IRC log of #zope3-dev for Friday, 2005-02-18

srichterthat's a shame00:00
projekt01You have URL's like http://localhost:8080/folder/contents.html/++help++contexthelp.html00:00
projekt01This means we get the context at http://localhost:8080/++help++contexthelp.html00:00
projekt01Adapting the folder/contes.html view as the context00:01
srichterbut that does not matter00:01
srichterbecause the OnlineHelp class is a ContainmentRoot00:01
srichterthat means everything is relative to ++help00:01
projekt01Yes, but...00:02
projekt01You accessing the contectual help on the URL00:02
srichtermmh, I don't understand what the problem is00:03
srichterok, so what?00:03
projekt01This points to the topic say, ++help++/first topic00:03
projekt01The child of first topic is xxx00:03
projekt01If you access it via the contextual topic view the URL doesn't it to the child00:04
projekt01One more example:00:04
projekt01The URL of the contextual topic is:00:04
projekt01The URL of the child is:00:05
projekt01The cookie tree can't hanlde this jkind of URL00:06
srichteroh, I see00:06
srichterso really, only the first level is a problem00:06
srichterit would be better if the first URL were: ++help++/contexthelp.html00:06
projekt01No you can access contextual topics on every context00:06
srichteroooh, I see00:07
srichterso contexthelp.html should forward you to something like http://localhost:8080/++help++/topic00:07
projekt01Contexthelp.html is a view adapting the context, The context is something like localhost:8080/site/folder00:08
projekt01++help++ makes the adaption00:09
projekt01For the contextual help view00:09
srichterJ1m: what do I have to do again, so that I can use "..." as a wildcard in doctests?00:19
srichterJ1m: nevermind00:20
*** _projekt01 has joined #zope3-dev00:29
*** projekt01 has quit IRC00:29
*** tvon|desk has joined #zope3-dev00:52
benji_york_zopesrichter, I'm getting a functional test failure...00:52
benji_york_zope    ZopeXMLConfigurationError: File "/home/benji/test/src/zope3/src/zope/app/apidoc/bookmodule/book.zcml", line 5.200:52
benji_york_zope    ConfigurationError: ('Unknown directive', u'', u'bookchapter')00:52
srichterok, done00:54
srichterhowever, the bookmoduel depends on the onlinehelp00:54
srichterwhich Jim just mentioned you don't have00:54
benji_york_zopesame error00:56
srichterdo you have onlinehelp installed?00:56
benji_york_zopeI don't think so.00:57
srichterright, can you install it in your installation?00:57
srichterotherwise you have to comemnts out the bookmodule00:57
benji_york_zopeIs the bookmodule part of the standard instance skeleton?00:57
GaryPosterwhat configure.zcml imports the bookmodule?00:57
GaryPoster(hi stephan ;-) )00:58
srichterGaryPoster: hi00:58
srichteralso, you need to comment out the meta configure stuff for the bookmodule in package-includes/apidoc-meta.zcml00:58
J1mwe'll just load onlinehelp for now01:01
J1mso apidoc depends on onlinehelp.01:03
J1mThat's fair enough.01:03
J1mAs long as they are both optional.01:04
J1mWe want apidoc during development, so we'll load onlinehelp then as well.01:04
GaryPosterfor now, I tried   <include package=""/> and I still got ZopeXMLConfigurationError: File "/Users/gary/jicpac/src/Zope3-trunk/src/zope/app/onlinehelp/help/configure.zcml", line 8.201:04
GaryPoster    ConfigurationError: ('Unknown directive', u'', u'register')01:04
J1mYou have to load the meta.zcml too.01:04
GaryPostereh, I guess I need the meta first.01:04
srichterok, gotta go01:07
srichtersend me a mail, if you have troubles01:08
*** srichter has quit IRC01:08
*** benji_york_zope has quit IRC01:18
*** GaryPoster has quit IRC01:20
* tvon|desk wonders why the jscript menu is busted in the CSS skin01:30
*** srichter has joined #zope3-dev01:38
*** ChanServ sets mode: +o srichter01:39
*** mexiKON has joined #zope3-dev01:43
*** philiZZZ has quit IRC02:00
_projekt01Tvon|desk, do you mean the CSS skin I check in before?02:06
tvon|desk_projekt01: yeah, its fine... didn't realize that the nav menu needed some css to not die02:07
_projekt01Yup, I hope we can define the styles from scratch in the next couple days.02:08
tvon|deskI might toss something to the list tonight..02:09
tvon|deskif I dont go see a bad movie02:09
*** J1m has quit IRC02:28
*** hazmat has joined #zope3-dev03:06
*** hazmat has quit IRC03:53
tvon|deskIs anyone really attached to the javascript menus?04:14
srichternot me, but how else do you want to do sub-menues later?04:15
tvon|deskwithout reloading the whole page? dunno really...04:17
*** niemeyer has quit IRC04:17
srichterso it's a usability issue04:17
srichterI have seen JS menus that work well in Konqui as well ;-)04:18
srichterthat's all I care about04:18
tvon|deskthe current one is busted on konqi, right?04:19
tvon|desksuppose I should just install it...04:20
srichteroh, you mean the tree04:20
srichterwe really should use static tree for the tree04:20
srichterKonqui does not handle the current one04:20
_projekt01Let's develope a own js tree implementation with fits in (all) browser04:21
tvon|desksrichter: what tree were you talking about?04:22
srichteron the left side04:22
tvon|desk_projekt01: there are lots of trees out there, likely easier to find one that fits whatever licensing requirements we have04:22
srichterwe should not go there04:22
tvon|deskthird party code?04:23
_projekt01I now many of them, but they all doesn't fit for my wishes.04:23
_projekt01Support for opera, konquerror, support reload via http request in the background, support drag and drop for nodes etc04:24
srichterif the developer is not willing to check in the code under ZPL in our repository, then its a no go04:24
_projekt01And you find many trees which are build with javascript methods like tree = new Tree(); tree.addItem(level, node)04:25
*** `anthony has quit IRC07:03
*** `anthony has joined #zope3-dev07:32
*** apauley|away is now known as apauley08:10
*** srichter has quit IRC08:27
*** zagy has joined #zope3-dev08:39
*** stub has joined #zope3-dev08:56
*** Theuni has joined #zope3-dev09:28
*** d2m has quit IRC09:51
*** d2m has joined #zope3-dev09:55
*** `anthony has quit IRC09:56
*** hdima has joined #zope3-dev10:53
*** mgedmin has joined #zope3-dev11:07
*** gintas has joined #zope3-dev11:12
*** J1m has joined #zope3-dev11:15
*** MalcolmC has joined #zope3-dev11:16
*** bskahan has quit IRC11:43
*** zug has joined #zope3-dev11:57
*** rejj has quit IRC11:57
*** zug is now known as rejj11:57
*** AJC has quit IRC11:58
*** apauley has quit IRC12:33
*** gintas has quit IRC13:04
*** gintas has joined #zope3-dev13:06
*** alga has joined #zope3-dev13:13
*** zug has joined #zope3-dev13:24
*** rejj has quit IRC13:24
*** zug is now known as rejj13:24
*** srichter has joined #zope3-dev13:35
*** ChanServ sets mode: +o srichter13:35
*** Theuni has quit IRC13:36
*** niemeyer has joined #zope3-dev14:10
*** J1m has quit IRC14:18
*** rejj has quit IRC14:19
*** `anthony has joined #zope3-dev15:03
*** tarek has joined #zope3-dev15:11
*** rejj has joined #zope3-dev15:16
*** stub has quit IRC15:20
*** stub has joined #zope3-dev15:26
*** regebro has quit IRC15:29
*** mgedmin has quit IRC16:12
*** vlado has joined #zope3-dev16:18
*** gintas has quit IRC16:22
*** AJC has joined #zope3-dev16:27
tareksrichter: hello16:28
srichtertarek: hello16:29
tarekabout my mail on zope3-dev, maybe i can explain it here if you have a minute ?16:29
srichter(thought I am not the most familiar guy with mailers)16:30
tarekok, so what i have understood of the mail queuing thing16:30
srichtermmh, marius is not here today; he wrote the thing16:30
tarekoh ok do you want me to wait for him instead ?16:30
srichterI think so16:31
srichterI think I understnad your problem; I am just not qualified to answer16:31
*** J1m has joined #zope3-dev16:32
srichtertarek: how many SMTP servers are you dealing with?16:33
srichtertarek: this seems a very unusual scenario16:33
tareki am making a webmail that let each user specify his or her smtp16:33
tarekat this time i have a simple call to a regular webmail16:34
tarekat this time i have a simple call to a regular sendmail16:34
tarekbut i wanted to used the threaded queue16:34
tarekbecause it rocks16:34
tarekno i need to figure out how to be able to give smtp infos on the fly16:35
tarekand in *one* queue16:35
tareki think it is not possible as ti is right now16:35
srichterso here is what I would do: write your own IMailer interface that extends the current one and overrides this particular method16:36
srichterthis way you do not have to mess with the existing implementation16:36
tarekyes indeed, that what i wanted to do, but i was not sure16:36
tareki wnated to implement IMailQueueProcessor with my own IMailer interface for the mailer object16:37
tarekthanks for the tip16:38
*** hdima has quit IRC16:43
mexiKONsrichter, you got any idea why benji york committed to a tag??16:45
mexiKONnevermind, i see he committed to the trunk as well16:45
srichterthey work on the tag until they can receover from my merge16:46
mexiKONi figured something like this16:47
mexiKONi was just stumbled by the fact that he was committing to a tag16:47
srichteryeah, I think he was not even thinking about it :-)16:47
J1mit was an accident16:52
J1mand a test to see if mexiKON was paying attention.16:52
srichterthe checkin police strikes again! :-)16:53
mexiKONi actually am still reading checkins but most of the time i don't have the time to "strike" when i see something16:54
mexiKONbut srichter is covering me well :)16:54
*** vlado has quit IRC17:07
*** vlado has joined #zope3-dev17:09
*** GaryPoster has joined #zope3-dev17:26
*** mgedmin has joined #zope3-dev17:49
*** __gotcha has quit IRC18:20
*** __gotcha has joined #zope3-dev18:21
*** tvon|desk has quit IRC18:33
*** bskahan has joined #zope3-dev18:51
*** AJC has quit IRC19:31
*** tvon|desk has joined #zope3-dev19:33
*** __gotcha has quit IRC19:36
mexiKONsrichter, btw, in favour of making a 3.1 release easier, i think the new i18n message impl shouldn't be in it19:43
mexiKONwhen i was still working on it i ran into some problems that i need to further discuss19:44
*** MalcolmC has quit IRC19:44
mexiKONplus, nothing in the trunk uses the new implementation anyways19:44
mexiKONso, it's currently just sitting there on its own19:45
*** benji_york has joined #zope3-dev19:48
srichtermexiKON: wasn't the recent work on it solving the last problem?20:12
mexiKONi never quite finished it20:15
srichterwhat is missing?20:19
mexiKONnothing's missing20:20
mexiKONit's the way you work with messages20:20
mexiKONcreating new ones20:20
mexiKONsince they are immutable20:20
mexiKONthere were some unresolved issues20:20
J1msrichter, is there a simplified api for registering utilities?20:33
srichterin there is20:33
srichterreally, much of ztapi needs to go somewhere else20:33
srichterits generally useful20:33
J1msome if it should be deprecated.20:34
srichteryeah, I noticed a lot of overlap between placefulsetup, ztapi, ...20:34
srichterI did not feel qualified to do so, though20:34
J1mI want to get rid of placefulsetup someday.20:35
J1mI never use it myself any more.20:35
srichterit has useful stuff in it20:35
srichterlike setUpAnnotation20:35
srichteror setUpTraversal20:35
srichterI really like these shortcuts20:35
J1msrichter, is there a simplified api for registering *local* utilities?20:35
srichterI always found it annoying to do that myself20:35
srichterI think it is in placefullsetup20:36
srichternope, it is not20:36
srichterlet me search further20:36
srichteroh, it is in setup.py20:36
srichterthat's where all the other useful functions are I just mentioned....20:37
srichterdef addUtility(sitemanager, name, iface, utility, suffix='')20:37
srichterI think having a fully-colored API like zope.component would be nice20:38
srichterI started this already20:38
mgedminsrichter, what do you mean by "fully-colored"?20:41
srichtermgedmin: an API that provides all the tools you would want20:42
srichterlike zope.component does now20:42
*** vlado has quit IRC20:42
srichterregisterLocalUtility(), registerLocalAdapter(), ...20:43
srichtermgedmin: so you are going to implement zcml:condition this weekend? :-)20:44
mgedminsrichter, yes20:44
srichterthat's so cool20:44
mgedminimplementation hints are welcome, of course ;)20:45
srichtermmh, I was thinking about it and I was really at a loss20:45
mgedminI suppose I'll take a look at the old zcml:condition that was removed20:45
srichterI guess you would have to hard code the handler somehow20:45
mgedminfind out the places that need to be modified20:45
srichterwas the old code available as an attribute on any directive (J1m)?20:46
J1mare you talking about zcml:condition?20:50
J1mzcml:condition can apply to any directive.20:51
srichterand that was already implemented?20:51
J1mmgedmin, I hope you'll start with the work that Fred did.20:51
J1mThat's why I have the revision number in my rfc last night.20:52
J1mThat's why I gave the revision number in my rfc last night.20:52
mgedminJ1m, I was planning to start by reading that checkin you mentioned yesterday20:52
J1mThere are two parts:20:52
J1m1. controllng zcml20:52
J1m2. evaluating expressions20:53
J1mYou should be able to reuse 1.20:53
* mgedmin was just bitten (indirectly) by
J1mYou'll have to redo 2.20:53
J1mshould be easy to fix20:53
srichteroh, this seems very easy20:54
srichterthe implementation was slick20:54
mgedminby changing MultiSelectWidget to always return set.Set objects?20:54
J1mmgedmin, no, by creating flavors appropriate for the field type.20:54
J1mThey can probably subclass that one and let it still do most of the heavy lifting.20:55
* mgedmin greps throught configure.zcmls20:56
mgedminMultiSelectWidget is only registered for (ISet, IVocabularyTokenized)20:57
mgedminso the field type is always ISet -- or do I misunderstand what you mean by "field type"?20:57
J1mIn that case, the field type is always ISet.20:57
J1mPerhaps this thing should also be registered for IList (with unique) and ITuple (with unique),20:58
J1mDoes this thing let you control ordering?20:58
mgedminno, but there's an OrderedMultiSelectWidget that inherits from MultiSelectWidget20:59
mgedminerr, no it doesn't20:59
mgedminclass OrderedMultiSelectWidget(ItemsMultiEditWidgetBase):20:59
mgedminso it seems to me that the simplest fix would be20:59
mgedmindef _toFieldValue(self):20:59
Workbliamgedmin, i tried fixing it, yet it breaks some tests21:00
mgedmin    return sets.Set(ItemsMultiEditWidgetBase._toFieldValue(input))21:00
Workbliaoh you mean one level higher21:01
mgedminWorkblia, two levels lower21:01
mgedminWorkblia, you changed MultiDataHelper._toFieldValue directly, which is used for lists as well21:01
Workbliai see21:02
Workbliawill you do it or should i try fixing it ?21:02
mgedminI'm working on something else at the moment21:03
Workbliai am working on widgets then21:03
WorkbliaMultiCheckBoxWidget is it Set or List based ?21:13
Workbliaand how to check ?21:13
J1msrichter, the bootstrap code is in poor shape.21:20
J1mIt's rather brittle.21:20
srichterI know21:20
srichterit is disgusting21:20
Workbliacan i draw an ascii pentagram in widget code to protect me ?21:22
* mgedmin was wondering the other day how he could hook into bootstrap code and create the root folder differently21:23
* mgedmin was also wondering whether he could use zcml overrides to replace an event handler21:23
J1mYou can't currently override subscribers.21:24
SteveAi totally removed the zcml that references bootstrap code for launchpad21:25
SteveAseeing as we're not depending on zodb for root objects21:26
mgedminand I can't control the order of subscribers either21:27
SteveAi think the bootstrap code should be run on creating an instance, and by the upgrade script on upgrading an instance.  not automatically at startup.21:27
SteveAmgedmin: i had an idea once about each bootstrap subscriber sending an "IThatBootstrapTaskFinised" event21:28
SteveAthen you can subscribe to the events you depend on, and see when they are done21:28
J1mmgedmin, the bootstrap code actually depends on the subscriber ordering.21:28
SteveAto make this easier, i had the idea of extending the events system to allow you to subscribe to "IFooFinished event while IBarFinished event still being processed" events.21:29
J1mAs an implementation accident, subscribers of the same type are execited in order of registration.21:29
mgedminI'd like to have schoolbell running as a standalone app21:29
* SteveA waves hands some more21:29
mgedminwith a SchoolBellApplication instance as the ZODB root object21:29
* SteveA waves hands about ZODB root not being root of zope publishing, and a zcml:stick-an-app-at-this-url directive21:30
mgedminbut if I <include package="" />, I have no chance to insert a bootstrap subscriber that would run before the default ones21:30
mgedminand, I believe,'s configure.zcml includes the meta zcml file that defines <subscriber>21:31
* mgedmin murmurs something about "vapourware" and "hard deadlines"21:31
SteveAmgedmin: in launchpad, i don't include, but include my own file that includes much of what includes21:31
mgedminanyway, I think I have workable a solution for my problem21:31
SteveAnotably excluding bootstrap stuff21:32
mgedmin(I can open the ZODB manually, stick my object into root, and then publish the DatabaseOpened event myself)21:32
mgedminI just wanted to know if there was a "nicer" way21:32
mgedminor, alternatively, I wanted the zope 3 developers know about this "stick-my-application-instead-of-RootFolder" use case21:32
SteveAthe root folder should be rooted with a "stick the zope3 root folder at /" directive.  you should be able to override this directive.21:33
SteveAfancy working on that instead? ;-)21:33
J1mMy long term plan is to allow multiple root objects with some sort of mapping, as y'all have done.21:35
*** tvon|desk is now known as tvon21:36
*** tarek has quit IRC21:39
J1mI'm adding an event that gets generated when a new local site is created21:47
J1mHm, but I'm not sure that this is what I want. :(21:48
J1mSteveA, the problem with your idea of doing this when the instance is created is that you might want to add things later.21:49
*** bradb has quit IRC21:54
SteveAat what point later?21:58
J1many point. You decide that you want your root to have a foo utility.  You want to be able to arrange that existing instances can get this.22:01
J1mYou may decide you want more *after* the instance has been created.22:01
SteveAthat's the same problem as modifying any persistent data.  it is an upgrade problem.22:04
J1mNo, I don't think so.22:04
J1mSuppose you install a new application that wants something in the root.22:04
J1mThere is no previous generation.22:04
* SteveA resolves to think upon this :-)22:05
mgedminwhat about being able to run an "install" script when installing a schema manager?22:06
mgedminI was wondering about this when I studied generations22:06
* SteveA gets ready to leave the office22:06
*** niemeyer has quit IRC22:07
*** SteveA has quit IRC22:12
*** stub has left #zope3-dev22:38
J1mme tries mgedmin's install idea23:00
*** bskahan has quit IRC23:16
*** bskahan has joined #zope3-dev23:17
*** tvon has quit IRC23:18
*** tvon has joined #zope3-dev23:27
*** tvon|x31 has joined #zope3-dev23:33
*** vlado has joined #zope3-dev23:46
*** bskahan has quit IRC23:49
*** bskahan has joined #zope3-dev23:51
*** tvon has quit IRC23:54
*** tvon|x31 has quit IRC23:54

Generated by 2.15.1 by Marius Gedminas - find it at!