| *** Aiste has quit IRC | 00:13 | |
| *** Aiste has joined #schooltool | 00:18 | |
| *** hazmat has quit IRC | 00:33 | |
| *** hazmat has joined #schooltool | 00:34 | |
| *** jinty has joined #schooltool | 00:54 | |
| jinty | btw: python-tz is sitting in the new que for debian. It is out of my hands now. Though there might be someone at cannonical who can expidite matters. | 01:00 |
|---|---|---|
| *** TheShizNig has joined #schooltool | 01:36 | |
| *** jinty has quit IRC | 02:36 | |
| *** TheShizNig has quit IRC | 02:53 | |
| *** hazmat has quit IRC | 04:55 | |
| *** hazmat has joined #schooltool | 07:04 | |
| *** hazmat has quit IRC | 07:04 | |
| *** SteveA has joined #schooltool | 08:52 | |
| *** vinsci has quit IRC | 08:58 | |
| *** d2m has quit IRC | 09:30 | |
| *** vinsci has joined #schooltool | 10:50 | |
| *** Aiste has quit IRC | 10:51 | |
| *** thisfred has joined #schooltool | 12:00 | |
| *** mgedmin has joined #schooltool | 12:48 | |
| *** d2m has joined #schooltool | 13:50 | |
| *** gintas has joined #schooltool | 14:16 | |
| *** Aiste has joined #schooltool | 14:55 | |
| *** Aiste has quit IRC | 17:42 | |
| *** mgedmin has quit IRC | 17:45 | |
| *** gintas has quit IRC | 18:35 | |
| *** thisfred has quit IRC | 18:51 | |
| *** Aiste has joined #schooltool | 19:00 | |
| *** mgedmin has joined #schooltool | 19:04 | |
| *** jinty has joined #schooltool | 19:20 | |
| *** tvon has joined #schooltool | 19:26 | |
| *** alga has joined #SchoolTool | 19:30 | |
| tvon | K, so it had been suggested to me that there was a better way of doing the dublin core than the mixin, but if there is I haven't found it | 19:49 |
| tvon | that said, should I create an interface for mixin classes? | 19:49 |
| *** bskahan has joined #schooltool | 19:51 | |
| mgedmin | tvon, could you be more specific? | 19:53 |
| *** jinty has quit IRC | 19:57 | |
| tvon | mgedmin: in my mail to the list about zopedublincore, I had an interface defined for the mixin class | 19:59 |
| mgedmin | I'm not sure such a mixin is a good idea | 20:00 |
| tvon | ah, okay | 20:00 |
| mgedmin | the question is: is 'title' data or metadata | 20:00 |
| mgedmin | or both? | 20:01 |
| tvon | in most cases I think it's both | 20:01 |
| * mgedmin nods | 20:01 | |
| th1a | In the abstract, it's metadata. | 20:01 |
| mgedmin | in that case I tend to think that there should be a regular 'title' attribute | 20:01 |
| mgedmin | on an object | 20:01 |
| mgedmin | without any magic | 20:02 |
| mgedmin | and if we want the same title to be accessible as DC metadata | 20:02 |
| mgedmin | then there should be an adapter that provides IZopeDublinCore for this object | 20:02 |
| mgedmin | proxies the 'title' attribute directly to context | 20:02 |
| mgedmin | and uses the "standard" annotatable dublin core adapter to proxy all other attributes | 20:03 |
| mgedmin | but I might be wrong in this | 20:03 |
| mgedmin | (I have never actually done that) | 20:03 |
| th1a | I'm a little dubious. | 20:03 |
| mgedmin | your approach was the exact opposite | 20:03 |
| mgedmin | and perhaps it is simpler | 20:03 |
| mgedmin | although it adds an extra requirement to the content object | 20:04 |
| mgedmin | you can no longer use that component without metadata | 20:04 |
| mgedmin | and I think unit tests become more difficult, because you cannot use the components without setting up IZopeDublinCore adapters | 20:04 |
| * mgedmin shrugs | 20:04 | |
| tvon | hrm | 20:05 |
| mgedmin | do we need zope dublin core for schoolbell's content objects? | 20:05 |
| tvon | need? I don't think so | 20:06 |
| th1a | Hm... | 20:06 |
| th1a | Mulling this over... | 20:10 |
| th1a | Dublin Core is for content objects, and applications aren't content objects. | 20:11 |
| th1a | application objects, that is. | 20:11 |
| * mgedmin thinks | 20:12 | |
| mgedmin | if things that can be added to a Zope 3 Folder are content objects | 20:12 |
| mgedmin | then the top 'SchoolBellApplication' object is a content object | 20:12 |
| mgedmin | this is starting to smell like philosophy... | 20:13 |
| th1a | They are philosophical questions that are very important to SchoolTool, actually. | 20:13 |
| th1a | By content object I guess I'm thinking of an artifact, like a document, photo, etc. | 20:14 |
| th1a | Those are the things Dublin Core was intended for. | 20:14 |
| * mgedmin nods | 20:15 | |
| th1a | I mean, you can say the dc:creator for mgedmin is his parents, but that's not really the intent. | 20:16 |
| mgedmin | it could be useful, though | 20:16 |
| mgedmin | "who added this group to the system?" | 20:16 |
| tvon | yeah | 20:16 |
| mgedmin | and "when" | 20:16 |
| tvon | "content" can be anything really | 20:16 |
| mgedmin | all persons/groups/resources are IAttributeAnnotatable | 20:17 |
| mgedmin | so they get DC metadata for free | 20:17 |
| tvon | I tend to consider it to be anything that can be added, manipulated and removed from the system | 20:17 |
| th1a | I dunno. | 20:17 |
| th1a | I think stretching it too far can introduce more confusion than clarity. | 20:17 |
| th1a | The dc:creator isn't the user who added an object, it is the person who created it. | 20:18 |
| th1a | Also, explicit is better than implicit. | 20:18 |
| tvon | the metadata is on the object in the system though, not wha the object represents outside of the system | 20:19 |
| tvon | er, "what" | 20:19 |
| th1a | Not in my opinion. | 20:20 |
| tvon | hrm | 20:20 |
| th1a | The metadata about a book in a library system is about the book itself, not the entry in the database. | 20:20 |
| tvon | this is true | 20:21 |
| th1a | In the long run, I think it is important to handle a lot of stuff like metadata rather than data. | 20:24 |
| th1a | For example, I think it is important to think of grades as being metadata attached to the work, rather than data standing on its own. | 20:24 |
| th1a | A grade for an individual assignment, specifically. | 20:25 |
| tvon | metadata makes my head hurt | 20:27 |
| th1a | Well, I don't think we need to treat the title and description of application objects as Dublin Core, so that should make your head feel better. | 20:28 |
| * mgedmin just realized that "application object" became an ambiguous term | 20:31 | |
| mgedmin | were you referring to ISchoolBellApplication objects, or IPerson/IGroup/IResource objects? or both? | 20:32 |
| th1a | Ooh... | 20:32 |
| th1a | I was just thinking about IPerson/IGroup/IResource. | 20:32 |
| th1a | Whereas we're really worried about events here, aren't we? | 20:33 |
| th1a | Calendar events. | 20:33 |
| tvon | I've mostly been thinking about person/group/resource.. it applies to events too thogh | 20:33 |
| mgedmin | are we? | 20:33 |
| tvon | though | 20:33 |
| alga | metadiscussions on metadata | 20:34 |
| th1a | I don't think DC applies to events. | 20:35 |
| alga | If we are not a CMS, is DC relevant at all? | 20:36 |
| th1a | Not generally. | 20:37 |
| alga | guys, what do you think about a (marker) interface for all schoolbell objects? | 21:14 |
| th1a | What do you mean? | 21:14 |
| alga | it would be handy to register the menu views, ACL, login, logout, etc | 21:14 |
| alga | a common interface that all the objects under SchoolBellApplication would implement | 21:15 |
| alga | right now we have several views registered for * | 21:15 |
| alga | but they break if you call them on a non-schoolbell object | 21:16 |
| * mgedmin considers this | 21:17 | |
| mgedmin | these views are defined for the schoolbell skin only, right? | 21:18 |
| *** Aiste has quit IRC | 21:22 | |
| *** Voblia is now known as Awayblia | 21:40 | |
| alga | that's an ideaa | 22:04 |
| alga | I missed that in fact | 22:04 |
| * mgedmin adding event.description support to icalendar code | 23:26 | |
| alga | are you sure you're not eating Etria's lunch? | 23:40 |
| th1a | It appears so. | 23:42 |
| mgedmin | I was tying up loose ends in schoolbell.calendaring.icalendar | 23:43 |
| mgedmin | and there was a TODO | 23:43 |
| mgedmin | relax, it was something like 5 minutes of work | 23:43 |
| *** SteveA has quit IRC | 23:45 | |
Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!