| *** bskahan has quit IRC | 00:53 | |
| *** mgedmin has joined #schooltool | 01:36 | |
| mgedmin | I think there's a bug in the Session product | 01:36 |
|---|---|---|
| mgedmin | we've moved schooltool.org to new hardware last week | 01:36 |
| mgedmin | it is quite unlikely that the new server is flaky in the same way as the old one | 01:37 |
| * mgedmin goes to restart zope on schooltool.org | 01:37 | |
| tvon | IIRC that bug was supposed to be fixed the latest zope 2... not sure though | 01:42 |
| mgedmin | the bug, if there is a bug, has got something to do with acquisition | 01:43 |
| mgedmin | the last line in the traceback is | 01:43 |
| mgedmin | the last line referred to in the traceback is | 01:43 |
| mgedmin | if getattr(ob, '__of__', None) and getattr(ob, 'aq_parent', None): | 01:43 |
| mgedmin | and it raises SystemError: error exit with no error set | 01:43 |
| tvon | ah, not the error I was thinking about | 01:43 |
| mgedmin | what error were you thinking about? | 01:44 |
| tvon | the key part of the error said something about SessionObject or something along those lines | 01:45 |
| tvon | I'd get it at random on a few sites..restarting always got rid of it | 01:45 |
| * mgedmin checks | 01:45 | |
| mgedmin | schooltool.org is running on 2.7.1 | 01:45 |
| mgedmin | zope.org says the latest stable release is 2.7.3 | 01:45 |
| mgedmin | I always assumed it was an indication of a hardware problem | 01:47 |
| mgedmin | random memory corruption or some such | 01:48 |
| mgedmin | it looks like I was wrong | 01:48 |
| * mgedmin upgraded schooltool.org to zope 2.7.3 | 01:52 | |
| mgedmin | let's see if that error reappears | 01:52 |
| tvon | heh, its a suck error since you basically just have to wait to not see it | 01:52 |
| mgedmin | yes :( | 01:53 |
| mgedmin | ok, good night | 01:53 |
| *** mgedmin has quit IRC | 01:53 | |
| tvon | night | 01:54 |
| *** d2m has quit IRC | 02:05 | |
| *** tvon has quit IRC | 03:09 | |
| *** tvon has joined #schooltool | 06:20 | |
| *** tvon has left #schooltool | 06:55 | |
| *** d2m has joined #schooltool | 09:14 | |
| *** mgedmin has joined #schooltool | 10:42 | |
| *** mgedmin has quit IRC | 12:05 | |
| *** mgedmin has joined #schooltool | 12:07 | |
| *** alga has joined #SchoolTool | 12:18 | |
| *** alga has quit IRC | 12:55 | |
| *** mgedmin has quit IRC | 13:10 | |
| *** d2m has quit IRC | 14:23 | |
| *** mgedmin has joined #schooltool | 14:34 | |
| *** jinty has joined #schooltool | 15:36 | |
| *** tvon has joined #schooltool | 15:54 | |
| *** SteveA has joined #schooltool | 16:10 | |
| *** bskahan has joined #schooltool | 17:50 | |
| bskahan | Hi all | 17:51 |
| th1a | Hello. | 17:51 |
| tvon | Howdy | 17:53 |
| mgedmin | hi | 17:54 |
| jinty | hey | 17:54 |
| th1a | OK, let's get started. | 17:57 |
| *** alga has joined #SchoolTool | 17:58 | |
| th1a | First off, one thing I always forget to explicitly ask: Now that you're done with one contract, when will you be ready to start the next? | 17:58 |
| *** gintas has joined #schooltool | 17:59 | |
| bskahan | I can start now, tvon can start next monday | 17:59 |
| mgedmin | next monday | 18:00 |
| mgedmin | (I think) | 18:01 |
| th1a | OK. | 18:01 |
| th1a | Let's talk about what you'll be doing then. | 18:01 |
| mgedmin | our closest goal is SchoolBell 1.0 by March, right? | 18:02 |
| th1a | Right. | 18:02 |
| mgedmin | my vision for that release is: | 18:02 |
| bskahan | there's a schooltool target for march/april as well, right? | 18:02 |
| th1a | Yes. | 18:02 |
| mgedmin | - missing features that are necessary for a calendaring app (timezones, all-day events) | 18:03 |
| th1a | "SchoolTool Calendar" | 18:03 |
| mgedmin | (I mean, we should have those features that are missing now) | 18:03 |
| mgedmin | - easy to deploy (apt-get install) and use (slick UI) | 18:03 |
| th1a | SteveA expressed some concern about opening up a can of worms with timezones. | 18:04 |
| tvon | I think it is a can of worms, but we need to open it. | 18:04 |
| mgedmin | let's talk about timezones a bit later (I also think we need to open it) | 18:04 |
| mgedmin | - source code separated from schooltool into reusable Zope 3 components | 18:04 |
| th1a | Right. | 18:04 |
| alga | I think what we have by now is much more of a can of worms | 18:04 |
| alga | timetables, exceptions, calendars, compositions | 18:05 |
| mgedmin | - commitment to never break databases on upgrades | 18:05 |
| jinty | please | 18:05 |
| tvon | we need to decide on how we will handle migrations | 18:05 |
| mgedmin | zope 3 generations, probably | 18:06 |
| alga | and lots of discipline ;-) | 18:06 |
| mgedmin | http://dev.zope.org/Wikis/DevSite/Projects/ComponentArchitecture/DatabaseGenerations | 18:06 |
| tvon | is there an intention to drop the current "search" method in the code and use the z3 catalog? | 18:06 |
| mgedmin | (shorter link: http://dev.zope.org/Zope3/DatabaseGenerations ) | 18:06 |
| bskahan | an addition to the list of calendar features - event descriptions | 18:06 |
| tvon | oh yes | 18:06 |
| mgedmin | is there a wiki page "goals for schoolbell 1.0"? | 18:07 |
| mgedmin | if not, could we create it? | 18:07 |
| th1a | Yes. | 18:07 |
| bskahan | good idea | 18:07 |
| mgedmin | I think event descriptins fall under my "missing features" | 18:07 |
| * bskahan nods | 18:08 | |
| th1a | I'll make the wiki page. | 18:08 |
| tvon | mgedmin: did you catch my catalog question? | 18:08 |
| alga | do you mean busysearch? | 18:09 |
| mgedmin | we did not have intentions to start using z3 catalog | 18:09 |
| tvon | let me check | 18:09 |
| mgedmin | "first make it work, then make it fast" | 18:09 |
| * tvon nods | 18:09 | |
| mgedmin | here's an extra goal for schoolbell 1.0: | 18:09 |
| mgedmin | - make it reasonably fast | 18:09 |
| gintas | isn't it? | 18:10 |
| th1a | http://www.schooltool.org/bounties/schoolbell/OnePointOh | 18:10 |
| mgedmin | this means: write a unit or functional test with a large amount of data, see if it works fast enough | 18:10 |
| mgedmin | I do not know if it is or if it isn't, that's why we need some large calendars | 18:10 |
| tvon | we need some generated calendars | 18:10 |
| gintas | I would call that 'write benchmark tests' | 18:10 |
| mgedmin | irclogs2ical.py anyone? | 18:11 |
| gintas | :) | 18:11 |
| tvon | ooh | 18:11 |
| mgedmin | my irclogs2html.py has a LogReader class that gives you tuples (timestamp, event_type, text_or_something_else) | 18:11 |
| mgedmin | http://mg.pov.lt/blog/irc-logs.html | 18:11 |
| tvon | thats obscene..would make a great test | 18:12 |
| mgedmin | let's not get too obsessed with speed | 18:12 |
| mgedmin | if icalendar clients (sunbird, evolution, ical) are unbearably slow for a calendar, I think schoolbell can afford to be slow too | 18:12 |
| tvon | yeah | 18:13 |
| tvon | where does the gradebook fall into this? | 18:13 |
| gintas | well, the difference is that there is usually one client on a computer | 18:13 |
| mgedmin | schoolbell will have no gradebooks, I think | 18:14 |
| gintas | with schoolbell, we might have tens of simultaneous clients | 18:14 |
| tvon | st I mean | 18:14 |
| gintas | (I meant to say simultaneous connections) | 18:14 |
| th1a | I don't think speed is a concern for SB 1.0 or SchoolTool Calendar. | 18:15 |
| mgedmin | th1a: http://www.schooltool.org/bounties/schoolbell/OnePointOh gives me 404 not found | 18:15 |
| th1a | I just moved it... | 18:15 |
| th1a | It is at http://www.schooltool.org/bounties/schoolbell/1-0/OnePointOh/ | 18:15 |
| gintas | th1a, I think mgedmin meant to say that we don't really know how fast schoolbell is, and it's important to find that out before deployment | 18:15 |
| * bskahan agrees | 18:16 | |
| th1a | Yes. | 18:16 |
| th1a | OK, so I've got: * TimeZones | 18:18 |
| th1a | * RefactorNotes | 18:18 |
| th1a | * Integrate EventAndBooking form | 18:18 |
| th1a | * EuroUS calendar styles | 18:18 |
| th1a | * AllDayEvents | 18:18 |
| th1a | * EventDescriptions | 18:18 |
| th1a | Up for discussion. | 18:18 |
| mgedmin | this is a list of must-haves for 1.0, right? | 18:18 |
| tvon | yeah | 18:19 |
| th1a | Possibilities. | 18:19 |
| tvon | oh | 18:19 |
| th1a | Plus the Z3 work. | 18:19 |
| bskahan | /*/*/calendar.vfb? | 18:19 |
| tvon | yeah | 18:20 |
| th1a | ? | 18:20 |
| bskahan | what is EuroUS calendar styles? | 18:20 |
| tvon | free-busy calendar | 18:20 |
| bskahan | for outlook/evolution | 18:20 |
| th1a | AM/PM vs 24 hr. Weeks starting on Sunday, etc. | 18:20 |
| bskahan | since outlook can't read ical, and evo can't write it | 18:20 |
| * bskahan nods | 18:20 | |
| th1a | What are you guys talking about? | 18:20 |
| bskahan | VFreeBusy, is the published free/busy format outlook uses | 18:21 |
| th1a | Ah. | 18:21 |
| bskahan | for scheduling meetings, etc | 18:21 |
| th1a | We'd just have to publish it? | 18:22 |
| bskahan | yes | 18:22 |
| th1a | How difficult would that be? | 18:22 |
| bskahan | depends on how closely it overlaps ical | 18:22 |
| mgedmin | easy, I think | 18:22 |
| alga | A simplified subclass of our ical view | 18:23 |
| mgedmin | how should it be published? | 18:23 |
| gintas | I don't think that it would be hard, but I'm only vaguely familiar with that format | 18:23 |
| mgedmin | a http url that outlook can download? | 18:23 |
| * mgedmin never used outlook | 18:23 | |
| gintas | mgedmin, yes | 18:23 |
| gintas | well, I'm guessing | 18:23 |
| bskahan | i think calendar.vfb vs. calendar.ics would work | 18:23 |
| gintas | I saw the same feature on KOrganizer | 18:23 |
| th1a | If we could do it in two days it would be ok, I guess. | 18:24 |
| gintas | I think two days is reasonable | 18:24 |
| mgedmin | th1a, could you add "database forwards compatibility" to the list | 18:25 |
| th1a | bskahan: can you take a closer look at it and make an estimate? | 18:25 |
| * bskahan nods | 18:25 | |
| mgedmin | and also "slick user interface", "easy deployment", "reasonable speed" | 18:25 |
| mgedmin | I would also like to elaborate on each of those issues | 18:25 |
| jinty | mgedmin: anyway of testing for if the database forwards comptibility breaks? | 18:26 |
| mgedmin | perhaps I should just go and edit the wiki page myself now, and then ask for comments? | 18:26 |
| th1a | mgedmin: go ahead. | 18:26 |
| gintas | wouldn't it make sense for PoV to do the VFB export? It should be similar to iCal export, which Marius did | 18:26 |
| mgedmin | jinty, it is not trivial, but doable | 18:26 |
| tvon | yeah, POV could do it faster | 18:26 |
| th1a | That's ok by me. | 18:26 |
| jinty | mgedmin: thanks | 18:27 |
| th1a | So what are the issues around adding timezone support? Have you guys thought about this at all? | 18:27 |
| * jinty goes back to kicking chroots | 18:28 | |
| gintas | I'm thinking this involves two things: iCalendar parsing and the web interface | 18:28 |
| tvon | system should store it all in UTC, convert it based on user settings for viewing/ical... | 18:28 |
| * mgedmin wants to finish thinking about high-level issues before descending into details too deeply | 18:29 | |
| gintas | still, it would be nice to have an idea of the scope of having decent timezone support | 18:29 |
| alga | don't you? | 18:30 |
| th1a | It isn't a problem I understand well. | 18:30 |
| gintas | I, for example, don't really see right now why Steve said that timezones are a can of worms | 18:30 |
| tvon | dos ical have any representation if timezones? | 18:30 |
| gintas | yes | 18:30 |
| alga | it's just that we use naive times now | 18:30 |
| alga | and we have to switch to UTC, or time with timezones in the server | 18:31 |
| alga | and make a TZ a preference of a user | 18:31 |
| tvon | I see | 18:31 |
| tvon | ideally an anon cookie of sorts as well | 18:32 |
| th1a | Does the SchoolTool know if its box is set to UTC or local time? | 18:32 |
| tvon | it can find out I think, or at least rely on a config file setting to do the math | 18:32 |
| alga | right now all times are naive, i.e. without timezones | 18:32 |
| gintas | tvon, are you sure a cookie is the best idea? OTOH, I like mgedmin's suggestion to stay focused on the big picture until we have the most important things figured out | 18:34 |
| tvon | okay, lets come back to the tz issue | 18:34 |
| gintas | I thought along the lines of a user preference | 18:34 |
| tvon | gintas: a cookie for anonymous is what I'm saying. I don't think you should have to have an account to view public calendars in your own tz | 18:35 |
| alga | right | 18:36 |
| bskahan | mgedmin: getting the _zope_proxy_proxy error on a fresh schooltool checkout again | 18:36 |
| tvon | odd | 18:37 |
| bskahan | was working last week | 18:37 |
| th1a | I guess I need a timeline of the Z3 work that'll be in SchoolBell 1.0. | 18:38 |
| mgedmin | bskahan, run make | 18:38 |
| mgedmin | what is "Integrate EventAndBooking form" | 18:39 |
| th1a | Right now we have a "create event" link and a "book a resource link" | 18:39 |
| bskahan | from a discussion tvon and I had about how to simplify them in the UI | 18:39 |
| th1a | which is really unnecessary, since a resource booking is implicitly part of an event. | 18:40 |
| bskahan | yeah | 18:40 |
| th1a | It's not a big job. | 18:40 |
| bskahan | mgedmin: it was make, tnx | 18:41 |
| th1a | The only thing on the list which is potentially controversial/time consuming is timezones. | 18:42 |
| alga | what about migration to zope3 publisher? | 18:42 |
| th1a | I want POV to keep working on that. | 18:43 |
| th1a | We need to decide where in the Z3 transistion we need to be for SB 1.0. | 18:43 |
| th1a | I'd like SB 1.0 to be Twisted-less. | 18:44 |
| th1a | SB 1.0 should clearly be a Z3 application. | 18:45 |
| th1a | That's a high priority. | 18:45 |
| bskahan | are you shooting for an installable Zope3 component by 1.0? | 18:46 |
| th1a | No. | 18:46 |
| mgedmin | waah, somebody edited the wiki page while I was editing it | 18:46 |
| th1a | From what I understand that's not likely in the timeframe. | 18:46 |
| tvon | heh, saw that coming | 18:46 |
| th1a | I thought it locked. Don't some versions of ZWiki lock? | 18:47 |
| gintas | how is "a Z3 application" different from "an installable Zope3 component"? | 18:47 |
| gintas | do you mean by "Z3 app", that it uses parts of zope3 as a library? | 18:47 |
| bskahan | gintas: yes | 18:47 |
| bskahan | as opposed to installs into Z3 using zcml, et al | 18:48 |
| tvon | installable z3 component to me means that you can drop it into instnace/lib/python and link the zcml in to site-includes | 18:48 |
| th1a | That's what we're shooting for in the long run--an installable component. | 18:48 |
| gintas | from what I understand, Albertas' work is the leading to an installable z3 component | 18:48 |
| gintas | s/ the // | 18:48 |
| mgedmin | please take a look at the wiki page | 18:49 |
| gintas | alga, can you elaborate on your progress? | 18:49 |
| mgedmin | I tried to deriver high level goals based on my understanding of mark's wishes expressed during the london meeting | 18:49 |
| th1a | http://source.schooltool.org/svn/branches/alga-zope3-branch/zope3views.txt | 18:49 |
| mgedmin | approximately "school bell is the best calendaring thingy" | 18:50 |
| * mgedmin now updating task list | 18:50 | |
| mgedmin | s/deriver/derive/ | 18:50 |
| th1a | That looks correct. | 18:51 |
| bskahan | how much work is required to be DB neutral? | 18:52 |
| bskahan | or, what do you mean by not enforcing a storage model | 18:53 |
| mgedmin | schoolbell defines interfaces | 18:53 |
| mgedmin | and provides sample implementations | 18:53 |
| bskahan | ok | 18:53 |
| gintas | that sounds *really* complex to me | 18:53 |
| bskahan | after I re-read the sentence it made more sense | 18:53 |
| mgedmin | applications are not required to use calendar/event classes that come with schoolbell if they want to use, e.g., calendaring composition or recurrence expanding code provided in the schoolbell library | 18:53 |
| mgedmin | I started doing this in Mataro, while at the Ubuntu conference | 18:54 |
| mgedmin | it was pretty straightforward | 18:54 |
| mgedmin | gintas, what sounds complex to you? | 18:54 |
| gintas | it was a misunderstanding | 18:55 |
| mgedmin | ok | 18:55 |
| gintas | well, maybe not, but I thought that we want to be able to use something else and drop ZODB entirel | 18:55 |
| gintas | entirely | 18:55 |
| gintas | is that what you had in mind? | 18:56 |
| mgedmin | I can see how that sentence in the wiki can be confusing | 18:56 |
| mgedmin | we will have the schoolbell application | 18:56 |
| mgedmin | and schoolbell -- a collection of libraries | 18:56 |
| gintas | ah, ok | 18:56 |
| th1a | The bulk of this work needs to be done by POV. Can you guys sketch out a timeline for getting it done? | 18:56 |
| mgedmin | supporting multiple storage models in the schoolbell application is YAGNI | 18:56 |
| gintas | yes, that's what I thought | 18:56 |
| mgedmin | that sentence refers to the library | 18:56 |
| gintas | ok | 18:56 |
| th1a | YAGNI? | 18:56 |
| mgedmin | You Ain't Gonna Need It | 18:57 |
| mgedmin | popular in zope3 circles | 18:57 |
| th1a | Yes. | 18:57 |
| mgedmin | so, do we agree on the scope? | 18:57 |
| mgedmin | so, do we agree on the scope for the SchoolBell 1.0 release? | 18:57 |
| mgedmin | have I missed anything? have I included too much? | 18:57 |
| gintas | I'd like to have a bit more details on the "finish migration to Zope3" part | 18:58 |
| mgedmin | yes | 18:58 |
| th1a | I have no clue how long the Zope 3 migration work will take, so I don't know if it is too much. | 18:58 |
| gintas | we had a list somewhere, but I don't remember if we decided to do everything in that list eventually | 18:58 |
| * mgedmin is currently focusing on scope and ignoring time constraints | 18:59 | |
| th1a | The scope is good. | 18:59 |
| th1a | OK. Our hour is up... One quick question... | 19:00 |
| alga | my estimate is that migration to Zope 3 views will require ~ 48 man/days | 19:00 |
| alga | this does not include security | 19:00 |
| mgedmin | if we notice that we are short on time, we can drop off least important items from the list | 19:01 |
| th1a | Yes. | 19:01 |
| alga | yes | 19:01 |
| alga | so prioritization is important | 19:01 |
| gintas | 48 man-days for the views only? that sounds a bit excessive | 19:01 |
| mgedmin | gintas, have you looked at alga's document on view migration? | 19:01 |
| alga | see my estimates for steps in the doc | 19:01 |
| alga | it is in ideal programming time, I assumed 0.6 ratio to developer time | 19:02 |
| * mgedmin 's gut feeling is that we can release schoolbell 1.0 after a month of work, but history tells that he is always too optimistic | 19:03 | |
| th1a | We've got more than that, so it's ok. | 19:03 |
| tvon | heh, initial_estimate * 2.5 | 19:03 |
| mgedmin | th1a, what is your one quick question? | 19:03 |
| th1a | Refactoring notes. | 19:03 |
| th1a | Were we talking about Z3 annotations for that? | 19:04 |
| bskahan | no, just using facets | 19:04 |
| th1a | Oh. So we could do that now. | 19:04 |
| * bskahan nods | 19:04 | |
| tvon | oh? I thought it was annotations | 19:04 |
| mgedmin | I expect that we will do quite a bit of refactoring while splitting schooltool into independent components | 19:04 |
| th1a | I was wondering if we needed to wait before refactoring notes. | 19:05 |
| gintas | yes | 19:05 |
| mgedmin | th1a, I don't think so | 19:05 |
| th1a | OK. | 19:05 |
| bskahan | is there a string reason to use annotations over facets? | 19:05 |
| mgedmin | however it would be a very good idea to work on the trunk and coordinate closely | 19:05 |
| jinty | mgedmin: please also think of separating pictures and page templates from .py files | 19:05 |
| bskahan | s/string/strong/ | 19:05 |
| gintas | the 'yes' was to to mgedmin's note | 19:05 |
| gintas | I don't think it's a very good idea to use both annotations and facets | 19:06 |
| jinty | makes life a lot easier | 19:06 |
| gintas | they are similar in their function somewhat, right? | 19:06 |
| bskahan | gintas: that's the impression I'm under | 19:06 |
| bskahan | but I could be missing something | 19:06 |
| gintas | well, unless we plan to get rid of facets altogether, I suggest sticking with them | 19:07 |
| mgedmin | ok, let's wrap up then | 19:07 |
| tvon | I thought facets were "hard core" annotations, but I dont know much about z3 specifics that we arent using | 19:07 |
| mgedmin | otherwise we will keep talking all night | 19:07 |
| tvon | yeah | 19:07 |
| th1a | Good meeting, folks. | 19:07 |
| th1a | I guess next Monday we'll be ready to discuss our work plans for the week. | 19:08 |
| mgedmin | parting thought on facets and annotations: both are fine, annotations are more lightweight, facets are more domain-oriented; perhaps schoolbell should use only annotations while schooltool should use both | 19:08 |
| bskahan | not a bad idea | 19:08 |
| alga | are we officially done with the meeting? | 19:11 |
| th1a | Yes. You are dismissed. | 19:11 |
| * mgedmin disappears | 19:14 | |
| *** gintas has quit IRC | 19:16 | |
| jinty | th1a: on the release of 0.9? | 19:18 |
| th1a | Yes. I was just thinking we needed to discuss that. | 19:18 |
| th1a | bskahan is cleaning up a few things. | 19:18 |
| th1a | And merging into the trunk. | 19:19 |
| jinty | There is a way of having schooltool 0.9 packages the same as 0.8... I thought it up last night... but its scary. | 19:19 |
| bskahan | th1a: what do you want to change in the branch | 19:19 |
| bskahan | vs. trunk | 19:19 |
| th1a | The release is coming from the branch, I guess. | 19:20 |
| bskahan | yes | 19:20 |
| jinty | has to, the trunk is way too broken | 19:20 |
| bskahan | so for things to be in 0.9 they need to change their | 19:20 |
| th1a | OK. | 19:20 |
| jinty | from a packaing pov | 19:20 |
| alga | a packaging whom? :-) | 19:21 |
| jinty | point of view:) | 19:21 |
| alga | just kidding | 19:22 |
| jinty | what my first thought was when i saw pov | 19:22 |
| th1a | I guess you don't need to do anything else to the branch. | 19:22 |
| th1a | Hm. | 19:22 |
| * tvon goes to walk a dog | 19:22 | |
| th1a | Hm. | 19:23 |
| th1a | So the only breakage in the database we introduced was changing the name of the root group? | 19:23 |
| bskahan | there's a new attribute on the person object as well | 19:24 |
| th1a | OK. | 19:24 |
| bskahan | existing persons wouldn't have it | 19:24 |
| th1a | Right. | 19:24 |
| th1a | And that would cause an error? | 19:25 |
| bskahan | yes | 19:27 |
| th1a | OK. | 19:27 |
| bskahan | when the calendar went to look up colors for overlays | 19:28 |
| th1a | Yeah. | 19:28 |
| * jinty thinks a lot of false bug reports | 19:30 | |
| jinty | is this something that should be reverted for 0.9? | 19:31 |
| th1a | No, I don't think so. | 19:31 |
| jinty | ok, then anything else? | 19:32 |
| th1a | So if I write a script that pulls out the essential data from a current instance via the REST interface, and one that loads that data into a new instance, can the Debian install machinery do that automatically? | 19:32 |
| th1a | Run those? | 19:32 |
| jinty | qualified yes. | 19:33 |
| bskahan | th1a: you could access the ZODB directly for that | 19:33 |
| th1a | Oh yeah. | 19:33 |
| th1a | How would that work? | 19:33 |
| * mgedmin delurks | 19:34 | |
| * jinty actually thinks - cannot guarentee the previous server is running | 19:34 | |
| th1a | Could we just add the attributes directly? | 19:34 |
| mgedmin | if there are only those two issues (other name for the root group, new attribute on persons) | 19:34 |
| mgedmin | then it should be simple to add a couple of lines of code to upgrade the database | 19:34 |
| th1a | jinty: right, so just manipulating the ZODB might be preferable. | 19:34 |
| mgedmin | without worrying about infrastructure like zope 3 generations | 19:35 |
| th1a | OK. Let's do that. | 19:35 |
| jinty | Point, before we get carried away: I never uploaded packages of 0.9rc1 anywhere. | 19:35 |
| mgedmin | in Person.__setstate__ call the inherited method and then if not hasattr(self, 'newattribute'): self.newattribute = defaultvalue | 19:35 |
| th1a | jinty: that's ok. | 19:36 |
| mgedmin | in ensureAppExists add a check if there's a group named 'root' and if there is, rename | 19:36 |
| mgedmin | rename might be tricky: remove from group container, change __name__, readd - or something like that | 19:36 |
| * mgedmin goes back to lurking | 19:37 | |
| th1a | bskahan: could you do that? (we'll add it to your next contract) | 19:37 |
| jinty | th1a: I mean that people who build packages themselves - well, it's shouldn't be too well supported. | 19:38 |
| th1a | Hm? | 19:38 |
| jinty | s/it's/it/ | 19:39 |
| bskahan | I can do it, do we need to if 0.9rc1 was never published? | 19:39 |
| th1a | We're worried about migration from 0.8. | 19:39 |
| bskahan | oh, that's a different ball of wax | 19:39 |
| th1a | Oh. Sorry. | 19:40 |
| jinty | ouch | 19:40 |
| mgedmin | too many changes? | 19:40 |
| bskahan | maybe not | 19:40 |
| bskahan | but I don't know them off the top of my head | 19:41 |
| th1a | There's very little new functionality. It's all UI-oriented. | 19:41 |
| bskahan | since 95% of the work was in the schooltool.browser its quite possible | 19:41 |
| * bskahan nods | 19:41 | |
| * bskahan goes to look | 19:41 | |
| jinty | while he is looking: upload of 0.9 packages | 19:43 |
| jinty | personally I lean towards it, on the basis that schooltool has only ever been in debian-unstable and ubuntu-universe | 19:45 |
| th1a | Lean towards what? | 19:45 |
| jinty | towards uploading 0.9 to debian/ubuntu | 19:46 |
| th1a | We're eventually going to be uncoupling SchoolTool and SchoolBell releases permanantly. | 19:47 |
| th1a | Actually, in a couple months. | 19:47 |
| jinty | Great, but untill I can actually put them in separate source packages without copying the entire code base, we should think of something. | 19:48 |
| th1a | OK. | 19:49 |
| th1a | So what are you thinking for ST 0.9? | 19:49 |
| jinty | The idea I thought of last night won't work. | 19:50 |
| jinty | But unstable is well unstable - so perhaps just upload them as they are. | 19:51 |
| jinty | at least it gives people a chance to test schoolbell. | 19:51 |
| th1a | We can just quickly move the UI improvements over to ST and release them together. | 19:52 |
| jinty | WOW! | 19:52 |
| jinty | you mean trunk? | 19:52 |
| th1a | I suppose I do. Haven't I suggested this before? | 19:53 |
| bskahan | yes | 19:53 |
| tvon | I suspect something will break, but we could give it a whirl | 19:53 |
| bskahan | I'm not into the idea | 19:53 |
| bskahan | its completely untested that way | 19:53 |
| tvon | yeah | 19:53 |
| * jinty wonders if his memory is broken | 19:53 | |
| bskahan | depending on when you want to release | 19:54 |
| th1a | A week? | 19:54 |
| bskahan | have to ask POV | 19:54 |
| bskahan | I have no idea how close trunk is to being ready to release | 19:54 |
| th1a | OK. Maybe not trunk. | 19:54 |
| th1a | That's not what I was thinking of. | 19:54 |
| jinty | and to ui | 19:54 |
| * jinty has his fingers crossed | 19:55 | |
| th1a | This is all too confusing. | 19:55 |
| jinty | maybe we take a step back | 19:55 |
| jinty | summary: | 19:56 |
| th1a | If you run the SchoolTool server from the SB ui branch, does it work as schooltool? | 19:56 |
| bskahan | yes | 19:56 |
| tvon | bskahan: what of periods in the ui? | 19:56 |
| mgedmin | trunk always works (unless someone breaks it) | 19:56 |
| tvon | heh | 19:56 |
| mgedmin | the tricky bit would be to merge the ui changes without making parts of schooltool functionality inaccessible | 19:57 |
| * jinty tries testing packages built from trunk in chroot | 19:57 | |
| th1a | That has to be done now anyhow. | 19:57 |
| * jinty waits for things to compile - a lot | 20:02 | |
| *** SteveA has quit IRC | 20:06 | |
| jinty | hmmm - ImportError: No module named ZConfig | 20:07 |
| mgedmin | jinty, trunk? | 20:08 |
| jinty | yes | 20:08 |
| mgedmin | what command produced this error? | 20:08 |
| tvon | Should I tell anyone if I get a portal_status_message: Email notice error: 'Connection unexpectedly closed' | 20:08 |
| jinty | buildschooltool | 20:08 |
| tvon | from the collector on members.p.o | 20:08 |
| tvon | er, wrong room | 20:08 |
| jinty | basically installing built packages into a clean chroot and trying to start the server | 20:09 |
| mgedmin | ah | 20:09 |
| mgedmin | the source layout changed | 20:09 |
| mgedmin | we no longer have zope modules in src/ | 20:09 |
| mgedmin | we have the full zope 3 checkout as a sibling to src | 20:09 |
| mgedmin | so you can run schooltool from a source checkout | 20:09 |
| mgedmin | but debian packages will not work | 20:09 |
| * mgedmin wonders if anyone has already built zope 3 packages for debian | 20:10 | |
| mgedmin | if we could only depend on python2.3-zope3 ... | 20:10 |
| jinty | experimental stage only - zopex3 | 20:10 |
| jinty | have a copy from ubuntu on my machine | 20:11 |
| jinty | am not sure about debian | 20:12 |
| jinty | So, to get the debian packages from trunk to work, I only have to include Zope3 and put it on the path? | 20:14 |
| mgedmin | afaics yes | 20:14 |
| jinty | ok, and the current src/zope directory? | 20:15 |
| *** bskahan has quit IRC | 20:16 | |
| mgedmin | there should be no src/zope directory | 20:18 |
| jinty | true - sorry!!! | 20:18 |
| jinty | btw: anyway I can easily separate the arch-indep files from modules? (I discovered .py files are not arch indep) | 20:20 |
| jinty | i.e. contents of src/schooltool/browser/www | 20:21 |
| mgedmin | patch the Template class | 20:24 |
| jinty | thanks. | 20:24 |
| mgedmin | override get_path_from_prefix | 20:24 |
| mgedmin | also change read_file in the same module (src/schooltool/rest/__init__.py) | 20:25 |
| mgedmin | recursively grep the source tree for __file__ just to be safe | 20:25 |
| jinty | Running the functional tests on the server in a chroot enough to catch everything? | 20:26 |
| mgedmin | our functional tests for the web interface are not very complete | 20:27 |
| jinty | ah, ok then, leave it to the users | 20:27 |
| mgedmin | I suggest looking at a couple of pages with a browser | 20:27 |
| jinty | ok | 20:28 |
| mgedmin | if you see text, images and css, then it works | 20:28 |
| mgedmin | I do not think there are any other arch-indep files in there | 20:28 |
| mgedmin | maybe zconfig's schema.xml | 20:28 |
| mgedmin | maybe our relaxng schemas, but patching read_file() should fix them | 20:28 |
| jinty | Ok, i'll give it a try, probably for 0.10 | 20:29 |
| jinty | I don't need to be super clean. Just shut up lintian a bit. | 20:29 |
| jinty | th1a: I should be able to release schoolbell/tool from trunk after a couple of days work. | 20:31 |
| jinty | Thanks marius. | 20:32 |
| jinty | th1a: I need to go and look at a new flat, speak later about the release. | 20:33 |
| *** jinty has quit IRC | 20:33 | |
| th1a | OK. I was in the shower. | 20:35 |
| *** SteveA has joined #schooltool | 21:11 | |
| th1a | tvon: ayt? | 21:22 |
| tvon | yup | 21:23 |
| th1a | Did you get my Jabber message? | 21:25 |
| tvon | mhm, no | 21:25 |
| tvon | my jabber connection to you is always a bit gooney, usually very laggy | 21:26 |
| tvon | lemme reconnect | 21:26 |
| tvon | since I'm on a shoddy coffee shop connection here | 21:26 |
| th1a | Ah. Just got a phone call from Mark. They need your bank account info to do the wire transfer. | 21:26 |
| tvon | ah, okay | 21:26 |
| tvon | where should I send it? | 21:26 |
| th1a | You can send it to Mark directly, I suppose. | 21:27 |
| tvon | email? | 21:27 |
| th1a | Yes. They tried to call your office, I guess. | 21:27 |
| tvon | oh, Brian should be there but he poofed a lil while ago, not sure where he ran off too | 21:28 |
| tvon | could call my cell perhaps | 21:28 |
| th1a | I think the bank tried to call your office. | 21:28 |
| tvon | oh | 21:28 |
| tvon | hrm | 21:28 |
| th1a | Just email Mark. | 21:28 |
| tvon | okay, address? | 21:28 |
| tvon | btw, are we going to bundle svn trunk with our 1.0 release or are we hoping that zope 3.1 is out by then? | 21:50 |
| th1a | I'm unclear on that. | 21:51 |
| mgedmin | we'll see | 21:51 |
| th1a | I'll be out for a bit. | 21:51 |
| * tvon nods | 21:51 | |
| *** bskahan has joined #schooltool | 22:04 | |
| bskahan | reading the log from my connection being down | 22:06 |
| bskahan | mail on zope-dev list today made it sound like 3.1 won't be out in time for our 1.0 | 22:07 |
| *** alga has quit IRC | 22:09 | |
| * bskahan can't wait for bazzar | 22:18 | |
| bskahan | now I've missed everyone | 22:24 |
| bskahan | if we're going to release from trunk should I spend time looking at 0.8 -> 0.9 migration? | 22:25 |
| bskahan | or just keep working on merging? | 22:25 |
| * bskahan concludes merging | 22:28 | |
| *** tvon has quit IRC | 22:32 | |
| *** d2m has joined #schooltool | 22:32 | |
| *** th1a has quit IRC | 22:42 | |
| *** tvon has joined #schooltool | 22:54 | |
| bskahan | http://suppressingfire.dnsalias.org/~fb/chambery.vfb sample .vfb | 23:40 |
| *** th1a has joined #schooltool | 23:41 | |
| th1a | Either way, 0.8 to 0.9 conversion would be nice, if we can get it with an afternoon's work. | 23:43 |
| bskahan | th1a: I don't think that's going to happen if 0.9 is coming out of trunk | 23:47 |
| th1a | I guess that makes sense. | 23:47 |
| th1a | Hm... | 23:48 |
| th1a | I would be happier with SB/ST 0.9 out of the branch, with 0.8 > 0.9 migration. | 23:49 |
| th1a | But I suppose we're just delaying the inevitable | 23:49 |
| th1a | I should just write export/import scripts. | 23:50 |
| th1a | That's a good project for me. | 23:50 |
Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!