| *** jfroche has joined #schooltool | 00:20 | |
| *** didymo has joined #schooltool | 00:31 | |
| *** jhancock_ has quit IRC | 00:32 | |
| *** jfroche has quit IRC | 00:55 | |
| *** didymo has quit IRC | 02:48 | |
| *** didymo has joined #schooltool | 05:31 | |
| *** Fujitsu has quit IRC | 07:57 | |
| *** th1a_ has quit IRC | 09:11 | |
| *** Fujitsu has joined #schooltool | 09:15 | |
| *** didymo has quit IRC | 14:17 | |
| *** ignas has joined #schooltool | 14:46 | |
| *** th1a_ has joined #schooltool | 15:19 | |
| *** mgedmin has joined #schooltool | 15:51 | |
| ignas | th1a: are you in there? | 18:15 |
|---|---|---|
| ignas | aelkner: are you in there? | 18:31 |
| *** jhancock has joined #schooltool | 18:33 | |
| th1a_ | ignas: I am here. | 18:50 |
| ignas | a question about our cooperation with Ian | 18:51 |
| th1a_ | Yes. | 18:51 |
| ignas | from what I understand they want to be using the gradebook aelkner is working on | 18:51 |
| th1a_ | That seemed to make the most sense. | 18:51 |
| ignas | are they going to be modifying it to suit their need? | 18:52 |
| ignas | or they will be just using it | 18:52 |
| th1a_ | Hopefully more or less just using it. | 18:52 |
| th1a_ | They've really got a pretty simple case. | 18:53 |
| th1a_ | I *think*. | 18:53 |
| ignas | because one way to do this is to move jelkner gradebook to trunk | 18:53 |
| th1a_ | Yes. | 18:53 |
| ignas | move schooltoo.jelkner2007 out of the branch | 18:53 |
| ignas | and make aelkner use a buildout of schooltool.jelkner2007 | 18:53 |
| th1a_ | That might be simpler. | 18:53 |
| th1a_ | I'll defer to your judgement. | 18:54 |
| ignas | the downside is that all changes made by aelkner will have to go to trunk ... | 18:54 |
| th1a_ | Right. | 18:54 |
| ignas | all changes to gradebook i mean | 18:54 |
| th1a_ | Well, it seems unlikely that he'd do more than break the gradebook. | 18:54 |
| ignas | and aelkner will have to change the way he does things (start using buildout with 2 parts) | 18:54 |
| ignas | and Ian (or programmers working for him) will have to use a buildout as well | 18:55 |
| th1a_ | I'll defer to your judgment, but my sense is that fewer branches would be simpler. | 18:55 |
| ignas | which will limit them to overriding/extending schooltool, as they don't have commit access | 18:55 |
| ignas | if it only affected the way i do things - it would be OK | 18:56 |
| ignas | i am just worried about how much this will affect the others | 18:56 |
| th1a_ | I think aelkner would be happy to not be in a branch. | 18:56 |
| ignas | i see | 18:56 |
| ignas | ok, then i will just "copy" schooltool.jelkner2007 and make it into something usable as a development egg | 18:57 |
| ignas | and try explaining how things should be done the new way | 18:58 |
| *** Fujitsu has quit IRC | 20:32 | |
| *** aelkner has quit IRC | 20:33 | |
| ignas | th1a: should we put tizard in a separate package or do it as schooltool.tizard? | 20:34 |
| *** aelkner has joined #schooltool | 20:37 | |
| ignas | aelkner: there? | 20:45 |
| aelkner | yes | 20:46 |
| aelkner | i'm reading your discussion with tom on gradebook | 20:46 |
| *** jhancock_ has joined #schooltool | 20:47 | |
| ignas | svn co svn+ssh://source.schooltool.org/svn/schooltool/trunk/jelkner2007-buildout | 20:47 |
| ignas | make run | 20:47 |
| ignas | bin/test | 20:47 |
| ignas | and see if everything works | 20:47 |
| ignas | hmm | 20:47 |
| ignas | ok, it won't work because there is no up to date schooltool egg | 20:48 |
| aelkner | i see | 20:48 |
| ignas | looking at what i can do | 20:48 |
| *** th1a_ has quit IRC | 20:49 | |
| aelkner | i agree with you and tom in the limit to the risk involved in me changing only gradebook in trunk | 20:49 |
| ignas | ok | 20:51 |
| ignas | mgedmin will clean the buildbot | 20:51 |
| ignas | and as soon as you see buildbot package the schooltool egg | 20:51 |
| mgedmin | ignas: ssh muskatas && sudo clear-schooltool-build-slave | 20:52 |
| ignas | try commands i have listed above | 20:52 |
| ignas | in like 30-40 minutes ;) | 20:53 |
| aelkner | you mean me? | 20:53 |
| ignas | yes | 20:53 |
| ignas | i mean you | 20:53 |
| ignas | do the svn co svn+ssh://source.schooltool.org/svn/schooltool/trunk/jelkner2007-buildout | 20:54 |
| ignas | stuff | 20:54 |
| aelkner | i checked the code out already so that i can see what you are doing | 20:54 |
| ignas | as soon as you see buildbot build a new schooltool nightly egg | 20:54 |
| aelkner | then do the make | 20:54 |
| ignas | svn up then as i have changed it a bit | 20:54 |
| aelkner | ok | 20:54 |
| ignas | if you did make with an old schooltool egg | 20:54 |
| aelkner | i didn't | 20:54 |
| ignas | run "bin/buildout -n" | 20:54 |
| ignas | to update all the eggs (including schooltool) | 20:54 |
| ignas | ok, going home, will be online in an hour or so | 20:55 |
| *** ignas has quit IRC | 20:55 | |
| *** jhancock__ has joined #schooltool | 20:56 | |
| *** jhancock has quit IRC | 21:01 | |
| *** mgedmin has quit IRC | 21:05 | |
| *** jhancock_ has quit IRC | 21:11 | |
| *** ignas has joined #schooltool | 21:48 | |
| aelkner | ignas: shall i do the svn up followed by make now? | 22:03 |
| ignas | not yet | 22:03 |
| ignas | unit tests are ok now | 22:03 |
| ignas | but functionals seemed to have some problems with the gradebook | 22:03 |
| * ignas is trying to find out why these tests pass in a buildout, but not in the buildbot | 22:06 | |
| ignas | hmm, ouch, it seems that functional tests are passing in my clean schooltool checkout :/ | 22:12 |
| ignas | cool, i can just force schooltool nightly egg rebuild | 22:19 |
| ignas | ouch, it seems that it's broken | 22:19 |
| ignas | fixing | 22:23 |
| ignas | aelkner: try the buildout now | 22:31 |
| aelkner | one sec | 22:31 |
| aelkner | ok, doing the make now | 22:32 |
| ignas | cool :) this should have been "you can try out buildout now" ;) | 22:32 |
| aelkner | no prob. we are technical people, so we allow each other curt forms of communication :) | 22:33 |
| aelkner | btw, is make wrong? you instructed make run | 22:34 |
| aelkner | i see from the makefile that make means make build | 22:34 |
| ignas | no just that it will only "build" stuff | 22:34 |
| ignas | yep | 22:34 |
| aelkner | so i just make run when the build is done | 22:34 |
| aelkner | then bin/test | 22:34 |
| aelkner | right? | 22:34 |
| ignas | make test should work too | 22:35 |
| ignas | though it just runs bin/test | 22:35 |
| aelkner | ok | 22:35 |
| aelkner | so i looked over all of your diffs, and it's complicated what you are doing. could you summarize? | 22:35 |
| ignas | well - i just moved all the code out of src/schooltool/ | 22:36 |
| ignas | and made schooltool a "namespace" package | 22:36 |
| ignas | which means that you can have more than one egg | 22:37 |
| ignas | that has schooltool in it | 22:37 |
| ignas | like schooltool/jelkner2007 and schooltool/devtools | 22:37 |
| ignas | and schooltool | 22:37 |
| ignas | and they will work together without conflicting about which one has the "right" schooltool package | 22:37 |
| ignas | another part was moving jelkner2007 out of the branch | 22:38 |
| ignas | and making it an egg | 22:38 |
| ignas | so you could develop it while using schooltool egg | 22:38 |
| ignas | and added a buildout that checks out jelkner2007 | 22:39 |
| ignas | and allows you to run it | 22:39 |
| ignas | anything else unclear? | 22:40 |
| aelkner | so schooltool could be fixed in an egg form (release eventually), yet I could effect a change in jelkner2007 at my convenience | 22:40 |
| aelkner | but if i wanted to change the gradebook package? | 22:41 |
| ignas | you will have to work with schooltool trunk | 22:41 |
| ignas | unless i would make schooltool.gradebook a separate egg as well | 22:41 |
| ignas | though i should talk to jinty about that | 22:42 |
| ignas | as someone would have to package schooltool.gradebook | 22:42 |
| ignas | and make it available | 22:42 |
| aelkner | i agree that we wouldn't be risking anything letting me change gradebook in trunk as it would be isolated | 22:42 |
| aelkner | al long as i make sure i don't commit with failing tests | 22:42 |
| ignas | and avoid changing other parts | 22:43 |
| aelkner | right | 22:43 |
| aelkner | i would consult with you if i ever needed changes to other parts | 22:43 |
| aelkner | one thing | 22:43 |
| aelkner | you noticed i needed to change secutiry tests | 22:44 |
| ignas | yes | 22:44 |
| ignas | i saw that | 22:44 |
| aelkner | so if i need to do something like that in the future | 22:44 |
| aelkner | i'll send you a diff, right? | 22:44 |
| ignas | yes, a diff would be nice | 22:44 |
| aelkner | i can work with that | 22:44 |
| aelkner | so i ran make run, and it gave me a deprecation warning | 22:45 |
| ignas | that's ok | 22:45 |
| ignas | it's not in schooltool | 22:45 |
| aelkner | and that it couldn't use kilthread | 22:45 |
| ignas | that's ok too, it's wsgi thing | 22:45 |
| aelkner | also it didn't give the same four lines that it gives at startup | 22:46 |
| ignas | well - it's a different/new way of starting schooltool server | 22:46 |
| ignas | there even are keyword commands like | 22:46 |
| ignas | --start-daemon | 22:46 |
| ignas | --stop-daemon | 22:47 |
| ignas | and --status | 22:47 |
| ignas | for bin/start-schooltool-instance | 22:47 |
| ignas | all the thing specific for your schooltool instance are in "instance/" by default | 22:47 |
| ignas | so you can do: bin/start-schooltool-instance instance/ --start-daemon | 22:48 |
| ignas | to run schooltool in daemon mode | 22:48 |
| ignas | you can also create separate instances in different directories using | 22:48 |
| ignas | bin/make-schooltool-instance directory-for-other-instance | 22:48 |
| ignas | which makes it easier to work with different types of instances, or different databases of the same type | 22:49 |
| ignas | when you will get used to jelkner2007 in a buildout, tell me and i'll explain the next step - having a checkout of schooltool and a checkout of jelkner2007 in the same buildout | 22:49 |
| ignas | so you could make related changes to both of them | 22:49 |
| ignas | and would not have to wait for schooltool egg to get built | 22:49 |
| aelkner | yeah, my first step is to get comfortable enough with this to be able to give instructions to jeff for using it | 22:51 |
| ignas | using it is easy enough | 22:52 |
| ignas | svn co; make run | 22:52 |
| ignas | though if you will see the need for it - add the cando make debian-dependencies or how was it called | 22:53 |
| ignas | to the buildout make file | 22:53 |
| aelkner | i ran bin/test, and it only ran three tests | 22:55 |
| ignas | yep, all the tests you have in jelkner2007 | 22:55 |
| aelkner | right. so it doesn't run schooltool tests because they are run by the developer of the schooltool egg | 22:56 |
| ignas | yep | 22:56 |
| ignas | and you can't break schooltool tests | 22:56 |
| ignas | by things in jelkner2007 ;) | 22:56 |
| ignas | which is one of the upsides of eggs from my standpoint | 22:57 |
| aelkner | even if i try real hard? :) | 22:57 |
| ignas | monkeypatching would probably work out, but even then - only if jelkner2007 *is* included when running schooltool tests | 22:57 |
| aelkner | eggs are supposed to work just like jars in java | 22:58 |
| aelkner | it's a good idea | 22:58 |
| ignas | maybe ;) i have no idea how they are supposed to work ;) | 22:58 |
| aelkner | it's just these issues that jars addressed, so python eventually adopted them | 22:59 |
| aelkner | that's the way it seems to me at least | 22:59 |
| ignas | it's more like a set of generic problems with library packaging, python packages <-> java jars, while eggs are a replacement for perl's CPAN and ruby gems | 23:00 |
| aelkner | ah | 23:02 |
| aelkner | so i copied the Data.fs file from my jelkner-gradebook branch instance to the buildout instance | 23:03 |
| aelkner | make run returned many more deprecation warnings | 23:03 |
| aelkner | all having to do with session | 23:03 |
| ignas | some Zope guys are changing these things | 23:04 |
| ignas | other Zope guys are yelling at them for doing it ;) | 23:04 |
| ignas | SNAFU ;) | 23:04 |
| aelkner | :) | 23:04 |
| ignas | is it working ? | 23:05 |
| aelkner | yes it is. | 23:08 |
| aelkner | the old data works | 23:08 |
| aelkner | obviously you didn't change anything there | 23:08 |
| aelkner | but it's still nice to know that jelkner can move his Data.fs in without any problems | 23:08 |
| aelkner | of course, i need to write a generation script to create a dummy worksheet and move his activies down there | 23:09 |
| aelkner | before i hand this over to him | 23:09 |
| aelkner | ignas: what's the url for checking out using http? | 23:13 |
| ignas | hmm | 23:13 |
| aelkner | i tried replacing svn_ssh with http, but it didn;'t work | 23:13 |
| ignas | now generation might be tricky | 23:13 |
| ignas | svn co http://svn.schooltool.org/schooltool/trunk/schooltool | 23:13 |
| ignas | http://source.schooltool.org/ | 23:13 |
| ignas | so it would be http://svn.schooltool.org/schooltool/trunk/jelkner-buildout | 23:14 |
| ignas | or was it jelkner2007-buildout | 23:14 |
| ignas | but you will probably have to replace the url in the makefile too | 23:14 |
| aelkner | i'll see | 23:14 |
| ignas | because at the moment it is pointing at the ssh url | 23:14 |
| aelkner | should it? | 23:15 |
| aelkner | ignas: should i change the make file? | 23:17 |
| aelkner | and commit the change? | 23:17 |
| aelkner | i mean, there's no need to do the svn+ssh checkout from the buildout, is there? | 23:18 |
| *** jinty has joined #schooltool | 23:22 | |
| *** jinty has quit IRC | 23:22 | |
| *** th1a_ has joined #schooltool | 23:32 | |
| ignas | aelkner: yes you can | 23:33 |
| aelkner | will do | 23:33 |
| ignas | as long as you know that you have to switch to an ssh checkout if you want to work with it | 23:34 |
| aelkner | but would i want to work with the code from there? | 23:34 |
| aelkner | i mean, we're talking about the buildout pulling in another repository | 23:35 |
| aelkner | if i wanted to change the other repo, wouldn't i svn+ssh checkout that myself | 23:35 |
| aelkner | ? | 23:35 |
| ignas | yes | 23:45 |
| ignas | that's what i mean | 23:45 |
| ignas | meant | 23:45 |
| ignas | "make" checks out it automatically | 23:46 |
| ignas | so in a new buildout checkout - you will have to check out the branch before "make" does that, or replace the old checkout with a "working" one | 23:46 |
| ignas | though it is definitely better for a user like Jeff to have it set to http by default | 23:46 |
| aelkner | i'm thinking that jelkner-buildout is only for delivering to jelkner | 23:48 |
| aelkner | where | 23:48 |
| aelkner | if i want to change jelkner007, i do that in a different svn co | 23:48 |
| aelkner | in a different directory altogether | 23:48 |
| aelkner | what it sounds like you're saying is | 23:48 |
| aelkner | that i woudl be coding jelkner2007 from within jelkner-buildout and doing this manual co dance there | 23:49 |
| aelkner | wouldn't it be easier to just svn+ssh co jelkner2007 in a different dir? | 23:49 |
| aelkner | and trunk itself it yet another dir? | 23:50 |
| aelkner | ignas: am i making sense? | 23:52 |
| *** didymo has joined #schooltool | 23:58 | |
| *** jhancock__ has quit IRC | 23:58 | |
Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!