| *** ignas has joined #schooltool | 01:17 | |
| *** ignas has quit IRC | 01:18 | |
| *** SteveA has quit IRC | 02:02 | |
| *** SteveA has joined #schooltool | 02:07 | |
| *** dudanogueira has joined #schooltool | 03:10 | |
| dudanogueira | hello there! i had just installed the schooltool, and dont find where to admin the accounts | 03:10 |
|---|---|---|
| *** didymo has joined #schooltool | 03:14 | |
| *** wrobel has quit IRC | 03:20 | |
| dudanogueira | please, someone help-me | 03:37 |
| dudanogueira | Aim2, didymo vidasp ??? | 03:38 |
| th1a | dudanogueira: Hi. | 04:26 |
| *** dudanogueira has quit IRC | 07:35 | |
| *** wrobel has joined #schooltool | 09:33 | |
| *** jfroche has joined #schooltool | 10:28 | |
| *** thisfred has joined #schooltool | 12:27 | |
| *** mgedmin has joined #schooltool | 13:10 | |
| *** mgedmin has quit IRC | 13:43 | |
| *** Aiste has quit IRC | 15:09 | |
| *** ignas has joined #schooltool | 15:15 | |
| *** ignas has quit IRC | 15:34 | |
| *** SteveA has quit IRC | 15:35 | |
| *** gintas has joined #schooltool | 15:40 | |
| gintas | jfroche: ping | 15:41 |
| jfroche | hello | 15:41 |
| gintas | Hi, I'm a co-worker of Ignas. | 15:41 |
| gintas | I'm currently working on a new version of zc.datetimewidget, which will be used in SchoolTool, as far as I understand. | 15:42 |
| jfroche | ah great, fixing the first day of the week problem ? | 15:42 |
| gintas | Didn't get to that part yet. | 15:43 |
| gintas | Ignas said that the French translation was incomplete. | 15:43 |
| gintas | Is that correct? | 15:43 |
| *** alga has joined #SchoolTool | 15:43 | |
| *** ignas has joined #schooltool | 15:43 | |
| jfroche | uhm i had trouble with the utf8 in the translation | 15:43 |
| jfroche | and the header of the translation wasn't correct | 15:43 |
| gintas | I fixed the problems with encoding. | 15:44 |
| gintas | (well, for the most part) | 15:44 |
| gintas | What do you mean by the header of the translation? | 15:44 |
| jfroche | name of the translation, encoding name ... | 15:46 |
| gintas | Ah. That doesn't really seem to matter. | 15:47 |
| gintas | I'll try to push my branch out to trunk soon, so you should get the improvements. | 15:47 |
| jfroche | thanks, ping me when i can try it | 15:50 |
| jfroche | please | 15:50 |
| gintas | jfroche: I also fixed the 1st day of week problem | 16:07 |
| jfroche | great ! thanks | 16:07 |
| gintas | jfroche: I just pushed out the widget to trunk | 17:13 |
| gintas | There seems to be one problem with French l10n in there: Wednesday is 'Mar'. | 17:13 |
| gintas | It should probably be 'Mec' or 'Mer'. | 17:13 |
| jfroche | gintas: ok i ll try it out, are you making a new release/new egg ? | 17:13 |
| gintas | I don't know how to do those. | 17:14 |
| gintas | I just updated the Subversion trunk, that's all. | 17:14 |
| gintas | jfroche: should it be Mec or Mer? | 17:14 |
| gintas | I'd fix it right away. | 17:14 |
| gintas | jfroche: try 'svn update' on the widget if you can. | 17:15 |
| jfroche | It's Mer | 17:15 |
| gintas | OK. | 17:15 |
| gintas | Fixed. | 17:16 |
| gintas | If you find any more problems, let me know. | 17:16 |
| jfroche | ok thanks | 17:19 |
| jfroche | gintas: thanks it is perfectly working for me | 17:25 |
| gintas | jfroche: great | 17:25 |
| *** aelkner has joined #schooltool | 17:57 | |
| aelkner | th1a: how's itgoing? | 17:57 |
| *** wdickers has joined #schooltool | 17:59 | |
| wdickers | good morning | 17:59 |
| aelkner | oh, there you are. i was looking for you on gmail chat. | 18:04 |
| aelkner | have you been looking at Tom's progress? | 18:05 |
| *** mgedmin has joined #schooltool | 18:06 | |
| wdickers | Not yet, I'm checking it out now | 18:06 |
| aelkner | I just did the svn up to maddog's pyagentlib, and it brought down a lot of stuff. | 18:07 |
| wdickers | yeah, same here. I'm looking at his new request class | 18:08 |
| *** gintas has quit IRC | 18:08 | |
| wdickers | But what I would like to know is what place do I have in all this? | 18:10 |
| aelkner | That's a good question. We have a need to reach an end goal which is to have scholtool talking through the SIF. | 18:11 |
| th1a | Hi guys. | 18:12 |
| aelkner | You need to demonstate this for your project, right? | 18:12 |
| aelkner | Hi Tom. | 18:12 |
| wdickers | Yes, however it is difficult for me to work on a schooltool page if I don't know how the agents communicate using the ZODB | 18:12 |
| wdickers | hello Tom | 18:12 |
| th1a | Don't worry, there is still plenty to be done. | 18:12 |
| th1a | OK. Let me catch you guys up. | 18:12 |
| aelkner | Tom, I see your request class does what our messageTools intened to do. | 18:13 |
| th1a | Pretty much, yeah. | 18:13 |
| aelkner | Why didn't you reduce the code size by uiseing buildElement as we had done? | 18:13 |
| aelkner | That's two lines down to one for each element. | 18:13 |
| th1a | Look at testAgent.py | 18:13 |
| th1a | Um... I'm stubborn? | 18:13 |
| aelkner | Could be. | 18:14 |
| aelkner | My feeling is always, the less you have to look at, the faster you can absorb. | 18:14 |
| th1a | I'll take another look at how you did it. | 18:14 |
| aelkner | Cool. | 18:14 |
| wdickers | I agree with aelkner. If we ever need to change how we create elements, it would be easier with buildElement | 18:14 |
| aelkner | Anyway, looking at testAgent.puy | 18:14 |
| aelkner | We need to look at the Agent base class, too. | 18:15 |
| th1a | OK... well, the idea is you don't, really. | 18:15 |
| th1a | Or, not much. | 18:16 |
| aelkner | Only because TestAgent calls into it. | 18:16 |
| th1a | the register method is the most complete example. | 18:16 |
| wdickers | I see you've changed how exam.py is organized | 18:16 |
| th1a | exam.py just makes a stack of responses which are in the correct sequence for the doctest. | 18:17 |
| th1a | So if you change the order of either, they break. | 18:17 |
| th1a | But it is the easiest way to simulate responses. | 18:17 |
| wdickers | what is the sif_exceptions.pyagentlib_Error | 18:18 |
| th1a | That's the class of internal pyagentlib errors. | 18:18 |
| th1a | So that catches any subclassed pyagentlib error. | 18:19 |
| th1a | And the other one catches any SIF error. | 18:19 |
| wdickers | could you give an example? | 18:19 |
| th1a | If you go to agent.py... | 18:19 |
| th1a | line 73 raises a pyagentlib error. | 18:20 |
| th1a | Because the schema validation has caught an invalid outgoing message created by pyagentlib. | 18:20 |
| wdickers | ah, and then that raised error gets caught in the test | 18:20 |
| th1a | Also, if you want your SchoolTool agent to do something when an error occurs, you'll need to do the same kind of thing. | 18:21 |
| th1a | Log the error, etc. | 18:21 |
| th1a | Actually, the errors will log themselves if you set up a logger. | 18:22 |
| aelkner | Where does that happen? | 18:23 |
| th1a | So a lot of the TestAgent methods don't really do much of anything from SchoolTool's point of view. | 18:23 |
| th1a | Um... I would be looking at TinyZIS to remember how it works, but you basically just have to instantiate a logger object. | 18:23 |
| aelkner | I mean where is the exception caught? | 18:24 |
| th1a | The exception is caught in your Agent subclass, like in register() | 18:24 |
| th1a | like register() in TestAgent. | 18:25 |
| aelkner | I see. | 18:25 |
| aelkner | And the creation of the exception object itself lead to logging, right? | 18:25 |
| wdickers | http://www.red-dove.com/python_logging.html jere | 18:26 |
| th1a | Right. There is logging built into the exception classes. | 18:26 |
| wdickers | *here's the tutorial that I found on logging.py | 18:26 |
| th1a | If you look at the SIF_Error base class. | 18:26 |
| th1a | in sif_exceptions. | 18:26 |
| aelkner | I see it. | 18:27 |
| aelkner | In addition to logging, the erived classes populate them selves so that anyone catching them can use the info. | 18:28 |
| aelkner | I remember you passing these objects around in tinyzis. | 18:28 |
| aelkner | Will, remember that? | 18:28 |
| wdickers | why is the logging only in the sif errors? | 18:28 |
| aelkner | You had asked about that. | 18:28 |
| wdickers | The logging? I thought I asked about the conf file | 18:28 |
| wdickers | Or is the logs stored there as well? | 18:29 |
| aelkner | No, once you asked, "what is this ex parm in the makeAck" | 18:29 |
| th1a | The logs themselves aren't anywhere yet, but it will be set in the conf. | 18:29 |
| th1a | The log output goes to the console right now. | 18:29 |
| aelkner | We don't need to worry about logging right now. | 18:30 |
| aelkner | The console is fine. | 18:30 |
| th1a | So there are a couple places to start on the SchoolTool end. | 18:30 |
| wdickers | Oh yeah, I forgot about that | 18:30 |
| aelkner | I'm all ears. | 18:30 |
| th1a | We need to learn how to hook up SchoolTool to ZEO, which is easy, if ignas tells us which three lines of code to change. | 18:31 |
| th1a | ignas: ayt? | 18:31 |
| th1a | So we pretty much will just have to nag him for that info... if we try to do it we'll waste hours. | 18:31 |
| th1a | Other starting point: | 18:32 |
| wdickers | What is the ZEO? | 18:32 |
| th1a | ZEO is Zope Enterprise Objects, which basically allows you to simultaneously connect two applications to a ZODB instance. | 18:32 |
| th1a | So the agent and SchoolTool will be able to share the DB. | 18:32 |
| aelkner | Will, let's worry about that later. I want to hear Tom's high-level plan. | 18:32 |
| wdickers | gotcha | 18:32 |
| th1a | It is good to know what it means ;-) | 18:33 |
| th1a | Now, moving on... | 18:33 |
| th1a | One thing you can do is create an object in the ZODB that represents the state of the agent, | 18:33 |
| th1a | or, more precisely, the state of its relationship to the ZIS. | 18:33 |
| th1a | For example: registered or not registered with the ZIS. | 18:33 |
| th1a | So that, it would, for example, start out unregistered by default. | 18:34 |
| th1a | When you run register() successfully, the agent switches it to True (registered). | 18:34 |
| th1a | When you run unregister successfully it switches it to False. | 18:34 |
| wdickers | Wouldn't it be simpler to have a class for the relationship? It could have that boolean flag as well as contain the queue of messages | 18:35 |
| th1a | Not sure what you mean. | 18:35 |
| th1a | There are several more bits of data that will need to be tracked. | 18:36 |
| th1a | Which objects are subscribed to, which are published, etc. | 18:36 |
| wdickers | I mean have an object called AgentSchooltool or something that can contain all the necessary data instead of several unconnected variables | 18:36 |
| th1a | wdickers: I think that's what I had in mind. | 18:37 |
| aelkner | I think so, too. | 18:37 |
| aelkner | So AgentSchool | 18:37 |
| aelkner | derives from Agent and Persist, right? | 18:37 |
| th1a | No. | 18:37 |
| th1a | Er. | 18:38 |
| th1a | Maybe. | 18:38 |
| th1a | I guess it does. | 18:38 |
| aelkner | I thought that was our goal. | 18:38 |
| th1a | I have to put on my ZODB hat. | 18:38 |
| th1a | OK, here's what I'm thinking. | 18:38 |
| th1a | The Persistent part is a SchoolTool object. | 18:39 |
| th1a | The Agent opens the ZODB and manipulates that Persistent object. | 18:39 |
| th1a | But perhaps you're right. | 18:39 |
| th1a | The agent IS the persistent object. | 18:40 |
| aelkner | No, I'm sorry. Forgot the difference between stanalone agent and schooltool. | 18:40 |
| aelkner | t | 18:40 |
| th1a | OK. | 18:40 |
| aelkner | The agent runs as a standalone app. | 18:40 |
| th1a | Yes. | 18:40 |
| aelkner | If it is a schooltool agent, it doesn't need to be derived from Persist. | 18:40 |
| aelkner | It just needs to look in the ZODB. | 18:40 |
| aelkner | Using ZEO? | 18:41 |
| th1a | Right. | 18:41 |
| wdickers | Sorry guys, I gotta go. Thanks | 18:41 |
| aelkner | Wiat. | 18:41 |
| aelkner | Will, can you meet tonight. | 18:41 |
| *** wdickers has quit IRC | 18:41 | |
| aelkner | Crap, he's one. | 18:41 |
| aelkner | This was going well. | 18:41 |
| aelkner | gone. | 18:41 |
| aelkner | Tom, Will expressed concern over where he fits in. | 18:42 |
| th1a | That's the bell schedule for you. | 18:42 |
| aelkner | I thought we were dong right with this mindshare for starters. | 18:42 |
| th1a | Well, I think it was best to let me hump over the base class. | 18:43 |
| aelkner | Then he and I will eventually sit down to do something together. | 18:43 |
| aelkner | I agree. | 18:43 |
| th1a | I think it is more clear now what needs to be built on it. | 18:43 |
| aelkner | I suppose so. | 18:43 |
| aelkner | Could we develop AgentSchoolTool with tests without having a schooltool instance running? | 18:44 |
| th1a | One thing that we/you could do, pending a ZEO lesson, is make a simple agent that persists changes into its own ZODB. | 18:44 |
| aelkner | Is that the answer to my question? | 18:44 |
| th1a | Yes. That's pretty much the same idea as I just said. | 18:44 |
| th1a | No, it wasn't an answer ;-) | 18:45 |
| th1a | We were just thinking the same thing. | 18:45 |
| aelkner | Well, sort of. | 18:45 |
| *** ignas has quit IRC | 18:45 | |
| aelkner | About the tests: | 18:45 |
| aelkner | Right now you have doctests.txt for the base class? | 18:45 |
| aelkner | No the TestAgent. | 18:46 |
| th1a | It tests testAgent.py specifically. | 18:46 |
| aelkner | Could we have the filenames of the tests match the pytion files for simplicity? | 18:46 |
| th1a | One thing that's important is that since the incoming and outgoing messages are all tested against the XML schema, that itself is a pretty rigorous test for what I'm writing. | 18:46 |
| th1a | Actually, I suppose that's ok in this case. | 18:47 |
| th1a | in TinyZIS it pretty much had to be one big file because all the tests depend on each other. | 18:47 |
| aelkner | I mean, testAgent.txt, agentSchoolTool, txt, etc. | 18:47 |
| aelkner | .txt | 18:47 |
| th1a | Well, your schooltool agent should be in its own module. | 18:48 |
| th1a | Er. | 18:48 |
| th1a | Package. | 18:48 |
| aelkner | Own directory? | 18:48 |
| th1a | Yes. | 18:48 |
| th1a | And actually, I'm over halfway done with the base class. | 18:48 |
| th1a | It isn't going to do a heck of a lot. | 18:48 |
| aelkner | Yes, I think you were right to kick me off that effort. Your steaming along :) | 18:48 |
| th1a | I've got a kind of weird style on this, but it works. | 18:49 |
| aelkner | When you are done, or close to done, we could build the other package that uses the base package, right? | 18:49 |
| th1a | Build the SchoolToolAgent package? | 18:50 |
| aelkner | That's wat I meant. | 18:50 |
| th1a | We should start that right now. | 18:50 |
| aelkner | An get ignas to help wth the ZEO, right? | 18:50 |
| th1a | Yes. I'll send him an email. | 18:51 |
| aelkner | Ok, for Will's part, he should keep meting us to keep abreast of the poject. | 18:52 |
| aelkner | I feel responsible for his ongoing progress as he has a presentation to give some time in Feb I believe. | 18:53 |
| aelkner | Thing is I need to catch up myself. | 18:53 |
| aelkner | Could you be available by 11:00 am tomorrow to continue our three-way chat? | 18:54 |
| th1a | I should be here. | 18:54 |
| aelkner | Godd. I have to go to lunch now. Maybe I'll see you later this after noon. | 18:54 |
| th1a | OK. | 18:54 |
| *** aelkner has quit IRC | 18:55 | |
| *** thisfred has quit IRC | 20:51 | |
| *** gintas has joined #schooltool | 21:35 | |
| *** ignas has joined #schooltool | 21:37 | |
| *** th1a has quit IRC | 21:46 | |
| *** mgedmin has quit IRC | 21:58 | |
| *** povbot has joined #schooltool | 21:58 | |
| *** ignas has quit IRC | 21:58 | |
| *** gintas has quit IRC | 22:03 | |
| *** jfroche has quit IRC | 23:10 | |
| *** SteveA has joined #schooltool | 23:31 | |
| *** povbot` has joined #schooltool | 23:48 | |
| *** alga has quit IRC | 23:52 | |
Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!