| ignas | ccarey: are you there? | 00:13 |
|---|---|---|
| ccarey | yes i'm here | 00:13 |
| ignas | hi ;) | 00:14 |
| ccarey | hello | 00:14 |
| ignas | do you have any specific idea about what jelkner wants you to do? | 00:14 |
| ignas | because I am not sure I understand his goals for the conference sprint | 00:14 |
| ccarey | no i don't understand the specifics, but i think i get the general idea | 00:15 |
| *** aelkner has joined #schooltool | 00:15 | |
| ignas | aelkner: maybe you know any specifics about what jelkner want's ccarey to do in the conference? | 00:15 |
| aelkner | i saw you email | 00:15 |
| aelkner | actually, he was hoping to get ccarey started right away | 00:16 |
| aelkner | to explain | 00:16 |
| aelkner | cando is hosted on schooltol, yes | 00:16 |
| aelkner | however, its make file then loads schooltol | 00:16 |
| aelkner | svns it | 00:16 |
| aelkner | so its not really part of schooltool. | 00:16 |
| aelkner | whereas jelkner2007 is | 00:16 |
| aelkner | or sla-buildout | 00:17 |
| ignas | emm | 00:17 |
| aelkner | so he would like there to be a cando-builtout | 00:17 |
| ignas | cando is actually using schooltool egg | 00:17 |
| ignas | as for a buildout | 00:17 |
| ignas | I have one working on my machine ;) | 00:17 |
| aelkner | a cando-buildout? | 00:17 |
| ignas | well, yes | 00:17 |
| aelkner | you're too good :) | 00:17 |
| ignas | I have migrated to lxml | 00:18 |
| aelkner | so in other words, what jeff wants is already done | 00:18 |
| ignas | to make it possible | 00:18 |
| ignas | well - it's not available anywhere | 00:18 |
| aelkner | of course | 00:18 |
| aelkner | a small step | 00:18 |
| ignas | but you know - after doing it like 20 times, it takes like 15-20 minutes to get one going ;) | 00:18 |
| aelkner | but the job is 99% done | 00:18 |
| ignas | there are some things to make integration easier | 00:18 |
| aelkner | i bet you're quick at it | 00:19 |
| ignas | like fixing instances | 00:19 |
| ignas | to use schooltool.conf | 00:19 |
| ignas | making make-schooltool-instance work with cando properly | 00:19 |
| ignas | migrate to schooltool trunk | 00:19 |
| aelkner | ok, so maybe it's 85% done | 00:19 |
| ignas | nightlies that is | 00:19 |
| ignas | the important question is | 00:20 |
| ignas | do you like buildouts outside of packages or inside? | 00:20 |
| ignas | buildbot likes them inside ;) | 00:20 |
| ignas | because that saves one checkout | 00:20 |
| aelkner | i don't have an opinion | 00:20 |
| aelkner | yet | 00:20 |
| ignas | you only need to svn co stuff once | 00:20 |
| aelkner | as i'm not litterate | 00:20 |
| ignas | hmm | 00:20 |
| aelkner | i've carefully avoided the issue | 00:20 |
| ignas | i guess i'll just branch cando | 00:20 |
| ignas | and put the buildout inside it | 00:21 |
| ignas | because that's the way I like it | 00:21 |
| aelkner | my opinion is that you are going to know what's right | 00:21 |
| aelkner | and i would follow your suggestions | 00:21 |
| aelkner | as would anyone else | 00:21 |
| aelkner | so if ccarey could work with you | 00:21 |
| aelkner | he would gain the expertise | 00:22 |
| aelkner | he'd probably know more than i by the time he was done | 00:22 |
| aelkner | eventually, yes, i should develop fluency | 00:22 |
| aelkner | as this is an important topic | 00:22 |
| aelkner | but i just wanted to keep my brain cycles free | 00:22 |
| aelkner | for focused tasks like what i'll be doing for sla | 00:22 |
| aelkner | i'm not the greatest multi-tasker | 00:23 |
| ignas | i see | 00:23 |
| aelkner | but if you give me a project | 00:23 |
| aelkner | even a big on | 00:23 |
| aelkner | one | 00:23 |
| aelkner | i have a long career history of delivering | 00:23 |
| aelkner | and within agreed upon target dates | 00:23 |
| aelkner | so that's why i tend to avoid taking on too many varied types of tasks | 00:24 |
| ignas | makes sense | 00:24 |
| aelkner | maybe shortly befroe the sprint i'll get my mind around it | 00:24 |
| aelkner | so that i can contribute | 00:24 |
| aelkner | to the packaging work | 00:24 |
| aelkner | i'm also pushing heavily for the UI work that Tom also wants done | 00:25 |
| ignas | you should just focus on the UI then | 00:25 |
| aelkner | maybe that would be best | 00:26 |
| aelkner | we'll see | 00:26 |
| aelkner | leading up to pycon, we'll be able to plan that stuff | 00:26 |
| aelkner | but we have over a month until then | 00:26 |
| ignas | yep | 00:26 |
| aelkner | so could you please respond to jeff's email | 00:26 |
| aelkner | ccing ccarey as well | 00:26 |
| aelkner | that is, if you approve of the task | 00:27 |
| aelkner | ccardy is a great kid and will be at pycon | 00:27 |
| ignas | can I still keep ccarey if he'll be done with the buildout stuff before pycon? | 00:27 |
| ignas | ;) | 00:27 |
| aelkner | certainly :) | 00:27 |
| aelkner | jeff's loving this | 00:27 |
| aelkner | you've been so great with this kids, and we're going to be a really strong team | 00:28 |
| aelkner | stronger every month | 00:28 |
| aelkner | and keep in mind | 00:29 |
| aelkner | the whole goal of all of that training last summer | 00:29 |
| aelkner | and all of the dollars spent | 00:29 |
| aelkner | was the hope to get at least two or three kids | 00:29 |
| aelkner | who will grow into schooltool pros | 00:29 |
| ignas | yeah, i can recall that | 00:30 |
| aelkner | and we were successful | 00:30 |
| aelkner | so if you continue to use any one of them for anything | 00:30 |
| aelkner | it's only helping us all | 00:30 |
| aelkner | funding is between Tom and the state of Virginia | 00:30 |
| lisppaste5 | ignas pasted "A list of schooltool packages that use buildout" at http://paste.lisp.org/display/54904 | 00:31 |
| ignas | ccarey: if you could checkout a few of these | 00:31 |
| ignas | and look at the files in the top level directory | 00:31 |
| ignas | to see the pattern | 00:31 |
| aelkner | ccarey is here? | 00:31 |
| ignas | it would be very cool | 00:31 |
| ccarey | ok | 00:31 |
| ccarey | yes i'm here | 00:31 |
| ccarey | hi aelkner | 00:31 |
| aelkner | and here i was talking about you behind your back, or so i thought :) | 00:31 |
| aelkner | good thing i only had good things to say :) | 00:32 |
| ccarey | =) | 00:32 |
| aelkner | i'm so glad that you are involved again | 00:32 |
| aelkner | it would have been sad if we couldn't find a way | 00:32 |
| aelkner | i see you already did something with breadcrumbs | 00:33 |
| aelkner | that was more advanced even than anything we had done in the summer | 00:33 |
| aelkner | i was glad to see the progress | 00:34 |
| ccarey | thanks | 00:34 |
| aelkner | it's great that you're coming to pycon | 00:34 |
| aelkner | i told jeff that you would be right for the packaging work | 00:34 |
| aelkner | because of the initiative you showed with the breadcumb work | 00:34 |
| ccarey | speaking of that | 00:35 |
| ccarey | ignas, should i finish up the timetable task before the packaging work? | 00:35 |
| ignas | hmm, the way it's more convenient to you | 00:36 |
| ignas | packaging is easier | 00:36 |
| ignas | ;) | 00:36 |
| ccarey | ok | 00:37 |
| ccarey | i was getting along through the task and i might be ready for functional tests soon | 00:37 |
| aelkner | ignas, thanks. i'm done for the day. see you at the meeting tomorrow. | 00:37 |
| ccarey | bye aelkner | 00:38 |
| aelkner | aelkner has not quit cause he likes to keep the channel open even when he's away :) | 00:38 |
| ignas | ccarey: nice to hear it | 00:38 |
| * Fujitsu ponders submitting his first SchoolTool patch (having just returned from a holiday of much SchoolTool-hacking). | 00:41 | |
| ignas | Fujitsu: don't ponder ;) submit it! | 00:41 |
| ignas | ccarey: as for cando packaging - would you like me to do some of the things, or would you like to do it yourself while being guided (for extra learn by doing)? | 00:42 |
| ccarey | i would prefer the latter, would we have to schedule a meeting on #schooltool? | 00:43 |
| ignas | we can, just think of a time, i am up until at least "now" most of the days | 00:44 |
| ignas | now is 0:44 in here | 00:44 |
| Fujitsu | Hmm... the calendar(.ics|.vfb)? traversers in schooltool.app.rest.app are apparently also doing their stuff for normal HTTP requests, resulting in ToCalendarTraverser not actually be used, and calendar.ics not being iCal. | 00:45 |
| ccarey | ignas: so that would mean for me i would get home from school around 00:00 your time | 00:45 |
| ccarey | which isn't too good | 00:45 |
| ignas | i guess | 00:45 |
| ignas | i can stay up until 3am if you need that though, so there is a 3 hour window | 00:46 |
| ccarey | that would work if it's fine with you | 00:46 |
| ignas | ok, so just think of a week day or week days you'll be working on that so i would know when to log onto irc | 00:47 |
| ignas | Fujitsu: thinking | 00:47 |
| ignas | Fujitsu: if you can get past all the functional tests with a raise "I AM USED APPARENTLY" in the ToCalendarTraverser | 00:48 |
| ignas | then i'd assume that you are most probably right ;) | 00:48 |
| ignas | ccarey: i have created a branch for the buildout integration work, and will set up a buildbot to track the progress | 00:49 |
| Fujitsu | ignas: I've done a similar thing, but manually, not running the tests. Good idea. | 00:49 |
| ccarey | ignas: ok, wednesday and friday evenings are my best options, but i can do monday and tuesday as well | 00:50 |
| ignas | wednesday and friday it is then | 00:50 |
| ccarey | and will we start this week? | 00:51 |
| ignas | but i'll try to have an irc client open during the whole week while at home | 00:51 |
| ignas | ccarey: i don't know, i will set up a buildbot today, and i can try and help you with timetabling task as well | 00:52 |
| ccarey | ok, so i'll be on wednesday and friday this week in case | 00:52 |
| ignas | ok, cool | 00:53 |
| *** medo has quit IRC | 00:54 | |
| ccarey | ignas, it's dinner time for me, thanks | 00:55 |
| Fujitsu | ignas: Is it deliberate that it's trivial to create a booking conflict by adjusting the time/duration of an event after booking a resource to it, or would my patch to stop that be useful? | 00:56 |
| ignas | Fujitsu: you might want to look for bugs in the bugtracker | 00:57 |
| ignas | that relate to resource booking conflicts | 00:57 |
| Fujitsu | I've had a look (I filed one of them, and it's now assigned to me, but it's for a different issue). | 00:57 |
| ignas | i think one of them explains why doing a non-trivial detection of such errors is hard | 00:57 |
| ignas | but yes, i guess it would be usefult | 00:57 |
| Fujitsu | Ah yes, the recurrence one, right. | 00:58 |
| ignas | to at least warn people when we know about a conflict for sure | 00:58 |
| Fujitsu | I was going to try to detect recurrence conflicts eventually, but have for now nastiyl disabled recurrence in my dev branch. | 00:58 |
| Fujitsu | (this branch is going to eventually be deployed for resource booking for my school, so needs to not have conflicts) | 00:59 |
| ignas | i see | 01:00 |
| * ignas would like to get timetabling independent from calendaring just to make precise resource booking possible | 01:01 | |
| ignas | like - have bookings only for 30 minute periods or timetable periods | 01:01 |
| ignas | so the space would be narrow and discrete | 01:02 |
| Fujitsu | Yes, that would be nice. | 01:02 |
| ignas | not many schools need to book a classroom for 179 minutes | 01:02 |
| ignas | starting 23:57 | 01:03 |
| Fujitsu | The booking calendar does somewhat facilitate the timetable period bit. | 01:03 |
| ignas | yes, but it does not enforce it | 01:03 |
| Fujitsu | Though the following form could be simplified in that caase. | 01:03 |
| Fujitsu | Right. | 01:03 |
| ignas | calendar is not needed for booking and timetabling | 01:03 |
| ignas | it's just a way of looking at the state | 01:03 |
| Fujitsu | It is technically, isn't it? But it probably shouldn't be. | 01:03 |
| Fujitsu | Right. | 01:03 |
| ignas | it's not necessary to store the state that way | 01:04 |
| Fujitsu | I guess not, no. (sorry, haven't looked much into the timetable bits) | 01:04 |
| ignas | but it's a pie in the sky thing at the moment | 01:04 |
| Fujitsu | ignas: All of the ftests pass in trunk with the ToCalendarTraverser trap sprung. | 01:08 |
| Fujitsu | So it's not being used to traverse to any kind of calendar. | 01:08 |
| ignas | remove it then i guess | 01:09 |
| Fujitsu | I don't think that's the right solution... It seems wrong to rely on the REST stuff, and it still leaves URLs liike /calendar.ics not doing iCal. | 01:12 |
| ignas | oh | 01:12 |
| ignas | my thinking is getting slow at the moment | 01:12 |
| ignas | could you draft a spec or a functional test | 01:12 |
| ignas | thet would expose the usecases that do not work with the current setup | 01:13 |
| ignas | my memory about these places is a bit hazy at the moment | 01:13 |
| Fujitsu | I'm not sure why existing ftests aren't catching it, so I'll write one that does. | 01:16 |
| ignas | ok, thanks a lot | 01:17 |
| fsufitch | aelkner: ayt? | 02:49 |
| *** ignas has quit IRC | 03:07 | |
| *** ccarey has quit IRC | 04:10 | |
| *** didymo has joined #schooltool | 04:39 | |
| *** pcardune has joined #schooltool | 05:10 | |
| *** alga has quit IRC | 05:34 | |
| *** didymo has quit IRC | 06:07 | |
| *** pcardune has quit IRC | 11:52 | |
| *** didymo has joined #schooltool | 11:55 | |
| *** didymo has quit IRC | 11:55 | |
| *** subir has joined #schooltool | 11:57 | |
| *** didymo has joined #schooltool | 12:53 | |
| *** subir has quit IRC | 13:01 | |
| *** pcardune has joined #schooltool | 13:53 | |
| *** alga has joined #SchoolTool | 14:06 | |
| *** ignas has joined #schooltool | 14:10 | |
| *** pcardune has quit IRC | 14:45 | |
| *** pcardune has joined #schooltool | 15:50 | |
| *** jelkner has joined #schooltool | 15:51 | |
| *** th1a has joined #schooltool | 16:28 | |
| th1a | hi jelkner, aelkner, ignas, pcardune, Lumiere... | 16:29 |
| aelkner | hello | 16:29 |
| ignas | hi | 16:29 |
| pcardune | th1a: say, it this the meeting time? | 16:29 |
| th1a | Yes. 6:30 AM. | 16:29 |
| th1a | right pcardune? | 16:30 |
| pcardune | 9:30 pm here in bangkok... | 16:30 |
| th1a | Ah, still in Thailand. | 16:30 |
| pcardune | one more week | 16:30 |
| th1a | OK, so I guess the agenda today is updates and then what we mean by merging CanDo with SchoolTool. | 16:31 |
| th1a | First off, EduCon was a big success I think, and good for networking the US side of SchoolTool. | 16:32 |
| th1a | If we make Chris happy he'll be directing a lot of users to SchoolTool. | 16:33 |
| aelkner | looks like | 16:33 |
| th1a | OTOH, I was just lying in bed wondering if Zope 3.4 will ever be released and whether I need to start agitating for that (I imagine I do). | 16:33 |
| ignas | th1a: i think they have a release candidate already, if not - they are at least ready to do one | 16:34 |
| th1a | OK, I will try to prod the right people. Remind them that having it ready for Ubuntu is important to us (and perhaps even Zope 3!). | 16:35 |
| th1a | OK, that's what's on my mind. | 16:36 |
| th1a | aelkner? | 16:36 |
| ignas | is the launchpad ppa bug resolved already? | 16:36 |
| aelkner | i missed a couple of days yesterday due to being under the weater | 16:36 |
| aelkner | but i did get started on what they need at sla | 16:36 |
| th1a | I think it is "committed" but I'm not sure if it is deployed. | 16:36 |
| th1a | ignas ^ | 16:37 |
| ignas | th1a: i see | 16:37 |
| th1a | We appear to have our first set of deliverables defined. | 16:37 |
| aelkner | yes, the Narrative Report Cards and the Student Intervention (CSAP) stuff | 16:38 |
| aelkner | that and getting their students and teachers imported | 16:39 |
| aelkner | together with the extended demographics | 16:39 |
| aelkner | so, they will be using schooltool on a regular basis by April 15 | 16:39 |
| aelkner | even if their course stuff will wtill be handled in moodle | 16:40 |
| th1a | I'm happy with the goals. | 16:40 |
| aelkner | i know they will be glad to have the CSAP stuff automated | 16:40 |
| aelkner | as that is a real pain for them as it is now | 16:41 |
| th1a | Do aelkner and ignas need to talk more about how we're doing extended demographics. | 16:41 |
| th1a | ? | 16:41 |
| ignas | we have already talked about it I think | 16:41 |
| aelkner | actually, our chat friday has left me clear on what needs doing | 16:41 |
| ignas | but if there are any questions - I am always available | 16:41 |
| th1a | OK. | 16:41 |
| aelkner | so i got nothing else | 16:42 |
| aelkner | a short report this week | 16:42 |
| th1a | thanks aelkner. | 16:42 |
| th1a | ignas? | 16:42 |
| th1a | Actually, what are you doing first, aelkner? | 16:42 |
| ignas | not much, mostly continued doing some fixes to the buildbot/buildout infrastructure | 16:42 |
| ignas | added sample data for schooltool.stapp2008spring | 16:43 |
| ignas | that would generate grades and courses for 4 terms | 16:43 |
| ignas | (2 school years) | 16:43 |
| ignas | with each section spanning 1 term | 16:43 |
| ignas | and teachers teaching only 1 course with multiple sections | 16:43 |
| ignas | so if you will want to you can just check out stapp2008spring and try running it + generating sample data | 16:44 |
| th1a | It should be easier for me to make some screenshots now. | 16:44 |
| ignas | I am not too happy with the grade distribution at the moment though ;) there are as many "1" as "10"'s | 16:44 |
| ignas | but it's good enough at the moment | 16:44 |
| th1a | Need that bell curve. | 16:44 |
| aelkner | ignas: are you referring to my sample data? | 16:45 |
| ignas | no, i am referring to my sample data | 16:45 |
| ignas | schooltool.lyceum.journal | 16:45 |
| aelkner | you could take a look at what i did | 16:45 |
| aelkner | i made the grades go between 40 and 100 | 16:45 |
| aelkner | to make it look as much real-world as possible | 16:45 |
| ignas | real world woud have different persons have different averages though | 16:46 |
| ignas | and at the moment it's all around the same number | 16:46 |
| aelkner | i have different averages for everyone | 16:46 |
| ignas | nice | 16:46 |
| aelkner | it's random | 16:46 |
| ignas | hmm? | 16:47 |
| ignas | different as in - some people have 99% | 16:47 |
| ignas | and some have 52% | 16:47 |
| aelkner | right | 16:47 |
| ignas | I'll look at it | 16:47 |
| th1a | what else, ignas? | 16:48 |
| ignas | i think that's everything i did this week | 16:48 |
| aelkner | i didn't want to interrupt ignas | 16:48 |
| aelkner | but tom asked what i planned doing first | 16:48 |
| aelkner | basically, i'll get the demographics data in | 16:48 |
| aelkner | then, the Narrative Report Cards | 16:49 |
| aelkner | then CSAP | 16:49 |
| th1a | Makes sense. | 16:49 |
| th1a | ignas: What's coming up for you? | 16:51 |
| ignas | well - as I have the sample data ready I'll be going through schooltool | 16:52 |
| ignas | and looking for places to improve | 16:52 |
| ignas | like section lists | 16:52 |
| ignas | calendar overlays | 16:52 |
| ignas | links to the journal | 16:52 |
| ignas | mostly polishing to make 2008spring release usable | 16:52 |
| ignas | I would be glad if you would look at it too | 16:53 |
| th1a | Ah. OK. | 16:53 |
| th1a | Yes, I will. | 16:53 |
| ignas | and give some criticism | 16:53 |
| ignas | because the thing you will be seeing is what our users will be looking at | 16:53 |
| th1a | It is a lot easier to generate that criticism if I know you're going to be working on fixing those kind of problems. | 16:53 |
| ignas | i'll work on cando schooltool integration as well apparently | 16:54 |
| ignas | at least as a consultant ;) | 16:54 |
| th1a | I think we're to that point in the conversation. | 16:54 |
| th1a | jelkner: ayt? | 16:55 |
| th1a | So anyway, are we all getting to the same page about what "integration" means? | 16:56 |
| * pcardune raises his eyebrows | 16:57 | |
| th1a | Not merging CanDo into the SchoolTool monolith because we're disassembling the SchoolTool monolith. | 16:57 |
| ignas | true | 16:57 |
| ignas | my plans are: | 16:57 |
| ignas | refactor cando to act as schooltool.stapp2008spring | 16:57 |
| ignas | tizard | 16:57 |
| ignas | sla | 16:57 |
| ignas | and lyceum | 16:57 |
| ignas | add "plugins" related scripts to schooltool.paste | 16:58 |
| ignas | remove cando.paste and make cando use schooltool for instance-making | 16:58 |
| ignas | launch cando test and cando release buildbots | 16:58 |
| ignas | that use buildout to do these tasks | 16:58 |
| ignas | maybe even make it possible to refactor cando packages out of cando to be used separately | 16:59 |
| ignas | but that depends on the plans of cando project | 16:59 |
| th1a | Can you do the above pretty quickly at this point? | 17:00 |
| ignas | quite quickly i think, but i will be guiding ccarey rather than doing it myself | 17:00 |
| ignas | so that cando would have one person who knows the stuff | 17:00 |
| ignas | instead of "magic happens" and you get buildout | 17:01 |
| th1a | OK. Good. | 17:01 |
| ignas | the parts i am worried about are the - merge projects and get cando code reviewed parts that it seems jelkner is expecting | 17:01 |
| ignas | seems to me | 17:01 |
| aelkner | yeah, he was hoping to have you review the whole cando code base | 17:02 |
| * pcardune ducks | 17:02 | |
| aelkner | for things like test coverage | 17:02 |
| ignas | i mean - one of the points of splitting schooltool into parts was to make it less code to review and manage for me ;) | 17:03 |
| aelkner | and PEP 8 compliance | 17:03 |
| aelkner | that sounds fair | 17:03 |
| ignas | like - only aelkner needs schooltool.gradebook , so if it is a separate package - it's one thing i don't have to think that much about | 17:03 |
| th1a | I think the main thing is that they need to learn what ignas expects and try to clean up to those standards BEFORE ignas looks at it. | 17:03 |
| aelkner | good point | 17:03 |
| ignas | because apparently all the details and features just don't fit into my head, and i am not sure I can perform a quality review | 17:04 |
| th1a | (also, I'm willing to accept the idea that this is just a bad idea) | 17:04 |
| ignas | without knowing what cando does up to every single feature | 17:04 |
| th1a | Hm... | 17:04 |
| aelkner | that wouldn't be expected | 17:04 |
| aelkner | i think jelkner just wants your expertise to guide the process of getting cando | 17:05 |
| ignas | i can say PEP8 here, PEP8 here, but making meaningful suggestions is difficult unless I know future plans, requirements and etc. | 17:05 |
| aelkner | to comply with your coding standards | 17:05 |
| aelkner | really, it's the cando team's responsibility to worry about those things | 17:06 |
| th1a | This might not be realistic. | 17:06 |
| aelkner | agreed | 17:06 |
| aelkner | jelkner was trying to get something cheep :) | 17:06 |
| aelkner | cheap | 17:06 |
| ignas | i am afraid it won't comply :/ schooltool does not :/ it's a never ending battle ... I can try to imbue the same way of thinking the same criteria | 17:06 |
| aelkner | i think if you guide ccarey | 17:07 |
| aelkner | to review the code base himself | 17:07 |
| th1a | Well, I'd also say that if you could really just bring people in to do code reviews and easily fix things afterward, software would be a lot more reliable in general. | 17:07 |
| ignas | I can try | 17:07 |
| ignas | but it's up to them to keep the standards at some level | 17:07 |
| ignas | I think what Robert Collins suggested when talking about such codebases was: | 17:08 |
| ignas | Make sure that every commit you make is better than the average quality of the codebase. | 17:08 |
| ignas | in 3 years - the overall quality will be a lot better ;) | 17:08 |
| ignas | and i am afraid that it's the only way to do it - every time before commiting think whether the part that you are adding is better than the rest of the stuff | 17:09 |
| ignas | and never say - i | 17:09 |
| ignas | i'll improve it later | 17:09 |
| aelkner | cando's only issues are PEP 8 and test coverage | 17:09 |
| th1a | There seem to be PEP 8 checking scripts. | 17:10 |
| pcardune | I don't think PEP 8 is that big a deal, but test coverage yes... is a much bigger issue | 17:10 |
| ignas | aelkner: this is the part where I would disagree ;) | 17:10 |
| th1a | ignas: Right. Code review can also say the design is flawed. | 17:10 |
| th1a | Right? | 17:10 |
| ignas | which is why i think that pulling cando up is very difficult - if aelkner thinks it's allright - he can't fix it, you can only fix things that you think are broken ;) | 17:11 |
| ignas | th1a: yes, but to say that i must actually understand what was being done | 17:11 |
| th1a | Right. | 17:11 |
| jelkner | sorry, i'm back | 17:11 |
| ignas | th1a: which is a lot of work, and i am lazy ;) | 17:11 |
| th1a | It is a lot of work. | 17:11 |
| aelkner | jelkner: were your ears burning? | 17:12 |
| jelkner | we have money to pay young coders to fix things | 17:12 |
| jelkner | but they need adult supervision ;-) | 17:12 |
| aelkner | ignas shouldn't have to learn the inner workings of cando | 17:12 |
| aelkner | he has enough on his mind | 17:13 |
| th1a | http://svn.browsershots.org/trunk/devtools/pep8/pep8.py | 17:13 |
| ignas | pcardune: are there design decisions in cando that you would do differently in retrospect? | 17:13 |
| jelkner | that means aelkner would have to do it | 17:13 |
| jelkner | we need to: | 17:13 |
| jelkner | 1. identify clearly what needs to be fixed | 17:14 |
| * aelkner has his hands full with sla :) | 17:14 | |
| jelkner | 2. communicate that to chris carey who is willing to work on it | 17:14 |
| jelkner | 3. look over what he does and give him regular feedback so he can do it well | 17:14 |
| th1a | jelkner: The more we discuss this the less trivial the whole process sounds. | 17:14 |
| jelkner | th1a: ok, but it is *important* | 17:15 |
| jelkner | so what do you suggest? | 17:15 |
| th1a | OK, why is it important? | 17:15 |
| aelkner | may i suggest that cando is fine save for test coverage | 17:15 |
| jelkner | because if we don't handle it effectively now, then we will get bit in the ass when it comes time to deploy it later | 17:16 |
| th1a | So you just need to teach chris to check for test coverage and write tests. | 17:16 |
| ignas | aelkner: tests and pep8 are the things that we can fix and should fix, i agree with that | 17:16 |
| aelkner | ccarey could learn to be that expert for cando | 17:17 |
| ignas | aelkner: but the difference between cando and schooltool codebases is a bit bigger than that, even though schooltool codebase needs a lot of fixing as well | 17:17 |
| aelkner | it's a focused taskl | 17:17 |
| aelkner | that he could take weeks to do | 17:17 |
| th1a | They both can be at least partly checked by automated tools. | 17:17 |
| aelkner | ignas: what bigger differences? | 17:18 |
| ignas | aelkner: from what i have seen in cando, there are some things that could be done better, you can write the same things in more than one way most of the time | 17:19 |
| ignas | and in cando - there are places that were written without knowing some parts of Zope3 or python | 17:19 |
| ignas | and yes I understand that it works | 17:19 |
| ignas | but to me it's an inherent part of "codebase quality" | 17:19 |
| aelkner | i think in large projects, there has to be acceptance from time to time of code that could "be done better" as long as it is not done wrong | 17:19 |
| jelkner | ignas: you see "code smells" in cando and you are suggesting they be refactored, yes? | 17:20 |
| ignas | yes | 17:20 |
| aelkner | for instance? | 17:20 |
| jelkner | would it take too much of your time to communicate to to chris carey what they are? | 17:20 |
| ignas | but i think it's difficult to do unless people who are fixing them see it too | 17:20 |
| jelkner | ahh | 17:21 |
| ignas | it's just different perspectives | 17:21 |
| ignas | and i have come to think | 17:21 |
| jelkner | i understand the constraints we are all under in terms of time, workload, etc. | 17:21 |
| ignas | that cando should not be changed | 17:21 |
| ignas | i mean - if most of cando developers think that it's fine | 17:21 |
| ignas | why make them do it the way *I* like | 17:21 |
| ignas | so that's one of the reasons why I moved schooltool.gradebook from schooltool trunk | 17:22 |
| jelkner | because "you the man!" | 17:22 |
| jelkner | we need to have a review process that works | 17:22 |
| jelkner | we can't have too many different folks doing too many different approaches | 17:22 |
| jelkner | that would be bad | 17:22 |
| jelkner | particularly when working with young developers who need guidance | 17:23 |
| th1a | Well, I'm afraid this is how software development works in the real world. | 17:23 |
| th1a | Although I've never been in the real world. | 17:23 |
| ignas | yes, but you have to find the approach you like yourselves, because most of the things i have learned were learned by arguing with mgedmin for hours to decide which implementation of a feature would be the most elegant one | 17:23 |
| jelkner | i'm *very* comfortable with the idea that -we are a small group of developers and ignas is the ultimate decision maker on code questions | 17:24 |
| ignas | and all that i have learned was - my code sucks, i must work hard to improve ;) | 17:24 |
| ignas | jelkner: to make decisions I must know cando better than you do | 17:24 |
| ignas | which I just don't think is realistic | 17:24 |
| ignas | the people who have the most information should make decisions | 17:24 |
| ignas | even if they will not make the perfect one | 17:24 |
| ignas | they will learn | 17:24 |
| jelkner | that's fine too | 17:24 |
| aelkner | if schooltool is to be this apt-get installable thing, people all around the world could be developing their own plugins | 17:24 |
| jelkner | that would be your decision to delegate | 17:25 |
| aelkner | there's no way for anyone to review it all | 17:25 |
| jelkner | but then you would, i assume, let their code into schooltool | 17:25 |
| jelkner | that's what i want | 17:25 |
| jelkner | to be part of schooltool | 17:25 |
| ignas | nope ;) | 17:25 |
| aelkner | jelkner: you missed ignas' point earlier | 17:25 |
| ignas | no, really, i am trying to make schooltool smaller not bigger | 17:25 |
| aelkner | of tearing down the code monolith | 17:25 |
| ignas | to keep it maintainable | 17:26 |
| ignas | to keep separate parts | 17:26 |
| jelkner | i understand that | 17:26 |
| th1a | jelkner: Yes, ignas is right that we've been working on changing what it would mean to be "part of schooltool." | 17:26 |
| ignas | that anyone can work on, no matter how much I disaggree with their definition of quality | 17:26 |
| jelkner | what i mean by "part of schooltool" | 17:26 |
| jelkner | is shared responsibility for the code | 17:26 |
| jelkner | whatever packaging and distribution solution you come up with | 17:26 |
| jelkner | i want the concerns of cando to part of the calculus when thinking about schooltool | 17:27 |
| th1a | jelkner: I think that could be a gradual process. | 17:27 |
| jelkner | not a "that's there problem" | 17:27 |
| jelkner | th1a: we have *never* been able to do a package of cando | 17:28 |
| ignas | if our egg -> deb magic works that should be trivial | 17:28 |
| jelkner | great | 17:28 |
| ignas | but it's more "project goals thing" than code thing | 17:28 |
| th1a | jelkner: We're in really good shape as far as packaging. | 17:29 |
| jelkner | awesome | 17:29 |
| jelkner | then what will prevent cando from being released along with schooltool? | 17:29 |
| ignas | what do you mean by "along"? | 17:29 |
| jelkner | just like gradebook, sla build-out, etc.? | 17:29 |
| ignas | nothing I think, there are some small tasks to be performed | 17:30 |
| jelkner | as one of the components that are available with schooltool | 17:30 |
| jelkner | can chris carey do those tasks? | 17:30 |
| ignas | yes, that's the plan | 17:30 |
| jelkner | awesome | 17:30 |
| jelkner | what's the problem then? | 17:30 |
| th1a | The one big code review appears to be the problem. | 17:31 |
| ignas | understanding of responsibilities and expectations from both sides of the project? ;) | 17:31 |
| jelkner | yes | 17:31 |
| jelkner | i'm trying real hard to tear down the wall | 17:31 |
| th1a | It is not an artificial wall. | 17:31 |
| jelkner | th1a: explain | 17:32 |
| th1a | We just can't expect ignas to learn CanDo in the next six months. | 17:32 |
| jelkner | understood and reasonable | 17:32 |
| jelkner | does he trust anyone else enough to do the first round of review for him? | 17:33 |
| ignas | jelkner: there aren't any choices except aelkner | 17:33 |
| th1a | It still depends on what we're talking about. | 17:33 |
| ignas | you see - schooltool project is "Tom, Ignas and Alan" | 17:33 |
| th1a | If we specifically say: | 17:33 |
| th1a | 1) clean up PEP 8 implementation; | 17:33 |
| th1a | 2) improve test coverage. | 17:34 |
| th1a | Those are pretty mechanical. | 17:34 |
| jelkner | good | 17:34 |
| th1a | It is more a matter of attention to detail than needing a lot of advice. | 17:34 |
| aelkner | agreed | 17:34 |
| jelkner | aelkner recommended ccarey to me because he thinks ccarey is strong there | 17:34 |
| th1a | jelkner, you're a programming teacher. You should be able to do those things. | 17:35 |
| th1a | I mean, you're teaching test-driven development, right? | 17:35 |
| jelkner | th1a: i'm not even a programmer | 17:35 |
| jelkner | just a teacher :-( | 17:35 |
| aelkner | those you can do... | 17:36 |
| th1a | PEP 8 is just grammar. | 17:36 |
| th1a | Get out your red pen! | 17:36 |
| jelkner | i'm very familiar with pep 8 | 17:36 |
| jelkner | chris can do that by just being asked to do it | 17:36 |
| aelkner | and test coverage is something a user can get their hands around | 17:36 |
| jelkner | that's not a problem | 17:36 |
| jelkner | test coverage is more difficult | 17:37 |
| jelkner | but chris worked with aelkner this summer learning how to do that | 17:37 |
| aelkner | so he could master the app from a user's perspective (with jelkner's help) | 17:38 |
| aelkner | and then make sure the test coverage is there | 17:38 |
| th1a | OK... I have to get ready to fly back to Providence. | 17:38 |
| th1a | You guys can keep chatting. | 17:38 |
| th1a | Have a good week! | 17:38 |
| * th1a drops the bag of gravel. | 17:38 | |
| jelkner | tell jennifer it was great meeting her | 17:38 |
| *** th1a has quit IRC | 17:38 | |
| jelkner | aelkner: sure, wbrady can do that better than me | 17:39 |
| jelkner | so that isn't a problem either | 17:39 |
| aelkner | sounds like a plan | 17:39 |
| jelkner | so, the last practice question is: | 17:39 |
| jelkner | do i tell chris to contact ignas or aelkner for instructions | 17:40 |
| jelkner | ? | 17:40 |
| aelkner | how about netiher? have him contact will | 17:40 |
| aelkner | learn the app | 17:40 |
| aelkner | adn get writing tests | 17:41 |
| ignas | jelkner: i don | 17:41 |
| jelkner | bad answer | 17:41 |
| ignas | don't know, i have talked to ccarey yesterday | 17:41 |
| ignas | and we are planning meeting online wednesday and friday | 17:41 |
| jelkner | excellent | 17:41 |
| ignas | to talk about his current schooltool task | 17:41 |
| ignas | and about cando buildout integration | 17:41 |
| jelkner | awesome! | 17:41 |
| jelkner | problem solved | 17:42 |
| jelkner | ignas: you still here? | 17:57 |
| ignas | yes, kind of ;) | 17:57 |
| jelkner | i just had a good conversation on the phone with aelkner | 17:58 |
| jelkner | this is really an issue for th1a and i to work out, but here is the bottom line: | 17:58 |
| jelkner | when schooltool goes into ubuntu as a core package | 17:58 |
| jelkner | i want cando to go with it | 17:59 |
| jelkner | i think stating the goal that way helps provide a bit of clarity | 17:59 |
| jelkner | no code is perfect | 17:59 |
| jelkner | and all code can improve | 17:59 |
| jelkner | but do we feel it is ready enough to "stand by it" in that context | 17:59 |
| jelkner | that would be a good test | 18:00 |
| jelkner | however you setup the packaging system is fine with me | 18:00 |
| jelkner | i just want cando to be one of the packages | 18:00 |
| jelkner | does that make sense? | 18:00 |
| ignas | thinking | 18:02 |
| jelkner | take your time | 18:03 |
| ignas | i think it makes sense, at least it fits with what I was planning to do with cando | 18:07 |
| jelkner | great | 18:07 |
| jelkner | i'll follow up with tom tomorrow | 18:07 |
| jelkner | thanks! | 18:07 |
| ignas | glad to help | 18:08 |
| * jelkner prepares to sign off so he can do his grading... | 18:08 | |
| ignas | ok, bye | 18:08 |
| *** jelkner has quit IRC | 18:08 | |
| *** alga has quit IRC | 18:34 | |
| *** mgedmin has joined #schooltool | 18:34 | |
| fsufitch | aelkner: hi | 19:28 |
| fsufitch | ayt? | 19:29 |
| *** jfluhmann has joined #schooltool | 19:48 | |
| *** jfluhmann has left #schooltool | 19:49 | |
| *** povbot has joined #schooltool | 20:42 | |
| *** medo has joined #schooltool | 21:11 | |
| *** ignas has quit IRC | 21:25 | |
| *** pcardune has quit IRC | 21:47 | |
| *** ignas has joined #schooltool | 22:10 | |
| *** mgedmin has quit IRC | 22:28 | |
| *** lisppaste5 has quit IRC | 22:56 | |
| *** lisppaste5 has joined #schooltool | 23:07 | |
| *** didymo has quit IRC | 23:19 | |
| *** ignas has quit IRC | 23:22 | |
| *** medo has quit IRC | 23:47 | |
Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!