IRC log of #schooltool for Tuesday, 2006-05-23

*** th1a has quit IRC00:02
*** th1a_ has quit IRC00:06
*** didymo has joined #schooltool01:37
*** jinty has joined #schooltool02:27
*** jinty has quit IRC04:40
*** th1a has joined #schooltool06:53
*** jinty has joined #schooltool11:05
*** faassen has joined #schooltool11:35
*** alga has joined #SchoolTool12:02
*** Aiste has quit IRC12:33
*** Aiste has joined #schooltool12:54
*** thisfred has joined #schooltool13:16
*** th1a has quit IRC13:17
*** mgedmin has joined #schooltool13:29
*** Pupeno has joined #schooltool13:34
PupenoGood morning.13:34
povbot/svn/commits: * jinty committed revision 6087:13:44
povbot/svn/commits: Add the paths as sites as well as appending to sys.path. This is so that eggs can be found from instances.13:44
faassenPupeno: hello14:17
Pupenofaassen: hi.14:17
*** PupenoK has joined #schooltool14:17
*** PupenoK has left #schooltool14:19
*** Pupeno is now known as PupenoK14:20
*** PupenoK is now known as Pupeno14:21
*** tiredbones has quit IRC14:49
*** tiredbones has joined #schooltool14:50
*** jinty has quit IRC15:00
*** ignas has joined #schooltool15:03
*** Pupeno_ has joined #schooltool15:58
*** jinty has joined #schooltool16:03
*** th1a has joined #schooltool16:08
*** Pupeno has quit IRC16:11
*** Pupeno_ is now known as Pupeno16:25
* th1a shuffles some papers around.16:29
* mgedmin coughs discreetly16:29
* faassen messes around with a bag of gravel.16:29
th1aGood day everyone.16:30
th1asrichter:  You awake?16:30
ignasfaassen, why can't i see students in the table.html view on demo.schooltool.org16:30
faassenignas: I don't know. let me look.16:30
faassenwhat's the demo password?16:31
th1ajinty:  Are you here?  There?16:31
th1aOh... what permission do you need to see the table?16:31
faassenI have no idea.16:32
ignasfaassen, student001, student00116:32
faassenI think schooltool.view.16:32
th1aSo a student or teacher should be able to see it?16:32
faassenI think so, but I haven't done anything about security whatsoever.16:33
faassenignas: that's mysterious.16:33
faassenignas: I haven't seen that failure condition before. weirdly enough the batching works.16:33
ignassearch has that too16:33
faassenignas: it could be a permissions issue.16:33
ignasor hasn't16:33
th1aIt seems likely.16:34
faassensearch doesn't work either.16:34
faassenperhaps zc.table filters and doesn't show things you aren't supposed to view.16:34
faassenthough I thought it's all schooltool.view16:34
faassenyou could try with a user that's got schooltool.edit16:34
th1aI don't recall the manager password.16:34
th1ajinty knows it.16:35
faassenI haven't done any testing concerning access with this code.16:35
faassenI've just tested with schooltool's manager user.16:35
faassenbut it could be I have some thing tweaked wrong. still, mysterious that it doesn't show a thing.16:35
povbot/svn/commits: * faassen committed revision 6088:16:35
povbot/svn/commits: Refactored utility setup code so it's cleaner and less repetitious and is tested.16:35
povbot/svn/commits: Use it to set up intids and catalog in the demographics package now.16:35
faassenthere is no use of checkPermission in zc.table.16:36
faassenbut it would be something more subtle is going on.16:37
th1aOK.  I posted a quick note to the list so we don't confuse people further.16:37
th1aWell, let's go on, and have some updates.16:38
th1afaassen, would you like to go ahead?16:38
th1aIt seems like things are progressing.16:38
faassenwell, I finally got some fairly uninterrupted time to work on schooltool the last weeks.16:38
faassenalso, we hired someone to help me work on the project and he's been helping me the end of last week and will help for the coming period.16:39
faassenhe's on the channel, Pupeno16:39
th1aThat's good.16:39
th1aHi Pupeno!16:39
PupenoHello everybody.16:39
faassenso part of my time this week was pairing with him; we had him flown over here (from Spain). he's back now but he's working with me still.16:40
jintyth1a: pong16:40
faassenfor the time being he'll feed the patches through me and I'll review them before they get checked in.16:40
th1ajinty:  Glad you could make it.16:40
jintysorry i'm a bit late, lost track of time up writing a letter...16:41
faassenanyway, concerning the demographics work.16:41
faassenhm, lots of smaller UI tweaks that are still ongoing.16:41
faassenhooked in various new egg dependencies to get the catalog operational.16:42
faassengot the appropriate intids and catalog utilities installed. that was a lot of painful debugging as the schooltool setup procedure is somewhat unique.16:42
faassensince then I've refactored the setup code further so we have an easy way to register utilities.16:43
faassenit's just been checked in as schooltool.utility16:43
faassenthat is, local utilities.16:43
faassenanyway, got the search screen to work.16:43
faassenestablished a new widget for date entry (and noticed problems with it), and got a widget for password entry ported from document library.16:44
jintythis reminds me, we need to figure out if we want eggs to be downloaded and installed automatically for tarball users.16:44
faassenbeen moving some of the more generic UI related code (formlib stuff) into schooltool.skin16:44
ignasjinty, an interactive approach maybe ? make eggs?16:44
* faassen coughs.16:44
faassenanyway, I also sent a few comment and questions to the list.16:45
faassenI also started writing a document on my customization thoughts concerning schooltool with demographics in particular.16:45
faassenbut I'm not quite happy with it all yet.16:45
faassenI want to discuss one topic here concerning utility-based customization.16:45
faassenwe're preparing to replace the index.html of persons/ soon with the table.html based layout.16:46
faassen(independent from customization)16:46
faassenas we're almost at feature parity (plus got some features of our own)16:46
faassenfirst of course we need to reproduce the case of the mysteriously invisible table.16:47
th1afaassen:  If you want to get into a longer discussion we should get POV's report first.16:47
faassenth1a: yes, let's do the rest first. anyway, I also want feedback from someone on the UI.16:47
faassenth1a: what columns are needed, myriads of UI tweaks, etc.16:48
th1ath1a:  Do you have time after the meeting?16:48
faassenth1a: I presume th1a does, but I do too. :)16:48
th1aTalking to myself again.16:48
faassenI do that too.16:48
faassenanyway, that was my report. in summary, finally gaining some steam.16:48
th1aOK.  ignas/mgedmin?16:48
faassenall the preparatory work is starting to show for itself.16:49
Pupenowhat/who is POV ?16:49
ignasProgrammers of Vilnius16:49
faassenProgrammers of Vilnius, that's ignas and lots of people.16:49
faassenmost people except th1a and srichter that do checkins.16:49
ignasme, mgedmin, vidas, gintas, alga, Aiste ...16:49
faassenright, endless amounts of lithuanians. great eurovision entry by the way. :)16:49
ignaswell, I have finished estimating and planning Access control story (at last)16:50
ignaswe have implemented the core of the new security policy16:50
ignascreated a branch for migration16:50
ignasform the old one to the new one16:50
ignasand made functional tests pass (or commented outdated ACL view tests so we would not forget to replace them with new ones)16:51
ignaswe are using the crowd model described by SteveA in the last europython16:51
th1aHave I lost my connection here?16:52
ignaswith some parts tuned to be schooltool specific16:52
faassenth1a: ignas is doing a lot of talking.16:52
faassenth1a, can you hear us?16:52
th1aNow I can.16:53
ignasthats kind of it16:54
ignasfaassen, you can take a look at my last check in to get a grasp of what you have avoided ;)16:54
th1aAh yes.  You had told me about the crowds when I was in Vilnius.16:54
faassenignas: what have I avoided? :)16:54
th1aignas:  How do you feel about the decision now?16:54
ignasth1a, same way gintas does - it's a lot of work, but it has to be done earlier than later16:55
ignasany delay would complicate matters even more16:55
ignass/than/rather than/16:55
th1afaassen:  OK, so you have some concerns about utilities?16:56
faassenignas: do you mean avoided work on the security policy or what?16:56
faassenignas: I'm trying to figure out what I avoided. :)16:56
th1aI think that's what he meant.16:56
ignasfaassen, yes :)16:56
faassenth1a: it's not really a concern.16:56
faassenignas: I didn't avoid that actively. :)16:56
faassenwell, it ties into the customization discussion.16:56
faassenit's a half-question.16:57
faassenanyway, customization by extension packages in Zope 3 can be done using a number of different strategies.16:57
faassenone strategy I'll call 'zcml overrides', the other strategy I'll call 'skin + local utilities'16:57
faassenthough of course they're not mutually exclusive.16:57
faassenzcml overrides is the idea you have an extension package, you add a lot of extra ZCML, override existing ZCML in the core and perhaps even have to disable some ZCML in the core..16:58
mgedminI've tried zcml overrides and decided it is a death-trap16:58
faassento override views, add annotations to effectively change the schema of content objects, etc.16:58
faassenyeah, I don't like zcml overrides either, though I'm also including 'zcml additions'16:58
faassenbut any approach that really wants to use this is probably going to have to both add and override existing ZCML.16:59
faassenthen there's the skin + local utilities approach.16:59
faassenI'm quite familiar with the local utilities pattern, as we use it in Silva, and CMF uses it too.16:59
faassenbasically you have a 'core' schooltool that installs default utilities for various things.16:59
faassen(for instance, in case of demographics, creating 'person' objects)16:59
faassenand then you have an extension (Schooltool US, say)17:00
faassenthat replaces the default local utility with one of its own, that makes sure that Schooltool US Person objects are created.17:00
faassenthat's just one example. it could be any policy, not just object creation.17:00
faassenanyway, utilities can take care of content creation pluggability, skin tend to take care of user interface pluggability.17:01
faassenskins are of course also a familiar Zope 2 style pattern used in many contexts (CMF-based sites in particular)17:01
faassenanyway, I was just pondering the installation of extensions.17:01
faassenthe advantage of a ZCML override approach is that you hook in the extension, done.17:01
faassenno ZODB changes needed.17:01
faassenbut swapping in new local utilities takes ZODB changes.17:02
faassenand it's unclear what a good pattern for that is. does it need a UI? does it happen automatically when you start up schooltool?17:02
faassenanyway, these are points to consider.17:02
faassenwhat I'll do in any case is demonstrate the factory-utility based approach for Person objects in the schooltool core.17:03
faassento control creation of Person objects centrally.17:03
ignasi was kind of thinking that overrides might be avoided most of the time by providing enough sockets/hooks to plug in17:03
ignasa.k.a. subscribers/viewlets17:03
ignasinstead of overrides17:03
faassenthat's something I hadn't thought of yet. I tried to study what's going on with viewlets in schooltool but I'm still somewhat lost. I get the basic idea.17:04
faassenbut the implementation is an awful lot of ZCML.17:04
faassenwhich seems to feature schooltool-specific ZCML attributes too in the viewlet registration business, if I read the code right.17:04
ignasyes but it seems way friendlier on unit tests, and customizability17:04
ignasYMMV though17:05
faassenI agree that it would be more friendly than just plain ZCML overrides.17:05
th1aOK.  We should probably move on.17:05
th1aBut that's definitely something that will have to be worked out in the longer run.17:06
faassenI'm also considering programmer friendliness.17:06
faassenwhile ZCML + viewlets + annotations are worthwhile technologies undoubtedly.17:06
faassensubclassing + skins are probably more familiar technologies for the average Python programmer.17:06
faassenanyway, we'll see.17:07
th1aI think some experimentation is justified.17:07
th1aNow... our upcoming alpha release.17:07
jintyer, I guess that's my cue17:07
th1afaassen:  What's your time frame for finishing this up?17:08
faassenth1a: hah, what do you need to have done? :)17:08
th1aCan we think about starting to package a week from today?17:08
faassenth1a: yeah, I'm planning for that.17:08
faassenth1a: note that thursday is a day off here, but I already calculated that in.17:08
th1aIt seems like we're that close.17:09
faassenth1a: there are a lot of smaller todo items, but I expect we'll be done with hopefully most if not all o fthem next week17:09
faassenth1a: and I'm planning to focus on schooltool until then too.17:09
ignaserr, i am not sure about having access control 100% ready in a week17:09
th1aignas:  I'm not expecting that.17:09
th1aJust don't leave the trunk in a broken state a week from today.17:10
ignasi am working on a branch at the moment17:10
th1aSo perhaps you'll need to plan on staying in the branch until after next Tuesday?17:10
th1ajinty:  What else do we need to consider?17:11
th1aHow were're going to handle all our new external packages?17:11
th1aThe goal here is to have a SchoolTool tarball and .debs.17:12
faassenjinty: I saw you did some egg related checkin today, what did that mean?17:12
faassenjinty: I tried moving eggs into src from Zope3/src but failed (as per schooltool-dev)17:12
faassen(also on the external packages front)17:12
th1ajinty... ?17:14
th1aIn the meantime... any concerns about going to the zc. date widget?17:14
th1aI've always felt ours was quite sketchy, so I'm happy to be rid of it.17:15
th1aOK.  Guess that's settled then.17:16
faassenwe will hunt and destroy your old widget and replace it.17:16
faassenit may take some doing a we'll likely convert the forms involved to zope.formlib17:16
faassena -> as17:17
faassenas it's much easier to use special widgets with zope.formlib.17:17
faassenanyway, it isn't in use in many places, so we should be okay.17:17
th1aSounds good.17:17
th1aHm... while we see if jinty reappears...17:18
th1aI ended up spending quite a bit of time yesterday exploring the new UI code for the "$100 Laptop."17:18
th1aThere is a team at Red Hat working on it.17:18
th1aIt is Python & GTK.17:19
th1aSort of a tabbed interface for a small display.17:19
th1aYou can see a bit of it here:
mgedminhow large is the display on that device?17:19
th1aBigger than your 770.17:19
th1aSmaller than a laptop.17:19
mgedminI meant in pixels17:20
th1aThey're making extensive use of Zeroconf, which makes sense since they're using mesh networking.17:20
th1amgedmin:  Not sure.17:20
ignasth1a, any ideas for the PoV deadline ?17:21
th1aThey also have their own chat protocol going.17:21
th1aignas:  Prior to June 17?17:22
th1aAnyhow, my point is that for the $100 laptop we may eventually see the stunning return of a REST client for SchoolTool.17:23
ignascan you remind me why June 17 ?17:23
faassenyay, REST. :)17:23
th1aAlso that it will probably be easy to build chat notifications for the system into SchoolTool.17:23
th1aThat's just when srichter and I are doing a SchoolTool workshop in Maine.17:24
th1aIt would be nice to go through the current code with them.17:24
th1aSpend a whole day with actual users trying to use it.17:24
th1aSo that's one landmark in the future.17:25
faassenwow, that's cool.17:26
th1aHaving lost jinty, I guess we can wrap this up.17:26
faassenth1a: any idea what could be a followon for infrae after demographics?17:26
faassenth1a: I'd like to know how to plan for june.17:26
faassenth1a: (and I need to let you know we'll be out for about a week during europython, early july)17:26
th1aOr, I can release the rest of you to continue discussing the UI with faassen.17:26
faassenth1a: okay. :)17:26
faassenrelease them!17:26
th1aClearly I should not try to apt-get upgrade during these chats.17:27
mgedmin640x480, looks like17:27
mgedminmy nokia 770 has a larger screen17:27
th1aLarger in pixels.17:28
th1aYes, the $100 laptop will require some UI tweaking for many apps.17:28
th1aOK... have a good week, folks.17:28
* th1a bangs the virtual gravel.17:29
th1afaassen:  Let me fire up my SchoolTool.17:29
th1afaassen:  How do I get to the demographics for an individual student?17:30
faassenth1a: pupeno reproduced the issue on his install, by the way - it seems to be security related.17:30
th1aFor all non-managers?17:30
faassenth1a: go to the table screen and then click 'edit' (or display)17:30
faassenth1a: I have no idea, we're still investigating what causes this.17:30
faassenth1a: it's non-intentional.17:30
jintysorry all, the person who is going to live in my room for two months arrived to pay the deposit17:31
* jinty lives in a madhouse17:31
th1ajinty:  OK.  We'll have to discuss the release later.17:31
mgedminis there a ftest for the demographics stuff that looks at the table with a non-manager browser?17:32
jintyth1a: he just left, I can now17:33
th1afaassen:  It seems to me that this particular search form is pretty simple.17:33
th1aThat is, in the future there will be search forms that will be used in generating reports.17:33
th1aWhere you'd, say, want to compare girls and boys.17:34
Pupenomgedmin: not yet.17:34
faassenth1a: sure, we have a very simple search form, just for finding people.17:34
th1aBut in this case, you're just worrying about finding and creating individuals or small groups.17:34
faassenth1a: the search form is only about finding people, small or large amounts.17:35
th1aSo is this search form/widget going to be reused, or is this just for editing demographics?17:35
th1aOne definite is parents' name.17:36
faassenI'm trying to understand the question.17:36
faassenthe search form right now involves looking for people (to view them or edit them)17:37
th1aOK.  I'll focus on the task at hand.17:37
faassenthe table + search infrastructure is written in a reusable way.17:37
faassenso we could start wiring that up elsewhere so we can find other things.17:37
faassenso we could have table based browsing of courses or sections, or whatnot.17:37
th1a* parents names17:37
th1a* ID17:38
th1aI think that's sufficient.17:39
th1aTo start.17:39
faassenokay. :)17:39
faassenwe'll add those.17:40
faassenid being student id, right?17:40
faassenwhat about columns? we can add sortable columns for just about anything.17:40
th1aWell, I suppose what I need to think about is how this interacts with a proper "next student" workflow.17:43
faassenthere are all kinds of UI possibilities.17:43
faassenthere's also the add student workflow.17:43
faassenI did a bit of work to make sure we have a last modified date.17:43
faassenand we'll sort the table to show the recently modified students on top.17:43
faassenso at least you can find what you just added. :)17:43
faassenbut what needs to be in the add form?17:43
th1aOK.  Here's one thing.  When you sort by name, it has to be by last name, even if it is displaying the first name first.17:44
th1aSorting by first name drives teachers insane.17:44
th1aI know that from experience.17:45
faassenright, we already have that on the todo.17:45
faassento have last name in there.17:45
faassenbut I'll make sure we do that. :)17:45
faassenwe can easily make it sort on last name, if people fil that in.17:45
faassenI mean, that's one question.17:45
faassenright now lots of the bits are optional.17:45
faassenincluding last name.17:45
faassenif you want sensible sorting on last name, you need to require it.17:45
faassenotherwise people will be even more insane. :)17:45
th1aI would sort by last name by default.17:45
faassenwell, we don't know the last name.17:46
faassenas it's not required.17:46
faassenso it could be empty.17:46
faassenand then what do you sort by?17:46
th1aOh, I see what you mean.17:46
th1aWe certainly could require a last name.17:46
faassenright, we'll do that.17:46
faassenwe'll add it to the add form as well then.17:46
faassenwe're redoing the add form anyway.17:46
faassenas that's still from the old person code.17:46
faassenright now the person package is doing less and less.17:46
faassennot sure how that should all be balanced out eventually.17:47
th1aThe fact of the matter is that there is very little reason to sort in unusual ways in this screen.17:47
faassenright, we'll just see what people will want eventually.17:47
th1aBut I can see some uses for more searches.17:47
faassenlike, modified datetime is useful, and last name.17:47
faassenwe have some other sortable columns, but that's because it's easy.17:47
faassenanyway, we can now easily add searches, change columns or sort order, etc.17:47
faassenand I'm trying to get the code as simple as possible.17:48
faassenso that's easy to make more of such screens.17:48
faassenand hopefully replace some of the existing screens with this stuff.17:48
faassenoh, we still have an open question about the lists.17:48
faassenlet me check my mail, to see what my questions are. :)17:48
th1aOK.  Yeah, let's keep it simple.17:48
faassenokay, we discussed search screen.17:49
faassenwe discussed columns.17:49
faassenI have some ide about the add form.17:49
faassenone question I didn't ask but should..17:49
faassenis how we want to manage the choice lists.17:49
faassenlike, ethnicity.17:49
faassenthings like that.17:50
th1aOh, right!17:50
faassenthat just ties into customizability and so on.17:50
faassenright now I just had something random in there.17:50
faassenbut we need a list of things in there.17:50
faassenand if these are school specific, well, then we need to demonstrate we can do demographics extension packages. :)17:50
faassenas we could simply have school specific forms entirely.17:50
faassenthe theory is that this is easy to do now.17:51
faassenbut we could find out in practice. :)17:51
th1aIdeally, these lists would be parsed from text files.17:51
th1aBut for now we can just hard code dictionaries or whatever.17:51
th1aThat's what I've been thinking.17:52
faassenwell, then you need to mail the list and let us know what should be in them.17:52
faassenI'm happy to hard-code.17:52
th1aI don't want to make this a test case for extensibility.17:52
faassenand I think hard-coding per school would work reasonably well.17:52
Pupenoth1a: why not having them stored in the ZODB accessed thru some general configuration page ?17:52
faassenright, that's another option.17:52
th1aI added a to-do.  Should have some reasonable defaults tomorrow.17:52
th1aPupeno:  That would work.17:52
faassenlet us know what you'd like.17:53
th1aHow much longer would it take to do it through the web?17:53
faassenth1a: you don't want to make this a test case for extensibility, but we do have it coming up in september. :)17:53
th1aI just don't think it is worth the time right now.17:53
faassenth1a: a day's worth of work I expect, perhaps two depending on what we run into.17:54
faassenth1a: anyway, any upcoming tasks for early june?17:54
th1aWell...srichter's travels have screwed us up a bit.17:55
th1aSince your gradebook work is blocked by his.17:55
th1aI'll have to work out the timeframe with him in more detail.17:56
faassenI also need to know what the gradebook work entails in more detail.17:56
th1aWe'll let you know.  Right ;-)17:56
faassenit if involves UI work we'll need to do quite different planning and a UI design first.17:56
th1aTrue enough.17:56
faassenas I'm not going to dive into it just like this. :)17:56
th1aOK.  I'll work on the plan.17:57
faassenanyway, other work we could do is more general maintenance work.17:57
faassenwe could spread search and tables to other areas, convert forms to formlib, etc.17:57
th1aRight.  I just have to watch my budget.17:57
faassenwriting simple code is hard work the first time around, but we got some patterns now.17:57
faassenanyway, I've got more than enough work in june otherwise.17:58
faassenon document library.17:58
th1aI'll let you know what I need soon.17:58
faassenthanks. :)17:58
faassenokay, added some todos, we'll go to work again, thanks!17:58
th1aThanks faassen!17:58
th1aNice meeting you Pupeno.17:59
faassenI'm sure glad Pupeno is helping me. :017:59
*** alga has quit IRC18:00
jintyth1a: should we talk about the release now?18:02
PupenoNice meeting you too.18:06
PupenoI am trying to debug a functional test in the app module, it is this one:
Pupenowhat do the ... mean there ?18:12
mgedmina wildcard that matches any amount of text18:24
mgedmin(if you define the ELLIPSIS doctest flag, which we usually do)18:24
mgedminthe diffs are not very usable in that case, though, which is a pity18:25
Pupenoso, I can ignore unmatches such as ?18:26
Pupenomore worring unmatches are like these ones: (I haven't touched anything related to that html element).18:27
mgedminone technique for debugging these is to divide the ftest into two parts18:28
mgedminby inserting '...' and a blank line18:28
mgedminso that effectively the test only checks for the first part, followed by anything18:28
mgedminthis lets you isolate the actual chunk that doesn't match18:28
mgedminyeha, ignore diffs in whitespace18:29
Pupenommhh... ok.18:29
mgedminsometimes it is simpler to rewrite the test18:31
mgedminand use, e.g., xpath queries to isolate just those chunks that you want to test18:32
mgedminschooltool.xmlparser has useful XMLDocument and HTMLDocument classes for this18:33
mgedminexample test: src/schooltool/app/rest/ftests/acl.txt18:33
mgedminschooltool.testing has an analyze module with a queryHTML function that many ftests use18:34
mgedminexample: src/schooltool/attendance/browser/ftests/attendance-forms.txt18:34
faassenth1a: that bug of table no-shows should be fixed now.18:38
povbot/svn/commits: * faassen committed revision 6089:18:38
povbot/svn/commits: Fix a wrong use of canModify.18:38
faassenth1a: you need to update the demo instance18:38
PupenoTo solve my problem I basically ignored every un-matching line except those directly related to my addition, the false positives distracted me.18:41
jintyfaassen: done18:41
srichtersorry guys; I totally missed the meeting time18:41
srichterI was thinking about it all morning too18:41
ignashi srichter18:41
srichterI was on my computer too, darn18:42
ignasif you have any questions about the new access control model just ask, i will probably answer them (with some latency as i'll be pair programming with gintas now)18:42
faassenignas: what is the repercussion for the grunt programmers like me?18:43
faassenignas: do I need to like, change permissions?18:43
srichterignas: I had more a comment, saying that I think it is not a good idea; but I am not up to speed to make a solid argument18:44
srichterignas: it is just one more piece of code that we have to maintain18:44
ignasfaassen, no reprecussions yet, it's in the branch at the moment18:44
ignassrichter, if we are lucky it will not eat any more time than modules like relationships do18:45
srichterlet's hope so18:45
ignasthe codebase is pretty small18:46
*** jinty has quit IRC18:46
faassenignas: I meant of course to ask "when the branch is merged" :)18:50
faassenignas: what are the repercussions for developers then?18:50
faassenignas: relationship model, there's a zc.relationship or something that uses an index, so if that is starting to cost time we could investigate that. :)18:51
faassenignas: oh, I think you meant developer time not speedtime. :)18:51
ignasfaassen, yes, developer time18:51
* mgedmin is studying zc.relationship right now18:52
povbot/svn/commits: * faassen committed revision 6090:18:52
povbot/svn/commits: Remove enc-type on GET, this is just plain wrong.18:52
ignasas for reprecussions, you will have to understand the new security model that's all18:52
faassenignas: oh, okay.18:56
faassenignas: that gives me the answer to my question entirely. :)18:57
faassenignas: so, as developers I will need to understand the new security model and adjust our code according to that understanding? :)18:57
ignasprobably you will not need to adjust your old code, just the new one ;)18:58
faassenadjust new code?19:00
faassenoh, you mean the security policy.19:01
povbot/svn/commits: * faassen committed revision 6091:19:02
povbot/svn/commits: Some CSS tweaks to make the forms render a bit better in case the title of the field is longer than fits on the screen. Previously this would make the next field appear in the wrong place.19:02
povbot/svn/commits: This fixes this, though the rendering is still imperfect, so this is a work in progress...19:03
ignasfaassen, i mean, you will have to add some code if you will want to add relationship based permissions to your objects, but it's much better than subscriber based hell19:04
povbot/svn/commits: * faassen committed revision 6092:19:07
povbot/svn/commits: Put in copyright headers.19:08
*** Pupeno has quit IRC19:12
faassenignas: good to hear.19:14
faassenignas: but no new permissions introduced?19:15
ignasnope, just all of them except for schooltool.edit/schooltool.view removed19:15
faassenignas: cool. :)19:19
faassenignas: that sounds good. :)19:19
faassensorry mr buildbot, I'm on it!19:19
povbot/svn/commits: * faassen committed revision 6093:19:27
povbot/svn/commits: Fix broken test.19:27
* srichter learned early how to shut buildbot up19:27
faassenrun all the tests? :)19:28
faassenI was feeding patches along that came from Pupeno and he already noticed the problem before I checked in, so I just did this by accident. :)19:29
*** Pupeno has joined #schooltool19:29
mgedmingintas had found a great way to slip behind buildbot's back: use svk!19:31
mgedminit always touches a property on the project root, or on /trunk19:32
mgedminwhich is outside the buildbot directory filter19:32
mgedminand buildbot had a bug that made it skip the entire checkin if a single change was outside the subtree of interest19:32
faassenheading home. :)19:34
faassensee you later!19:34
*** faassen has quit IRC19:34
*** jinty has joined #schooltool19:36
*** jinty_ has joined #schooltool20:11
*** ignas has quit IRC20:15
*** ignas has joined #schooltool20:15
*** jinty has quit IRC20:26
*** ignas has quit IRC21:01
*** thisfred has quit IRC21:27
*** jinty_ has quit IRC21:45
*** Aiste has quit IRC21:46
*** Aiste has joined #schooltool22:17
*** mgedmin has quit IRC23:02
*** mgedmin has joined #schooltool23:30
*** alga has joined #SchoolTool23:57

Generated by 2.15.1 by Marius Gedminas - find it at!