| *** aelkner_ has quit IRC | 01:25 | |
| *** ignas has joined #schooltool | 01:32 | |
| *** aelkner has joined #schooltool | 02:24 | |
| *** fsufitch has quit IRC | 02:55 | |
| *** pcardune has joined #schooltool | 03:58 | |
| *** didymo has joined #schooltool | 04:28 | |
| *** didymo has quit IRC | 04:33 | |
| *** lisppaste5 has quit IRC | 06:30 | |
| *** lisppaste5 has joined #schooltool | 06:40 | |
| *** subir has joined #schooltool | 08:03 | |
| *** subir has quit IRC | 08:40 | |
| *** pcardune has quit IRC | 09:41 | |
| *** pcardune has joined #schooltool | 09:44 | |
| *** pcardune has quit IRC | 10:27 | |
| *** mgedmin has joined #schooltool | 12:00 | |
| *** pcardune has joined #schooltool | 14:14 | |
| *** alga has joined #SchoolTool | 14:19 | |
| *** ignas has joined #schooltool | 14:59 | |
| *** mgedmin has quit IRC | 15:50 | |
| *** ignas has quit IRC | 15:50 | |
| *** mgedmin has joined #schooltool | 16:01 | |
| *** ignas has joined #schooltool | 16:12 | |
| th1a | ignas: So I'm trying to figure out something coherent to tell jelkner about what he should expect. | 16:20 |
|---|---|---|
| ignas | hmm | 16:21 |
| ignas | a good question, that I don't know the answer to at the moment | 16:22 |
| ignas | it seems that Cando has no "owner" at the moment | 16:23 |
| ignas | and the plan is to grow some interns to be responsible for cando development | 16:24 |
| ignas | but really, I don't know anything specific about it :/ | 16:24 |
| ignas | it seems that all the involved parties have different sets of needs in their heads | 16:24 |
| ignas | who is the person leading and managing the CanDo project? | 16:25 |
| ignas | not "giving money" and not "writing code", but deciding what code must be written and who will write it, and when + what will be released | 16:26 |
| ignas | or is the project completely committee driven at this point | 16:27 |
| pcardune | what happened to dwelsh? | 16:28 |
| ignas | don't know, but I definitely haven't heard his opinion about the schooltool cando integration | 16:28 |
| pcardune | I haven't heard anything from him in at least 4 months | 16:30 |
| ignas | what I will do (ccarey will do would be more precise) | 16:33 |
| ignas | is refactor cando to work just like lyceum school setup | 16:33 |
| ignas | with working translations | 16:33 |
| ignas | egg packaging | 16:33 |
| ignas | after every commit that passes tests | 16:34 |
| ignas | and make cando work with nightly schooltool builds | 16:34 |
| ignas | maybe even get ccarey to extract dashboard out of schooltool (it's cando functionality that has been stuck into schooltool) into a separate egg | 16:36 |
| ignas | so he would get some more buildout egg practice | 16:36 |
| * th1a realizes he covered up his irc window after asking a question. | 16:40 | |
| th1a | One issue here is that dwelsh and jelkner are essentially starting a new school. | 16:40 |
| th1a | I don't know if ignas has been informed of this. | 16:41 |
| ignas | nope | 16:41 |
| th1a | OK. Let's try to expand the context here then. | 16:41 |
| th1a | I guess it is my job to make sure everyone knows what is going on. | 16:41 |
| th1a | This is all pretty recent, regardless. | 16:41 |
| th1a | So yes, a proposal was accepted by the state or someone to close, re-design, and re-open dwelsh's school. | 16:42 |
| th1a | With dwelsh probably being essentially assistant director. | 16:43 |
| th1a | And jelkner working there. | 16:43 |
| th1a | Which means two basic things: | 16:43 |
| th1a | 1) dwelsh and jelkner are going to be super-busy; | 16:43 |
| th1a | 2) we'll have an even better test-bed for SchoolTool and CanDo going forward. | 16:44 |
| th1a | Overall it is good news. | 16:44 |
| th1a | Also, it is likely that more state funding for CanDo will turn up. | 16:44 |
| ignas | hmm | 16:45 |
| th1a | But basically, my perspective on what we're trying to do here, in the larger sense, is to give CanDo a set up which will be able to coast between gaps in funding without falling apart. | 16:45 |
| th1a | So the less forked they are from SchoolTool, the less likely that they'll get out of sync if nobody is minding the CanDo store for six months. | 16:46 |
| th1a | I definitely am not ok with ignas becoming the de facto manager of CanDo, and I don't think that's the intention. | 16:47 |
| th1a | Although I am in favor of having ignas dictate how CanDo should handle its releases, etc., so that they will be in sync with SchoolTool in a rational manner. | 16:47 |
| th1a | I think that's in line with what ignas wrote above. | 16:48 |
| ignas | kind of my plan, get CanDo into the same state schooltool.gradebook is in | 16:48 |
| ignas | if I will break it | 16:49 |
| ignas | I will fix it | 16:49 |
| ignas | but otherwise - i am not touching it | 16:49 |
| th1a | Yes. | 16:49 |
| th1a | I guess the difference with what jelkner would like is primarily having you do lots of code reviews, which I've concluded is unrealistic. | 16:50 |
| ignas | seems so | 16:51 |
| ignas | I will look at the checkins and will try to point out if there is something that can and should be fixed easily | 16:52 |
| ignas | but other than that it should be the responsibility of the programmers that are working on cando | 16:53 |
| ignas | it's still their project, not mine | 16:53 |
| ignas | so I am the federal government and they have their own State ;) | 16:53 |
| ignas | you being the president of course ;) | 16:54 |
| th1a | Of course. | 16:59 |
| *** pcarduner has joined #schooltool | 17:29 | |
| *** pcarduner is now known as pcardune_ | 17:29 | |
| aelkner | ignas: did you get my email? | 17:59 |
| aelkner | pcardune_: how's it going? | 18:07 |
| aelkner | th1a: ayt? | 18:10 |
| th1a | aelkner: Hi. | 18:11 |
| aelkner | hey, i saw your cando discussion | 18:11 |
| th1a | Yes? | 18:12 |
| aelkner | i was going to say | 18:12 |
| aelkner | definitely there's a problem with management of cando | 18:12 |
| aelkner | if the state of virginia would pony up the funds | 18:13 |
| aelkner | then maybe they could actaully pay someone to do that job | 18:13 |
| *** pcardune_ has quit IRC | 18:13 | |
| aelkner | as ignas said, play customer and decide what needs to be done | 18:13 |
| pcardune | hi aelkner | 18:13 |
| aelkner | hey pcardune | 18:13 |
| aelkner | things don't change much in cando land, do they? | 18:13 |
| aelkner | that could turn around later this year | 18:14 |
| aelkner | if the state gets as ecited about cando as they seem to start getting | 18:14 |
| aelkner | maybe by summer, they will decide that it's important enough to pay someone to manage it | 18:14 |
| aelkner | but tom is right about jelkner and dwelsh having their hands full | 18:15 |
| aelkner | there defintely will need to be a manager if it goes state-wide | 18:16 |
| aelkner | there's no way around it what with all of the release issues and user stories | 18:16 |
| aelkner | and interns to be managed | 18:16 |
| aelkner | there's a lot there that needs to be handled | 18:16 |
| th1a | Well, there won't really be new features this year. | 18:16 |
| aelkner | you mean the rest of the semester, right? | 18:17 |
| th1a | The tidying and consolidation is the main objective. | 18:17 |
| th1a | For the year, really. | 18:17 |
| aelkner | cause i know that the summer will bring many changes | 18:17 |
| aelkner | and by then we may have a commitment from the state for state-wide deployment | 18:17 |
| th1a | What changes? | 18:18 |
| aelkner | at the cando meeting last friday night | 18:18 |
| aelkner | we discussed the user stories at the top of the list | 18:18 |
| aelkner | things like global ids for comps | 18:18 |
| aelkner | right now there is a uri for each comp | 18:18 |
| aelkner | but no convention, so that Lee Caps would change the ids each time he rerelases the list | 18:19 |
| aelkner | that would be bad | 18:19 |
| aelkner | plus there's no convention for sols vrs comps | 18:19 |
| aelkner | that needs to be decided | 18:19 |
| aelkner | then of course there's the issue of data agregation | 18:20 |
| aelkner | that the state would want | 18:20 |
| aelkner | initially we thought SIS would be what we would use for that | 18:20 |
| aelkner | but we've decided that SIS is a bit overkill | 18:20 |
| th1a | SIF? | 18:20 |
| aelkner | hehe SIF + ZIS = SIS | 18:21 |
| aelkner | a simpler solution would be for all candos to report their data to a central cando instance | 18:21 |
| aelkner | and let the State access that instance | 18:21 |
| th1a | Well, that's very much stuff that is going to happen if the state pays for it or not happen. | 18:22 |
| aelkner | dwelsh made a good point when he said | 18:22 |
| aelkner | actually, we don't have to do anything | 18:22 |
| aelkner | let the states agregate the data manually | 18:22 |
| aelkner | that's done in business all the time | 18:22 |
| aelkner | then when they realize the cost in labor | 18:22 |
| aelkner | they could come to the conclusion that it would be worth money | 18:23 |
| aelkner | to pay cando developers to automate it | 18:23 |
| aelkner | looking at it that way, we don't have to beg them | 18:23 |
| aelkner | it will be their idea | 18:23 |
| aelkner | i think that makes good business sense | 18:24 |
| aelkner | it's amazing how much redendant data entry people are willing to put up with | 18:24 |
| aelkner | but also how much they appreciate when someone offers a aoutomatic solution | 18:25 |
| aelkner | so I think it's a good plan | 18:25 |
| aelkner | can't remember the other user stories | 18:25 |
| aelkner | oh, except that they want the schooltool admin interface to look more like cando | 18:25 |
| aelkner | which i think we can get into at the sprint | 18:25 |
| aelkner | and basically the need to keep the interns working on the project is important | 18:26 |
| aelkner | so that they retain what they learned and keep progressing | 18:26 |
| aelkner | so bug fixes this semester are a good outlet for that | 18:27 |
| aelkner | and ccarey working with ignas on packaging is also great | 18:27 |
| * pcardune hears crickets chirping | 18:28 | |
| aelkner | don't get the reference | 18:28 |
| aelkner | it's moring? | 18:28 |
| aelkner | are you in thailand | 18:28 |
| pcardune | yes | 18:29 |
| aelkner | there aren't any crickets in the city | 18:29 |
| pcardune | lol | 18:29 |
| th1a | Yes, you see, I have nothing to say about internal CanDo developments. | 18:29 |
| pcardune | th1a gets it | 18:29 |
| th1a | In that respect, they will continue to be separate projects. | 18:29 |
| aelkner | i head you | 18:30 |
| aelkner | hear | 18:30 |
| aelkner | th1a: i forget to mention that jelkner-builout depends on dashboard | 18:40 |
| aelkner | so removing it from schooltool would be a problem | 18:41 |
| aelkner | tell you what, i'll respond to your email with my thoughts | 18:42 |
| aelkner | ignas: ayt? | 18:50 |
| mgedmin | he's arguing in the next room | 18:54 |
| aelkner | sorry i'm missing that :) | 18:56 |
| th1a | aelkner: I think the idea is to just move dashboard so there aren't different copies of it (one in CanDo one in SchoolTool). | 19:02 |
| aelkner | cando doesn't have a dashboard | 19:03 |
| aelkner | we removed it since it isn't needed now that we have an easy to use UI | 19:03 |
| aelkner | really the dashboard was a workaround | 19:03 |
| aelkner | a role-based view if you will | 19:03 |
| aelkner | i look forward to our debate on the merits of role-based views | 19:04 |
| aelkner | judging from what you said at Educon it won't be a slam-dunk | 19:04 |
| aelkner | but an intersting debate when all parties converge on pycon | 19:04 |
| aelkner | it's only in jelkner2007 cuase there's no role-based view otherwise, and jelkner desparately wanted one | 19:05 |
| aelkner | he loves signing in as the teacher he is | 19:05 |
| th1a | There is no debate, really. | 19:05 |
| aelkner | and going right to a view of all his gradebooks | 19:05 |
| th1a | We just have to figure out how to do it without role-based views. | 19:06 |
| aelkner | i don't know how you envision one without the other | 19:07 |
| th1a | By "role-based view" I mean splitting off "index-teacher.html" and "index-student.html" | 19:07 |
| aelkner | oh no, that's not what we did | 19:08 |
| aelkner | we just had index.html behave based on the role | 19:08 |
| th1a | Well, don't call it that, then. | 19:08 |
| th1a | It is the same view. | 19:08 |
| aelkner | a rose by any other name... | 19:08 |
| aelkner | we just need another term | 19:09 |
| aelkner | how about role-reacting views | 19:09 |
| aelkner | i.e., the view reacts to the role of the user | 19:09 |
| th1a | How about "better views." | 19:09 |
| aelkner | and behaves accordingly | 19:09 |
| aelkner | better is too general | 19:09 |
| aelkner | the role does effect the view | 19:09 |
| th1a | I think we're all actually in agreement. | 19:10 |
| aelkner | role-sensitive views | 19:10 |
| aelkner | ? | 19:10 |
| pcardune | i like "better views" myself | 19:10 |
| aelkner | the idea being, all views should display that which the user would want to see | 19:10 |
| th1a | Less-crappy views are needed. | 19:10 |
| aelkner | mo better views | 19:10 |
| th1a | aelkner: Your point of view implies that there is a strategy in the current views. | 19:10 |
| aelkner | not implying, wishing | 19:11 |
| th1a | Nobody is arguing for, say, links to actions that you can't perform. Those are just bugs. We just need better views. | 19:11 |
| aelkner | enough said. that's what we all want | 19:12 |
| th1a | OK. | 19:12 |
| th1a | The confusion is partly because before aelkner got here we discussed entirely different views for different roles. | 19:13 |
| th1a | And I don't really know how you implemented what CanDo currently does. | 19:13 |
| aelkner | briefly put | 19:13 |
| aelkner | we created our "go" views as Stephan named them | 19:14 |
| aelkner | to behave differently depnding on the role of the user | 19:14 |
| aelkner | so the same index.html would display different links for students and teachers | 19:14 |
| aelkner | and it made the problem of "what's this for" all over the place go away | 19:15 |
| th1a | Yes... we are all in agreement. | 19:16 |
| aelkner | can't wait to see schooltool behaving like that | 19:17 |
| *** wbrady has joined #schooltool | 19:23 | |
| *** pcarduner has joined #schooltool | 19:26 | |
| *** pcarduner is now known as pcardune_ | 19:26 | |
| *** wbrady has quit IRC | 19:41 | |
| ignas | th1a: emm, not really, when I said that we could/should settle on roles, i meant that we'd have views for specific roles, like go.html | 19:47 |
| ignas | and go.html is the example of a role based view | 19:47 |
| ignas | because it riggidly dispatches on roles | 19:47 |
| ignas | not on capabilities | 19:47 |
| ignas | from what i have seen it goes like: if student: do stuff elif teacher: do other stuff | 19:47 |
| ignas | which is not extensible at all, because adding one more role makes you change the code | 19:48 |
| ignas | as for the dashboard | 19:49 |
| ignas | having it in separate egg | 19:49 |
| ignas | makes you add 'schooltool.dashboard' in your jelkner2007 setup.py | 19:49 |
| ignas | and you can have it, while the rest of schooltool just keeps on working without it | 19:49 |
| ignas | dashboard is very very rigidly defined, so it makes schooltool depend on schooltoo.gradebook, though schooltool is not using gradebook in any of it's default setups | 19:50 |
| ignas | if you have an "if" statement that decides what is being shown based on whether you are a teacher or a student and not based on "whether you have sections to instruct or you are a member of a section" it's a role specific view | 19:52 |
| ignas | aelkner: as for your question - no you don't need a custom update action, form library adapts the object to the interface | 20:03 |
| ignas | before assigning the values | 20:03 |
| ignas | so if field comes from IDemographics | 20:03 |
| ignas | before asigning the value the object will get adapted to IDemographics | 20:03 |
| ignas | if field comes from IPerson - the object will get adapted to IPerson | 20:03 |
| aelkner | that's great | 20:04 |
| aelkner | i love the way formlib wirks that way | 20:04 |
| ignas | kind of nice | 20:06 |
| ignas | but when you want a very custom form | 20:06 |
| ignas | and define a separate interface for all the fields | 20:08 |
| ignas | you have to define an adapter to that interface | 20:09 |
| ignas | even though your object has all the attributes | 20:09 |
| aelkner | why would you want a separate interface for all fields? | 20:10 |
| aelkner | isn't an intercace a colleciton of fields? | 20:10 |
| aelkner | so maybe you have a couple of interfaces | 20:10 |
| aelkner | to break the object up into chunks | 20:10 |
| ignas | yes it is, but when you want custom order with custom titles while skipping some fields | 20:10 |
| aelkner | but a separate interface for each field? | 20:10 |
| ignas | no not for each field | 20:10 |
| ignas | 1 interface for the new form | 20:11 |
| aelkner | all not each, my mistake | 20:11 |
| ignas | like if you want a formlib form to edit half of the person attributes, with some of the fields having different titles and constraints | 20:11 |
| th1a | ignas: Yes, you're right. I don't want the viewlets to be strictly based on global roles. | 21:02 |
| aelkner | ignas: a question about formlib | 21:09 |
| ignas | yes | 21:10 |
| aelkner | i'm getting this when trying to call up the edit form: | 21:10 |
| aelkner | ComponentLookupError: ((<zope.schema._bootstrapfields.Field object at 0x8890fcc>, <zc.resourcelibrary.publication.Request instance URL=http://localhost:7080/persons/manager/@@edit.html>), <InterfaceClass zope.app.form.interfaces.IInputWidget>, u'') | 21:10 |
| aelkner | Is there something i ned to register with widgets? | 21:10 |
| aelkner | does this look like anything you've gotten before? | 21:11 |
| ignas | never seen it | 21:11 |
| ignas | you might want to ask in Zope3-dev | 21:11 |
| aelkner | ok | 21:12 |
| mgedmin | what's the schema like? | 21:12 |
| mgedmin | could be you're trying to have an edit form on something that inherits IContained | 21:13 |
| mgedmin | and formlib is trying to create widgets for the __name__ and __parent__ fields | 21:13 |
| mgedmin | which is not going to work | 21:13 |
| mgedmin | the error message is bad | 21:13 |
| aelkner | that sounds like a good place to look | 21:13 |
| aelkner | i could pdb the lookup to see | 21:13 |
| aelkner | yes, it is contained | 21:14 |
| aelkner | so maybe i need to remove those fields | 21:14 |
| mgedmin | yep | 21:14 |
| mgedmin | form_fields = form.Fields(...).omit('__name__', '__parent__') | 21:14 |
| aelkner | right | 21:14 |
| aelkner | but basicperson doesn't do that | 21:14 |
| mgedmin | or have a separate schema for meaningful stuff and then implement both it an IContained | 21:14 |
| aelkner | and it works | 21:14 |
| mgedmin | or explicity select which fields you want to have in the form | 21:15 |
| ignas | aelkner: basicperson has a separate interface for a schema | 21:15 |
| ignas | i think | 21:15 |
| aelkner | oh yeah, that sounds familiar | 21:15 |
| aelkner | i'll try omitting those fields | 21:15 |
| ignas | oh IBasicPerson does not inherit IContained | 21:15 |
| aelkner | right | 21:17 |
| aelkner | i don't need as much interfaces as basicperson objects have on them, so I prefer to just derive from IContained | 21:18 |
| aelkner | yay, that did the trick | 21:19 |
| ignas | emm - IBasicPerson inherits from Interface actually ;) | 21:28 |
| aelkner | yeah, and you have lots of other interfaces for reading and writing to the person | 21:37 |
| aelkner | and IContained gets included in there | 21:38 |
| aelkner | i just wanted to keep it as simple as i feel my needs reflect | 21:38 |
| *** ignas has quit IRC | 22:05 | |
| *** mgedmin has quit IRC | 22:08 | |
| *** alga has quit IRC | 22:16 | |
| *** wbrady has joined #schooltool | 22:47 | |
| *** pcardune has quit IRC | 23:05 | |
| *** wbrady has quit IRC | 23:22 | |
| *** pcardune_ has quit IRC | 23:26 | |
Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!