| povbot | /svn/commits: * srichter committed revision 4796: | 00:06 |
|---|---|---|
| povbot | /svn/commits: Converted ICalendarOwner to an adapter to ISchoolBellCalendar; components that have a calendar implement IHaveCalendar. | 00:06 |
| tvon | so is srichter's branch replacing trunk before the next release? | 00:09 |
| srichter | something like that | 00:10 |
| srichter | or Marius merges the new code in one by one | 00:10 |
| povbot | /svn/commits: * srichter committed revision 4797: | 00:13 |
| povbot | /svn/commits: I just decided that extension will be okay for this use case. | 00:13 |
| povbot | /svn/commits: * tvon committed revision 4798: | 00:19 |
| povbot | /svn/commits: removing unused import | 00:19 |
| tvon | I probably shouldnt bother with commits like that if a merge is coming | 00:21 |
| srichter | well, that's hard to say | 00:22 |
| srichter | on the other hand we cannot stall development because of this | 00:22 |
| povbot | /svn/commits: * tvon committed revision 4799: | 00:38 |
| povbot | /svn/commits: more unused imports | 00:38 |
| *** jinty has joined #schooltool | 00:47 | |
| jinty | srichter: does zpkgtools dependency information contain version info? It doesn't look like it. | 01:00 |
| srichter | no | 01:00 |
| srichter | it is not meant to replace packaging software | 01:00 |
| srichter | it is meant to pull stuff together you need for a particular release | 01:01 |
| jinty | thanks, I'm just trying to digest it. to figure out the best way of working with it. | 01:02 |
| srichter | cool | 01:02 |
| srichter | I am glad you are having a serious look | 01:02 |
| jinty | we just got another debian bug about it... | 01:03 |
| srichter | I see | 01:03 |
| * jinty is sreious, but a little slow | 01:03 | |
| jinty | and can't spell | 01:03 |
| srichter | that's no problem; I have to look at it more closely too to get ST setup to use it | 01:04 |
| jinty | I would really like to be able to get the debian packaging and zpkgtools to work together at some level. But it is difficult. | 01:05 |
| srichter | mmh, that sounds interesting | 01:05 |
| srichter | that would be really cool, though | 01:05 |
| jinty | imagine not only producing a monolithic tarball, but a fleet of debian packages, all tested together with correct dependency info. | 01:06 |
| jinty | or something like that... | 01:06 |
| srichter | yep | 01:07 |
| jinty | anyway, I will go back to digesting | 01:07 |
| jinty | thanks | 01:07 |
| *** jinty has quit IRC | 01:54 | |
| *** gintas has joined #schooltool | 02:24 | |
| *** gintas has left #schooltool | 02:26 | |
| *** gintas has joined #schooltool | 02:26 | |
| *** gintas has left #schooltool | 02:32 | |
| *** gintas has joined #schooltool | 02:33 | |
| *** gintas has left #schooltool | 02:43 | |
| *** gintas has joined #schooltool | 02:43 | |
| tiredbones | /quit | 04:46 |
| *** tiredbones has quit IRC | 04:46 | |
| povbot | /svn/commits: * srichter committed revision 4800: | 04:58 |
| povbot | /svn/commits: Updated schooltool to use ISchoolBellCalendar adapter instead of IOwnerCalendar. | 04:58 |
| *** povbot` has joined #schooltool | 07:46 | |
| *** povbot has quit IRC | 07:46 | |
| *** Ricey_ has joined #schooltool | 08:29 | |
| *** Ricey has quit IRC | 08:29 | |
| *** povbot has joined #schooltool | 09:07 | |
| *** tvon has quit IRC | 10:57 | |
| *** tvon has joined #schooltool | 11:34 | |
| *** jinty has joined #schooltool | 12:17 | |
| *** gintas has joined #schooltool | 13:10 | |
| srichter | gintas: is marius already around? | 13:21 |
| gintas | srichter, I guess not | 13:22 |
| gintas | I'm at home right now | 13:22 |
| srichter | I see | 13:22 |
| *** th1a has joined #schooltool | 13:54 | |
| *** bskahan has joined #schooltool | 14:35 | |
| *** gintas has left #schooltool | 14:36 | |
| bskahan | http://people.etria.com/~bskahan/schooltool_0point11/rotating_schedule_thumb.png | 14:38 |
| bskahan | well, more generally | 14:38 |
| bskahan | http://people.etria.com/~bskahan/schooltool_0point11/ | 14:38 |
| th1a | Huh what? | 14:38 |
| bskahan | got started on screenshots and thumbnails but ran out of time | 14:38 |
| th1a | Oh. Thanks. | 14:39 |
| bskahan | np | 14:39 |
| * th1a is sitting outside POV, waiting for someone else to show up. | 14:39 | |
| bskahan | heh | 14:39 |
| bskahan | its 2pm there? | 14:39 |
| th1a | These guys look less ruthlessly efficient up close. | 14:39 |
| th1a | I woke up Marius at 2:00. | 14:39 |
| bskahan | >-) | 14:39 |
| th1a | It is 2:40 now. | 14:39 |
| th1a | With Aiste out of town, things start even later... | 14:40 |
| srichter | he he | 14:40 |
| bskahan | hah | 14:41 |
| srichter | yeah, get this guy over here; we have business to talk about | 14:41 |
| srichter | :-) | 14:41 |
| th1a | It is kind of funny watching her play mother hen to these guys. | 14:41 |
| srichter | I really want to switch to Python 2.4 for the trunk, ince my stuff merges | 14:41 |
| srichter | th1a: I am done with most of the cleanup tasks (except the merge and the DB compatibility, of course), so I am making really good progress | 14:43 |
| th1a | Well, we do need to switch our context now. The work in the trunk won't go into production until spring, so requiring users to switch to 2.4 at that point would seem to be even less of an issue than it is now. | 14:43 |
| srichter | th1a: but I think a lot of the clutter was due to a lack of (a) a dictator or checkin police and (b) story-based contracts | 14:43 |
| srichter | I think from time to time 10% of a contract should include cleanup | 14:43 |
| th1a | Yes. We've been racing to get these releases out for a while. We didn't really have time for a pause. Plus I knew we could do it now. | 14:46 |
| th1a | Having an outsider come in and look over everything has been particularly helpful, however. | 14:46 |
| srichter | I am particularly worried about tests at this point | 14:47 |
| srichter | they are almost totally unmanagable | 14:48 |
| srichter | We need to find a way to better pack them into text files, so they are compiled as part of the documentation | 14:48 |
| srichter | then 3rd party developers will tell us, if our description sucks in a particular place | 14:49 |
| th1a | Does development of Zope 3 itself handle tests better? | 14:49 |
| srichter | recently we have become much better with it | 14:50 |
| srichter | we have only resorted to doctests in *.txt files lately | 14:50 |
| srichter | Zope 3 sucks with ftests of course | 14:50 |
| th1a | I like doctests in text files. | 14:50 |
| srichter | but the ftests in ST are pretty okay and once we convert them to testbrowser-based tests, they will be very readable | 14:51 |
| srichter | me too | 14:51 |
| povbot | /svn/commits: * jinty committed revision 4801: | 14:51 |
| srichter | there are currently too many tests in docstrings in python in ST | 14:51 |
| povbot | /svn/commits: Ruthelessly patch build-debs.py: | 14:51 |
| povbot | /svn/commits: * Add support for building with pbuilder. * Add support for the dpkg-buildpackage options -v and -sa. | 14:51 |
| th1a | I definitely want to write testbrowser tests as acceptance tests. | 14:52 |
| srichter | what are "acceptance tests"? | 14:52 |
| srichter | as evaluation whether a story is implemented? | 14:52 |
| srichter | OT: did you make progress with the School WF? | 14:53 |
| th1a | Yes. | 14:53 |
| th1a | School WF? | 14:53 |
| srichter | School workflow | 14:53 |
| th1a | Oh. I've been too busy. | 14:53 |
| srichter | I see | 14:53 |
| th1a | This trip has been less of a vacation than I imagined. | 14:53 |
| th1a | But that's ok. | 14:54 |
| srichter | yeah, it can swing either way | 14:54 |
| srichter | mmh, another thought to the dev process: maybe having design meetings would help | 14:55 |
| th1a | Actually, I realized that the basic turn sequence in my game is about the simplest possible workflow: igo ugo. | 14:55 |
| th1a | So I'm going to get that working first to get a feel for workflow. | 14:55 |
| bskahan | heh | 14:55 |
| srichter | oh, that is cool | 14:55 |
| srichter | actually it is one step: Complete Turn | 14:55 |
| srichter | and basically the participant changes; I wonder whether this works | 14:56 |
| th1a | So I might be hitting you up for some help when I get to that. I have to get done with naval movement first. | 14:56 |
| srichter | ok, cool | 14:56 |
| srichter | I wonder whether we could use testbrowser acceptance tests to build up slowly a complete SchoolTool book | 14:59 |
| srichter | If we attack this correctly, it could be very cool | 14:59 |
| srichter | the developers would write the tests in a geeky fashion and you could put better text around them | 15:00 |
| srichter | and we compile it using apidoc | 15:00 |
| povbot | /svn/commits: * jinty committed revision 4802: | 15:00 |
| povbot | /svn/commits: * Re-build for upload to Debian. * On installation set the permissions of /var/{lib,log}/schoolbell to 0750. * Patch server startup script to change the umask to 007. | 15:00 |
| povbot | /svn/commits: NOTE: the patch to the server startup script is not contained in any repository. Yes, I know it is very evil and I should be using dpatch. But shortly the schoolbell packaging stuff will use arch and arch-buildpackage, then this problem dissapears. | 15:00 |
| th1a | I'm not sure how useful the testbrowser tests would actually be to outside developers or users. | 15:01 |
| srichter | well, in a book you have to take the user through steps anyways | 15:01 |
| srichter | If you want to create a new resource, click on the "Resources" link in the action menu. | 15:02 |
| srichter | 15:02 | |
| srichter | >>> browser.click("Resources") | 15:02 |
| srichter | 15:02 | |
| srichter | Now click on ... | 15:02 |
| bskahan | th1a: I don't think the current ftests are particularly useful as developer documentation | 15:03 |
| th1a | I don't think that would help actual users. | 15:03 |
| srichter | I think we could even massage docutils in ignoring those doctest entries completely and just show the text in certain rendering cases | 15:03 |
| bskahan | I haven't looked at the mechtests in depth enough to know if they're better though | 15:03 |
| th1a | I'm just saying I think browser tests are fantastic as a guide to what the application should do, and far more readable than the existing ones, | 15:04 |
| srichter | bskahan: they are a million times better! | 15:04 |
| th1a | but it doesn't seem like something you'd literally hang a book around. | 15:04 |
| srichter | bskahan: I refuse to write other style functional doctests (I always did, btw) | 15:04 |
| bskahan | srichter: :) | 15:04 |
| srichter | th1a: Jim mentioned several times that he would write a book in doctest style. | 15:05 |
| th1a | Or if you did, you'd mostly be taking out the doctests themselves, because "browser.click("Resources") still doesn't help actual humans. | 15:05 |
| bskahan | I wasn't saying they weren't, I just haven't used them (as a producer or consumer) | 15:05 |
| srichter | th1a: but it helps the documentation not getting outdated | 15:05 |
| th1a | Well, Philipp's book has a lot of doctest feel in it. | 15:05 |
| * jinty hates having to upload multiple files to plone one by one through a web interface and cries out for ftp... | 15:05 | |
| th1a | I'm only making a very small point here. | 15:05 |
| srichter | th1a: the main problem with documentation is information rotting | 15:05 |
| th1a | jinty: We are very, very close to getting a server of our own. | 15:06 |
| srichter | yeah, he started later and read Jim's comments :-) | 15:06 |
| bskahan | th1a: we managed to postpone enough to use plone 2.1 ;) | 15:06 |
| th1a | The only delay now is that Canonical seems to use renting servers as a strategy for getting new companies to support Ubuntu. | 15:06 |
| th1a | So when they need 'x' servers for non mission critical stuff, they approach a major host who doesn't support Ubuntu and say, | 15:07 |
| th1a | "we'll get X servers if you support Ubuntu for everyone." | 15:07 |
| th1a | At least that was my interpretation of the explanation. | 15:07 |
| srichter | he he | 15:08 |
| th1a | So we're waiting to see if ServerPronto will do it. | 15:08 |
| th1a | If not, we'll move to the next host on the list. | 15:08 |
| * bskahan is going to try upgrading the laptop to breezy colony-3 this weekend | 15:09 | |
| th1a | bskahan: I'd like you to help set up the new site. | 15:09 |
| * bskahan nods | 15:09 | |
| bskahan | I have some thoughts about it | 15:09 |
| th1a | That'll give you a little more work in the next month or so. | 15:09 |
| bskahan | mostly, getting the benefit that we should from plone | 15:10 |
| th1a | I want everything to fit into the Canonical model, which didn't really exist when we started. | 15:10 |
| bskahan | what's that? | 15:10 |
| th1a | Just the way they divide things up among Plone, Wiki's, forums, etc. | 15:10 |
| th1a | Using LaunchPad more. | 15:10 |
| th1a | Perhaps moving to Malone. I have to take another look at it. It looks more useful. | 15:11 |
| th1a | Than it used to. | 15:11 |
| bskahan | getting the full benefit from LP might require moving to bazaar[.|?] | 15:12 |
| bskahan | I'm actually opposed to the wiki + plone setup | 15:12 |
| th1a | After discussing it more with SteveA, I'd say it is only a matter of time until we move to bazaar. | 15:13 |
| * bskahan agrees | 15:13 | |
| th1a | We're going to move to doing all development on branches and only merging after code is reviewed. | 15:13 |
| * jinty wants an ftp site and a package pool and an account on the server and web space and some more stuff;) | 15:13 | |
| bskahan | ;) | 15:13 |
| th1a | Although we probably won't do that completely until we start using bazaar. | 15:14 |
| th1a | Why are you against wiki+Plone? | 15:14 |
| srichter | I am generally -1 for working on branches; it is always painful to merge | 15:14 |
| srichter | though I understand the quality issues | 15:15 |
| srichter | and arguments | 15:15 |
| jinty | isn't distributed revision control meant to solve some of that pain? | 15:16 |
| th1a | That's my understanding. | 15:16 |
| srichter | what is "distributed revision control"? | 15:17 |
| jinty | tla/baz/git... | 15:17 |
| bskahan | plone has really nice community oriented document handling, wikis are crud ;) | 15:17 |
| bskahan | more seriously, people end up using the wiki for things that should be done in a real CMS | 15:17 |
| bskahan | plone.org is a much more useful community site than the ubuntu wiki, imnsho | 15:17 |
| srichter | I'd rather have something like sandbox, we all work on and only move over revisions after review to the trunk | 15:18 |
| th1a | On LaunchPad they do a branch for each feature. | 15:19 |
| th1a | Lots of small branches. | 15:19 |
| srichter | it is amazing that Zope 3 did not have quality issues yet always working on the trunk | 15:19 |
| th1a | bskahan: I agree with you in theory, but in practice, outsiders don't contribute to Plone sites. | 15:20 |
| th1a | It is hard for people to understand the dual function of the community Plone site. | 15:21 |
| * bskahan nods | 15:22 | |
| bskahan | yeah, that's the barrier I'd like to get over | 15:22 |
| bskahan | plone.org seems to have worked it out, as well as some other sites. | 15:22 |
| th1a | Well, I think mirroring practices that are working for Canonical is a good idea. | 15:22 |
| povbot | /svn/commits: * jinty committed revision 4803: | 15:22 |
| povbot | /svn/commits: OOPS! | 15:22 |
| * th1a is still twiddling his thumbs outside POV. | 15:23 | |
| bskahan | lol | 15:23 |
| bskahan | th1a: do you know what wiki software they're using? | 15:23 |
| th1a | It is moin. | 15:24 |
| bskahan | th1a: one thing they have that's interesting ... their web forums are mirrors of mailing lists, if I understand correctly | 15:24 |
| bskahan | that I really like | 15:24 |
| th1a | Well... that's one thing I'm confused about. What URL are you looking at? | 15:24 |
| povbot | /svn/commits: * jinty committed revision 4804: | 15:24 |
| povbot | /svn/commits: port 4802, 4803 to schooltool. | 15:24 |
| bskahan | th1a: I'm not | 15:24 |
| bskahan | I remember a conversation in #ubuntu | 15:25 |
| bskahan | that brought it up | 15:25 |
| th1a | Oh. They've also got ubuntu-traffic, which we may be busy enough to need to replicate someday: http://people.ubuntulinux.org/~mako/ubuntu-traffic/ | 15:26 |
| bskahan | http://www.ubuntuforums.org/ | 15:26 |
| th1a | That's not Canonical, though, I don't think. | 15:27 |
| * bskahan nods | 15:27 | |
| th1a | That forum software isn't free, oddly enough. | 15:27 |
| bskahan | http://www.ubuntuforums.org/forumdisplay.php?f=29 | 15:27 |
| bskahan | ah | 15:27 |
| bskahan | lame | 15:27 |
| bskahan | I think that's a great way to get non-technical teachers in touch with developers | 15:28 |
| bskahan | since neo-luddite english teachers might use a web forum long before a mailing list | 15:29 |
| bskahan | ;) | 15:29 |
| povbot | /svn/commits: * srichter committed revision 4805: | 15:29 |
| povbot | /svn/commits: Updated views to the ST version, so we can remove the duplication, which was effectively only in the I18n domain. | 15:29 |
| th1a | I agree that having a forum is good. | 15:31 |
| th1a | Forum, FAQ, one of those release management packages for Plone. | 15:31 |
| povbot | /svn/commits: * srichter committed revision 4806: | 15:32 |
| povbot | /svn/commits: Removed some duplicated views. | 15:32 |
| *** bskahan has quit IRC | 15:35 | |
| th1a | If mgedmin doesn't get here soon, I'm going to have to go looking for a toilet. | 15:37 |
| srichter | LOL | 15:39 |
| *** th1a has quit IRC | 15:56 | |
| *** th1a has joined #schooltool | 15:57 | |
| srichter | th1a: found a bathroom? :-) | 15:58 |
| th1a | Marius arrived. | 15:58 |
| srichter | he he | 15:58 |
| *** gintas has joined #schooltool | 15:58 | |
| *** mgedmin has joined #schooltool | 15:58 | |
| th1a | Hoi gintas. | 15:59 |
| gintas | th1a: I've just uploaded the windows binary installers for SB and ST | 15:59 |
| th1a | OK. Cool. | 15:59 |
| th1a | Are you not coming in today? | 15:59 |
| gintas | I thought I'd take a day off | 16:00 |
| srichter | mgedmin: can I switch the refactor branch to Python 2.4 in anticipation that it will be used in the trunk (th1a okayed it) | 16:00 |
| mgedmin | ok | 16:00 |
| srichter | good | 16:00 |
| th1a | That's more of a jinty question. | 16:00 |
| srichter | jinty: Py 2.4 for the next release ok? | 16:00 |
| th1a | That is, the "next release" being in April. | 16:01 |
| srichter | th1a: I think his objections were already squashed, since all debian distros have Py 2.4 available | 16:01 |
| gintas | if we're talking about the April release, I'm sure that python2.3 support will be a very low priority | 16:01 |
| th1a | ... giving him one last chance ... | 16:01 |
| mgedmin | 3 ... 2 ... 1 ... | 16:02 |
| jinty | srichter: I think this is the third time I am saying yes | 16:02 |
| srichter | well, I just never heard it officially from the exec team | 16:03 |
| srichter | ok, it is though then :-) | 16:03 |
| * jinty wonders if anyone actualy reads his mail to the list | 16:03 | |
| th1a | I read and I forget. | 16:08 |
| * jinty sighs! | 16:10 | |
| povbot | /svn/commits: * jinty committed revision 4807: | 16:10 |
| povbot | /svn/commits: Add support for debian specific patches to build-debs.py. also add the umask patches to the repository. | 16:10 |
| povbot | /svn/commits: * srichter committed revision 4808: | 16:15 |
| povbot | /svn/commits: Renamed the traverser since the ICalendarOwner interface is gone. | 16:15 |
| *** Ricey_ has quit IRC | 16:19 | |
| *** Ricey has joined #schooltool | 16:20 | |
| srichter | mgedmin: why is the initial site.zcml code in a Python string and not in a file? | 16:25 |
| mgedmin | why not ;) | 16:27 |
| srichter | because it makes it harder to hook in new things | 16:27 |
| mgedmin | how come? | 16:28 |
| srichter | because if we want to support override.zcml or package-includes, we have to constantly change that string | 16:28 |
| srichter | so much easier to modify a file | 16:28 |
| mgedmin | that string is in a file | 16:29 |
| mgedmin | I fail to see the difference | 16:29 |
| mgedmin | actually, perhaps I do understand your point | 16:29 |
| srichter | :-) | 16:29 |
| mgedmin | if the people doing the customization aren't developers but site admins/etc, then yes, having a site.zcml makes sense | 16:29 |
| mgedmin | I do not object to it | 16:29 |
| srichter | ok | 16:29 |
| mgedmin | some history: initially ST/SB was a Z3 content object | 16:29 |
| mgedmin | I did some prototyping and got it running as a standalone server | 16:30 |
| srichter | I see | 16:30 |
| mgedmin | then I implemented the standalone server properly (with unit tests, etc) | 16:30 |
| mgedmin | and put everything dealing with that use case in a single file | 16:30 |
| srichter | I see | 16:30 |
| srichter | btw, have you tried using Zope's startup and just replacing the root folder witha schooltool site? | 16:31 |
| mgedmin | I was sort of breaking new ground at the time, and wanted a nicely contained example of "here's how you can build a Z3 based app instead of just a plugin for a Z3 web app server" | 16:31 |
| srichter | ok | 16:31 |
| srichter | right | 16:31 |
| mgedmin | but now the time has come for splitting it up, I think | 16:31 |
| srichter | ok, can we discuss this briefly? | 16:32 |
| mgedmin | re Z3 startup & replacing -- yes, but all the subscribers that create a site and populate it with utilities broke | 16:32 |
| srichter | mmh, that's aweful | 16:32 |
| srichter | do you remember why? | 16:33 |
| mgedmin | then I discovered that if I create the root object and run the subscribers, they do the right thing for us | 16:33 |
| *** gintas has left #schooltool | 16:33 | |
| srichter | mmh | 16:34 |
| srichter | mgedmin: here is the reason I ask: | 16:34 |
| srichter | I am thinking about the best way to be able to run ST abd SB from the same source code | 16:35 |
| srichter | basically I want to be able to say: | 16:35 |
| srichter | make run-st | 16:35 |
| srichter | make run-sb | 16:35 |
| srichter | do you think I could make SB/ST work with the Zope 3 startup routine? | 16:38 |
| mgedmin | if you could, that would be great for other standalone apps that want to use Z3 | 16:38 |
| srichter | yes :-) | 16:38 |
| mgedmin | take a look at z3 bootstrap events and subscribers | 16:38 |
| srichter | I am willing to make checkins into the Zope 3 trunk too | 16:39 |
| mgedmin | I mentioned once that they are ordering-dependent now | 16:39 |
| mgedmin | one subscriber creates the RootFolder and a site | 16:39 |
| * srichter wishes he would sit beside mgedmin and do some pair programming | 16:39 | |
| mgedmin | other subscribers then populate the site with attributes | 16:39 |
| mgedmin | if we <include package="zope.app" />, there's no way to introduce our own bootstrap event subscriber that could be called before the Z3 ones | 16:39 |
| mgedmin | if we replace the root folder with our app after Z3 subscribers, we'd have to call all the subscribers again, to get the site populated with utilities... wait a second... | 16:40 |
| srichter | ok, I am willing to change the Zope 3 trunk to fix that problem | 16:40 |
| mgedmin | maybe we could pry apart the old site from the old root folder and attach it on our app? | 16:40 |
| mgedmin | anyway, if we create the app object instead of the root folder, and then call the subscribers, everything works | 16:41 |
| srichter | uuuh, sounds good | 16:41 |
| mgedmin | suggestion: a global utility to create the app object | 16:41 |
| mgedmin | defined in zope.app zcml | 16:41 |
| srichter | but subscribers are not sorted | 16:41 |
| mgedmin | and we could override it in our zcml | 16:41 |
| mgedmin | the code that initializes ZODB could do | 16:41 |
| mgedmin | root_object = getUtility(IRootObjectFactory)() | 16:41 |
| mgedmin | zodb_connection.root()[PUBLICATION_NAME_OR_WHATEVER] = root_object | 16:42 |
| mgedmin | notify(BootstrapEvent(root_object)) | 16:42 |
| mgedmin | I'm probably mangling all the names that are actually used in Zope | 16:42 |
| mgedmin | but I hope you get the idea | 16:42 |
| srichter | I think this sounds like a winner | 16:42 |
| srichter | yep | 16:42 |
| srichter | hehe, there is even an "ugh.." comments there (probably from me), so I even have a good argument to remove it :-) | 16:47 |
| srichter | or from you? | 16:49 |
| mgedmin | if that's in Z3 core, then I doubt it's from me ;) | 16:50 |
| srichter | yeah, it is inside the root folder setup. | 16:50 |
| srichter | mgedmin: ok, now I got the utility all setup and working, what do you susggest is the best approach to this? | 17:05 |
| srichter | link schoolbell into Zope3/src and modify site.zcml and overrides.zcml? | 17:06 |
| srichter | I guess that's all that should be needed? | 17:08 |
| mgedmin | wha? | 17:09 |
| mgedmin | -1 | 17:09 |
| srichter | well, eventually, I want to do it as you have it setup now | 17:09 |
| srichter | but for testing purposes to see whether everything will work, what is the fastest way to success? | 17:10 |
| mgedmin | I am confused as to what your goal is | 17:10 |
| mgedmin | you want to run ST and SB from same source base | 17:10 |
| mgedmin | but with different versions of site zcml configuration | 17:10 |
| srichter | well that's step 2 | 17:10 |
| mgedmin | currently ST's main.py inherits SB's main.py and overrides the top-level zcml as a string | 17:11 |
| mgedmin | plus a few other things | 17:11 |
| srichter | right now I want to see whether I get the SB standalone server to utilise more of Zope 3's startup stuff | 17:11 |
| srichter | basically, custom conf file, site.zcml, and overriding IRootObjectFactory utility | 17:12 |
| mgedmin | how do you plan to hook schooltool.conf parsing? | 17:12 |
| srichter | that should give me the same functionality as the standalone server | 17:12 |
| srichter | dunno | 17:12 |
| srichter | this is step 1(b) | 17:12 |
| srichter | :-) | 17:12 |
| mgedmin | I want ST/SB to be able to use Zope 3 libraries from $PYTHONPATH (and share them with other apps), but not rely on having any existing Z3 instance directory | 17:13 |
| srichter | maybe monkey-patching z.a.server.amin.CONF_FILENAME | 17:14 |
| srichter | of course, we could write our own equivalent of z.a.server.main | 17:14 |
| mgedmin | if it is possilble to have a very tiny schooltool-server.py that imports zope.app.main and passes some custom things (path to our site.zcml?), then that's fine | 17:14 |
| srichter | I think this is better | 17:14 |
| srichter | right, this is the ultimate goal | 17:15 |
| mgedmin | we have a Z3 based app that would love to define its own ZConfig file in addition to zope.conf, or perhaps add some sections to zope.conf | 17:15 |
| mgedmin | currently all its runtime process configuration is done via custom zcml directives | 17:16 |
| mgedmin | which is a bit of a hack, but that was the path of least resistance | 17:16 |
| srichter | well, if we rewrite our own z.a.server.main.load_options function than we can plugin our own config schema | 17:17 |
| srichter | and conf filename | 17:17 |
| srichter | all this code is very short, so redoing this bit of code should be easy | 17:18 |
| *** alga has joined #SchoolTool | 17:46 | |
| srichter | mgedmin: partial success; I have successfully replaced the root folder with a schooltool application on a vanilla Zope 3 setup with a custom site.zcml | 17:49 |
| srichter | the skin and everything works too | 17:49 |
| srichter | just security does not work yet; seem something specific you guys do | 17:49 |
| srichter | ok, now I got it :-) | 17:56 |
| *** tvon has quit IRC | 18:02 | |
| *** number5 has joined #schooltool | 18:18 | |
| srichter | darn the problem is that schoolbell has so much custom setup | 18:22 |
| srichter | we could simplify a bit but I am not sure it is worth it now | 18:22 |
| *** number5 has quit IRC | 18:23 | |
| *** number_five has joined #schooltool | 18:23 | |
| *** number_five is now known as number5 | 18:23 | |
| srichter | I think merging takes priority now | 18:23 |
| mgedmin | I see no particular benefits to ST/SB from redoing the startup code | 18:25 |
| mgedmin | (other than splitting out site configuration into a separate zcml file, I mean) | 18:26 |
| th1a | It seems like a relatively low priority. | 18:26 |
| *** number5 has quit IRC | 18:37 | |
| *** number_five has joined #schooltool | 18:37 | |
| *** number_five is now known as number5 | 18:38 | |
| srichter | do you guys know a quick way to do a search and replace on many files? | 18:40 |
| mgedmin | search -- yes, search & replace -- not quite | 18:40 |
| mgedmin | I usually use vim's :grep | 18:40 |
| srichter | yeah, that's what I usually do | 18:41 |
| mgedmin | and then do global search/replace (with confirmation) in each file | 18:41 |
| srichter | but I will go insane if I do this for the merge | 18:41 |
| mgedmin | you can also do something like this in shell: | 18:41 |
| srichter | (shell is great) | 18:41 |
| mgedmin | find \! -path '*/.svn/*'|xargs grep -l PATTERN1|xargs perl -i -p -e 's/PATTERN2/REPLACEMENT/g' | 18:41 |
| mgedmin | (we use this in a script that does branding for one of our projects) | 18:42 |
| srichter | ok | 18:42 |
| srichter | I just found KFileReplace | 18:42 |
| srichter | let's see how this works | 18:43 |
| srichter | if it is no good, I use the script above | 18:43 |
| *** tiredbones has joined #schooltool | 18:57 | |
| *** jinty has quit IRC | 19:07 | |
| povbot | /svn/commits: * srichter committed revision 4809: | 19:14 |
| povbot | /svn/commits: Uugh, almost forgot to check this in before the merge. It basically minimizes the configure.zcml in st.browser. | 19:14 |
| povbot | /svn/commits: * srichter committed revision 4810: | 19:38 |
| povbot | /svn/commits: Okay, this branch is now officially broken until I get everything working again. The reason for the checkin is that I can then move more stuff around. | 19:38 |
| povbot | /svn/commits: Here you can see the initial layout as I envisioned it. | 19:38 |
| povbot | /svn/commits: * srichter committed revision 4811: | 19:39 |
| povbot | /svn/commits: Remove it. | 19:39 |
| *** tvon has joined #schooltool | 19:45 | |
| *** number5 has quit IRC | 19:50 | |
| mgedmin | _Bugs_in_Writing_, segment 133: "login" is a noun, "log-in" is an adjective, "log in" is a verb | 20:02 |
| srichter | I have a strange feeling this is pretty arbitrary :-) | 20:04 |
| mgedmin | not if you want to write good English :-) | 20:04 |
| srichter | what's the source of this info? | 20:04 |
| mgedmin | _Bugs_in_Writing_ by Lyn Dupré | 20:05 |
| srichter | oh, its a book? | 20:05 |
| th1a | If it is in a book, it must be true! | 20:05 |
| srichter | well, then I would at least give it some more credit | 20:06 |
| th1a | How does one use log-in as an adjective? The log-in field? | 20:06 |
| mgedmin | Steve Alexander recommeded this book to us | 20:06 |
| srichter | for example, about every company and big OS project comes up with their own translation guidelines and they often conflict each other | 20:06 |
| th1a | If SteveA recommended it it MUST be true!!! | 20:06 |
| srichter | LOL!!!!!!! | 20:07 |
| th1a | Actually, those make sense. | 20:07 |
| th1a | Although I don't entirely understand the hyphen. | 20:07 |
| srichter | yep, me neither | 20:08 |
| srichter | that's the reason I find it arbitrary | 20:08 |
| srichter | it is like FedEx | 20:08 |
| srichter | FedEx is the noun | 20:08 |
| mgedmin | _Eats,_Shoots_&_Leaves_ by Lynne Truss is a nice book about punctuation | 20:09 |
| mgedmin | I started understanding at least something when I read it | 20:09 |
| srichter | fedex presumely the verb, or should it be fed ex ;-) and fed-ex the adjective? | 20:09 |
| srichter | in america there are no set punctuation rules as I understand it | 20:10 |
| mgedmin | on the Internet there are no rules :) | 20:10 |
| srichter | you can follow the MLA, but there are other standards | 20:10 |
| * mgedmin didn't want to start a flame war | 20:10 | |
| srichter | no, I know | 20:11 |
| th1a | http://www.amazon.com/exec/obidos/tg/detail/-/020530902X/qid=1124384992/sr=8-1/ref=pd_bbs_1/102-5012198-0348910?v=glance&s=books&n=507846 | 20:11 |
| srichter | I was just very disappointed in college, when the English prof basically said: Here you use MLA, but outside the classroom other rules might count | 20:11 |
| mgedmin | what's MLA? | 20:12 |
| srichter | which is the opposite to German | 20:12 |
| th1a | Modern Language Association. | 20:12 |
| srichter | mgedmin: a book that lists all A-English rules | 20:12 |
| srichter | th1a: that book I have heard a lot about and I totally forgot about it | 20:12 |
| srichter | I always wanted to get it; it is supposed to be really good, right? | 20:13 |
| th1a | It is short and to the point. | 20:14 |
| th1a | That's the only thing I'd use. | 20:14 |
| srichter | ok, I'll pick it up at the book store | 20:15 |
| th1a | Probably doesn't cover things like "log in" though. | 20:15 |
| povbot | /svn/commits: * srichter committed revision 4812: | 20:16 |
| povbot | /svn/commits: Okay, some more moving around. I have not even attempted to untangle the wilderness of the browser pacakges. | 20:16 |
| povbot | /svn/commits: * srichter committed revision 4813: | 20:20 |
| povbot | /svn/commits: Applied my automated string replacements. KFileReplace is really nice! | 20:20 |
| th1a | OK, I'm out of here until Monday. | 20:59 |
| th1a | Have a good weekend, folks. | 20:59 |
| *** th1a has quit IRC | 20:59 | |
| *** mgedmin has quit IRC | 22:03 | |
| *** alga has quit IRC | 22:04 | |
| *** tvon has quit IRC | 22:25 | |
| povbot | /svn/commits: * srichter committed revision 4814: | 22:37 |
| povbot | /svn/commits: Okay, got schooltool.resource unittests running. (little yipee!) | 22:37 |
| *** tvon has joined #schooltool | 22:42 | |
| povbot | /svn/commits: * srichter committed revision 4815: | 22:45 |
| povbot | /svn/commits: Unit tests in schooltool.course are passing. | 22:45 |
| povbot | /svn/commits: * srichter committed revision 4816: | 22:47 |
| povbot | /svn/commits: schooltool.note unit tests pass. | 22:47 |
| povbot | /svn/commits: * srichter committed revision 4817: | 22:51 |
| povbot | /svn/commits: schooltool.level unittests pass. | 22:51 |
| povbot | /svn/commits: * srichter committed revision 4818: | 23:22 |
| povbot | /svn/commits: Unittests in schooltool.person pass. | 23:22 |
| tiredbones | How does one get a copy of the repository? | 23:28 |
| tiredbones | Never mind, I think I found it. | 23:29 |
Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!
