the room
[15:10:45] maas.maarten.zeeman joins the room
[15:11:25] maas.maarten.zeeman leaves the room
[16:23:38] arjan leaves the room
[16:26:16] arjan joins the room
[17:05:25] <arjan> I think it is also good to rename z_datamodel to z_domainmodel
[17:05:38] <arjan> "data model" refers to the generic system of rsc and edges
[17:05:54] <arjan> while domain model refers to a certain configuration of those
[17:39:17] Maas_ joins the room
[17:50:34] <Maas_> Sounds ok to me.
[17:59:23] Ilya Rezvov leaves the room
[19:04:45] Maas_ leaves the room
[19:06:55] Maas_ joins the room
[19:23:56] <Maas_> Reminds me I have to check something else... m_rsc:rid, and m_rsc:name_to_id usage. name_to_id can't fetch a resource which someone has given a name with just numbers. :-/
[19:26:31] <Maas_> in name_to_id should mostly be rid calls instead.
[19:53:09] <Maas_> Oops, found and fixed a problem in z_ids when generating random number in fixed=false mode.
[20:10:16] Maas_ leaves the room
[20:34:44] Maas_ joins the room
[20:46:44] andreas.stenius joins the room
[21:09:02] Maas_ leaves the room
[21:10:08] maas.maarten.zeeman joins the room
[21:53:07] Ilya Rezvov joins the room
[22:17:26] <Marc Worrell> Hei,
[22:18:19] <Marc Worrell> Should we add Facebook Open Graph meta tags to the Facebook module, or to the templates independently of the Facebook module?
[22:18:40] <Marc Worrell> I feel a bit to add it into the Facebook module, so you can opt to not have this stuff in your templates.
[22:21:37] <andreas.stenius> not sure I got what you said entirely, but it feels like if you disable mod facebook, it would be nice to not have any facebook related stuff mingling about... :)
[22:27:53] <Marc Worrell> yep, I thought the same
[22:28:07] <Marc Worrell> Will add it to mod_facebook.
[22:28:28] <Marc Worrell> "Open Graph Meta Tags" are the Google keywords :p
[22:28:54] <andreas.stenius> I've posted a little drawing in a google doc for a new acl module (sent out invitations, but I'm not quite finished.. but it's a start)....
[22:29:29] <Marc Worrell> :)
[22:30:52] <andreas.stenius> I see a minor error in the drawing, after reading your response to Greg... about that the user might not be a resource :p
[22:31:37] <Marc Worrell> normally, in Zotonic it is.
[22:31:53] <Marc Worrell> He is investigating his own non-rsc-id option
[22:32:02] <andreas.stenius> but doesn't affect the planned functionality. Phew. Uhm... welll... maybe it does. Since the user is a member of groups and ACL's by connecting them with edges....
[22:32:02] <Marc Worrell> and, of course, bumping into new territory
[22:32:07] <andreas.stenius> baha :d
[22:32:29] <Marc Worrell> in our case it is always a rsc id
[22:32:45] <andreas.stenius> yeah. Let's stick with that. :)
[22:32:51] <Marc Worrell> and I think that "if it is an integer, we should assume a rsc id" should be our rule of thumb
[22:33:28] <andreas.stenius> agreed. would complicate things too much if we didn't
[22:33:55] <Marc Worrell> heh, gatsy: http://arstechnica.com/security/2012/10/sha1-crypto-algorithm-could-fall-by-2018/
[22:34:21] <Marc Worrell> yep, we need to keep things decoupled, but also simple
[22:34:28] <andreas.stenius> ok, I ran out of steam on my acl writing... I'll see what comments you guys come up with after looking at it.. ;)
[22:34:41] <Marc Worrell> ok :)
[22:36:12] <andreas.stenius> interesting link...
[22:36:21] andreas.stenius reading...
[23:01:27] <andreas.stenius> ah, only scomps and cookbooks left to migrate over now :)
[23:01:36] <andreas.stenius> *only*... :p
[23:03:09] <Marc Worrell> An then we "only" need to write some extra stuff :p
[23:03:38] <andreas.stenius> yeah, then the "real" work begins ;)
[23:04:43] <andreas.stenius> thanks for the link to ars technica btw. Never stumbled upon it before... I like their style
[23:11:14] andreas.stenius leaves the room
[23:38:49] arjan leaves the room
[23:39:58] arjan joins the room
[23:48:03] <maas.maarten.zeeman> reading acl doc, sounds a lot like rbac, but with different terminology.
[23:49:42] <maas.maarten.zeeman> role => group, acl is sort of like the link between ((rcs, operation), role)
[23:59:42] <maas.maarten.zeeman> when a user logs in you can add the (current/effective) roles of the user to the session. Makes it easy to temporarily drop roles and such, and fast to check for permission on a resource.
Powered by ejabberd Powered by Erlang Valid XHTML 1.0 Transitional Valid CSS!