IRC log of #zope3-dev for Thursday, 2006-01-19

*** wrobel has quit IRC00:04
*** pcardune has quit IRC00:06
*** wrobel has joined #zope3-dev00:06
*** MJ has quit IRC00:11
*** Aiste has joined #zope3-dev00:15
*** sashav has quit IRC00:16
*** RockyBurt|away has quit IRC00:21
*** RockyBurt has joined #zope3-dev00:22
*** andrew_m has quit IRC00:28
*** MJ has joined #zope3-dev00:29
*** GaryPoster has quit IRC00:31
*** yota has quit IRC00:50
*** yota has joined #zope3-dev01:16
*** GaryPoster has joined #zope3-dev01:26
*** zmi_junkie has quit IRC01:28
*** tiredbones has quit IRC01:50
*** philiKON has joined #zope3-dev01:57
*** GaryPoster has quit IRC02:08
*** philiKON has quit IRC02:35
*** natea has quit IRC02:37
*** natea has joined #zope3-dev02:51
*** xenru|clone has joined #zope3-dev02:57
*** benji has quit IRC03:11
*** xenru has quit IRC03:15
*** zbir has quit IRC03:18
*** fluff has quit IRC04:07
*** RockyBurt has quit IRC04:15
*** RockyBurt has joined #zope3-dev04:16
*** xenru has joined #zope3-dev04:16
*** zbir has joined #zope3-dev04:19
*** wrobel has quit IRC04:24
*** wrobel has joined #zope3-dev04:25
*** xenru has quit IRC04:38
*** xenru|clone has quit IRC04:41
*** tahara has joined #zope3-dev04:48
*** stub has joined #zope3-dev05:07
*** RockyBurt is now known as RockyBurt|Zzz05:13
*** MacYET has joined #zope3-dev07:52
*** yota has quit IRC07:52
*** MacYET has left #zope3-dev07:52
*** RockyBurt|Zzz has quit IRC07:56
*** RockyBurt has joined #zope3-dev07:57
*** natea has quit IRC08:02
*** natea has joined #zope3-dev08:02
*** eins has joined #zope3-dev08:16
einshi all08:16
*** sashav has joined #zope3-dev08:19
*** Ragica has quit IRC08:20
*** sashav has quit IRC08:22
*** dobee has joined #zope3-dev08:31
*** wrobel has quit IRC08:44
*** wrobel has joined #zope3-dev08:46
*** dobee has quit IRC08:59
*** hdima has joined #zope3-dev09:01
*** MrTopf has joined #zope3-dev09:06
*** zagy has joined #zope3-dev09:10
*** MJ has quit IRC09:33
*** xenru has joined #zope3-dev09:37
*** sashav has joined #zope3-dev09:51
*** romanofski has joined #zope3-dev09:52
*** stub has quit IRC09:53
*** natea has quit IRC09:54
*** natea_ has joined #zope3-dev09:54
*** jinty has joined #zope3-dev09:57
*** j-w has joined #zope3-dev10:07
*** dobee has joined #zope3-dev10:11
einshi romanofski10:22
* romanofski waves to eins10:25
*** MJ has joined #zope3-dev10:27
*** drzoltron_ has joined #zope3-dev10:38
*** zagy has left #zope3-dev10:38
*** zagy has joined #zope3-dev10:39
*** srichter has quit IRC10:41
*** natea_ has quit IRC10:46
*** natea has joined #zope3-dev10:47
*** drzoltron_ has quit IRC10:50
*** drzoltron_ has joined #zope3-dev10:50
*** srichter has joined #zope3-dev11:00
*** Aiste has quit IRC11:07
*** tarek has quit IRC11:11
*** MrTopf has quit IRC11:16
*** Aiste has joined #zope3-dev11:24
*** Aiste has quit IRC11:24
*** BjornT_ has quit IRC11:40
*** BjornT has joined #zope3-dev11:45
*** andres_ has joined #zope3-dev11:46
*** andres has quit IRC11:47
*** andres_ is now known as andres11:47
*** drzoltron_ has quit IRC12:02
*** BjornT has quit IRC12:07
*** Aiste has joined #zope3-dev12:11
*** BjornT has joined #zope3-dev12:12
andresJ1m, srichter If you remember my problem about the speed impact of securityproxies, i just want to let you know, that the caching is working very nice. First time i iterate over all objects 14s second time. 2s. Perhaps problem is the combination of zope and sqlos/sqlobject. Is the cache for security lookups per request? If not, that would explain the thing. Sqlos objects are thrown away after each transaction in order to achieve transactional safety.12:12
andres So caching cant work.12:12
andres(Both time i iterate over all objects, they are alread cached within sqlos (via an iteration without security proxies), so its not a matter of that)12:13
*** tonico has quit IRC12:24
TheuniCan somebody confirm this statement: The security machiner differentiates between setattr and getattr, but does not handle 'calling a method' different than getattr?12:25
*** jinty has quit IRC12:25
*** stub has joined #zope3-dev12:30
*** andrew_m has joined #zope3-dev12:31
*** xenru has quit IRC12:35
*** andres_ has joined #zope3-dev12:55
*** andres has quit IRC13:01
*** andres_ is now known as andres13:01
*** andres_ has joined #zope3-dev13:06
*** dobee has quit IRC13:09
*** andres has quit IRC13:14
*** andres_ is now known as andres13:14
*** xenru has joined #zope3-dev13:17
*** BjornT_ has joined #zope3-dev13:22
*** ignas has joined #zope3-dev13:23
*** dobee has joined #zope3-dev13:26
*** vinsci has quit IRC13:27
*** mgedmin has joined #zope3-dev13:28
*** MrTopf has joined #zope3-dev13:31
*** J1m_ has joined #zope3-dev13:33
*** vinsci has joined #zope3-dev13:33
*** BjornT has quit IRC13:34
*** tonico has joined #zope3-dev13:41
*** faassen has joined #zope3-dev13:47
*** BjornT_ has quit IRC13:57
*** dobee_ has joined #zope3-dev13:57
*** dobee has quit IRC13:57
*** BjornT has joined #zope3-dev13:57
*** mkerrin has joined #zope3-dev14:00
*** dobee_ has quit IRC14:04
*** jinty has joined #zope3-dev14:14
*** RockyBurt has quit IRC14:14
*** RockyBurt has joined #zope3-dev14:15
*** natea has quit IRC14:20
*** natea has joined #zope3-dev14:20
*** _anguenot has joined #zope3-dev14:35
*** mgedmin has quit IRC14:36
*** alga has joined #zope3-dev14:36
*** J1m_ has quit IRC14:40
*** ChanServ sets mode: +o srichter14:53
srichterTheuni: I can confirm that14:54
*** philiKON has joined #zope3-dev15:13
srichterTheuni: the security mechanism only deals with attribute access and modification15:15
srichterthe way methods work is by getting (accessing) the method and then calling it15:15
srichteronce the method is accessed, you can do what you want; I think this is implemented by making methods rocks15:15
Theuniright. I just noticed an inconsistency in the way steve, jim and I wrote some stuff for the certification mentioning method calls as a third way15:16
srichterah, I see15:16
Theunii just wanted to clear up what the intended state was15:16
*** kobold has joined #zope3-dev15:21
*** benji has joined #zope3-dev15:34
*** benji has quit IRC15:34
*** benji has joined #zope3-dev15:34
*** tiredbones has joined #zope3-dev15:42
*** zbir has quit IRC15:49
*** kobold has quit IRC15:58
RockyBurthm, does anyone know if any work has been done (or needs to be done) to zope3 for it to support eggs at this point?16:02
*** zbir has joined #zope3-dev16:03
srichterRockyBurt: there is a huge discussion on zope3-dev right now16:08
RockyBurtsrichter: i know, thats why i was asking ;)16:08
srichterRockyBurt: some initial work has been done, but I do not know how far it came16:08
RockyBurtseems to me a lot of people are discsussing on a topic they know quite little about ;)16:09
srichterit is not a simple subject16:09
srichterdon't underestimate their understanding16:09
RockyBurtlol, but everyone who's discussing on the list all says they don't really know anything about eggs including the one arguing for them -- jim ;)16:10
srichterthey probably have a good high-level idea of what's going on; they just do not know the details and thus do not know the pitfalls16:10
benjiright, Jim and I have both put just a few hours into eggs and see the prommise, but we don't know if the details will fulfil the prommise or not16:17
RockyBurti have to admit i'm a little confused about the proposed direction, does J1m mean that eggs would be checked into svn?16:20
benjiI hope not :)16:24
*** ruda_porto has joined #zope3-dev16:25
*** mgedmin has joined #zope3-dev16:27
Theuniwe could add bacon, sausages and spam too :)16:45
*** vlado has joined #zope3-dev16:45
RockyBurtlol Theuni16:48
*** natea has quit IRC16:50
*** natea has joined #zope3-dev16:51
*** eins has quit IRC16:53
*** hdima has quit IRC17:14
j-wI'm reading through the zope.schema code and notice this line in
j-wclone = self.__class__.__new__(self.__class__)17:15
j-wisn't that equivalent to clone = self.__class__() ?17:16
*** bradb has joined #zope3-dev17:17
srichterI don't think __new__ and __init__ are necessarily equivalent17:21
srichterin fact it is not17:21
srichternew does nto require any constructor arguments17:21
srichterbut init does17:21
j-woh, ok17:21
j-wright, then I understand17:22
*** bradb has left #zope3-dev17:23
*** tarek has joined #zope3-dev17:28
*** agroszer has joined #zope3-dev17:34
*** sawdog has joined #zope3-dev17:39
*** danfairs has joined #zope3-dev17:59
*** stub has quit IRC18:01
*** tlotze has quit IRC18:09
*** mwegrzynek has joined #zope3-dev18:15
*** mwegrzynek is now known as mikka18:15
*** mikka is now known as mikka__18:15
*** j-w has quit IRC18:15
*** mleist has joined #zope3-dev18:16
*** yota has joined #zope3-dev18:18
*** natea has quit IRC18:18
*** mikka__ has quit IRC18:19
mleistHello, a few days before i asked:18:19
mleistis there any small kind of sourcecode which can do a "simple" trigger of an18:20
mleistecm.workflow direct in an object-method?18:20
mleistI have some xmlrpc-methods in which i have to modify the state of18:20
mleistin my old ""-source it looks like:18:20
mleistdef changeWorkflowFrom...18:20
mleist    adapter = IProcessInstanceContainer( self)18:20
mleist    wf_host1 = removeAllProxies(adapter.get( "wf_host1"))18:20
mleist    wf_host1.fireTransition(u"configured_offline")18:20
mleistany idea or link?18:20
philiKON_anguenot, ping18:20
*** romanofski has quit IRC18:20
_anguenotphiliKON, pong18:22
philiKONmleist, you want to talk to _anguenot  :)18:23
*** sashav has quit IRC18:23
agroszermleist, I can't tell you a better one, that looking at the ecm.workflow.demo, app.py18:23
agroszersorry, ecm.demo.workflow, app.py18:24
agroszerself.participant.activity.workItemFinished(self) is the keyword18:24
agroszerthat triggers the transition if every activity (I think) is finished18:25
*** BjornT has quit IRC18:27
mleistHm, the self.participant.activity.workItemFinished is part of the Application which modifies the workflow.18:29
Theunican somebody tell my why ObjectCreatedEvents seem to be triggered from view code all the time?18:29
agroszerTheuni: that is fired in the add form as I know18:30
Theunii saw that18:30
Theunii wondered why a file object wasn't annotated with creator and creation time when creating it manually. that was a bit surprising ...18:31
Theunianyway ... gotta catch a train. thanks.18:32
*** Theuni has quit IRC18:32
*** BjornT has joined #zope3-dev18:32
agroszermleist, that seems to be the new world order, zope.wfmc works the same way18:34
*** vlado has quit IRC18:36
mleistOk, I have to read the docs again (and again) but i don't see it. There is a trigger from outside (via XMLRPC) and is accepted by my Object. This Object will "call" an Application to modify the Workflow. I don't understand this last step - give me a chance ;-)18:38
agroszermleist, I also did not made to that point in my app, yet18:39
agroszerbut obviously that's the way18:39
agroszeryou have to figure out somehow how to get to the application object18:40
mgedminmleist, as far as I understand wfmc18:41
mgedminyou define an application class18:41
mgedminthat has a start method18:41
mgedminand in the start method you stash the instance somewhere to make it accessible18:41
mgedminthen from any other place you can access the instance of the application (which is usually called a "work item")18:42
mgedminannotations are one choice for stashing work items18:42
mgedminperhaps via an adapter18:42
agroszergood point18:42
agroszerone more idea: have a look at ecm.workflow.browser, edit.py18:44
agroszerhow it gets the workitem from the context18:44
* agroszer have to leave now18:46
*** agroszer has quit IRC18:46
mleistok, i will try it this way - thanks.18:47
*** BjornT_ has joined #zope3-dev18:47
*** zagy has quit IRC18:51
*** tiredbones has quit IRC18:53
*** tiredbones has joined #zope3-dev18:53
*** BjornT has quit IRC19:00
*** deo has quit IRC19:02
*** andres has quit IRC19:06
*** dobee has joined #zope3-dev19:11
*** MJ has quit IRC19:13
dobeesrichter: ayt?19:14
mleistfine, this seems to be a good way. now i have to take paper/pencil and draw my "big picture". Thanks for the moment.19:24
*** romanofski has joined #zope3-dev19:45
*** BjornT_ has quit IRC19:50
*** jinty has quit IRC19:50
*** BjornT has joined #zope3-dev19:55
*** zagy has joined #zope3-dev19:57
*** philiKON has quit IRC20:03
srichterdobee: now I am20:13
dobeehi, i read your mail and looked into zodb svn, but there is no blob support in the trunk ther is just a devel branch20:14
dobeedu you have any clue when the merge should happen of the Blobs package?20:15
srichterdobee: oh, I thought this code made it in the previous release20:15
srichterso we have to lobby (Theuni I think) to get it into the trunk20:15
srichterI think we need to send a mail to the list20:16
srichterJ1m: dobee is interested on working on large file support during the snow sprint20:16
dobeei also think then we have to pay attention on types of storage a zodb uses20:16
srichterapplication code should not need to deal with that20:16
J1mAs in large files stored in ZODB?20:17
srichterit is ZODB's responsibility to implement blobs efficiently for various storages20:17
srichterJ1m: yeah and efficiently getting htme in there and out of it20:17
dobeesrichter: well but you can not use FileStorage and Blobs in one db20:17
dobeeor am i wrong20:17
J1mOnce blobs are in place it should be trivial.20:18
J1mdobee, the ZODB blob storage is a storage adapter.20:18
J1mIt can be used with file storages.20:18
J1manyway, the hard part is getting ZODB blobs.20:20
srichterok, so not much we can do there20:20
J1mThere is a lot of work left to do on that.20:20
srichteranything that someone without much ZODB knowledge can pick up?20:20
srichterJ1m: btw, I need to talk to you somewhen soonish about other possible sprint tasks20:20
J1mNot on that.20:21
srichterright, that's what I thought20:21
srichterdobee: do you know if someone coming to the sprint is a ZODB hacker?20:22
dobeeafaik no, i asked jodok today20:22
srichterwe should not attack problems/tasks that we do not understand well20:23
*** hazmat has joined #zope3-dev20:23
srichterone thing that we could work on is the further simplification of the local component management20:23
srichterI feel comfortable with that once I synced with Jim20:24
J1mYup, but I really need to get the new component registry stuff done first. :/20:24
J1mThat is a project I plan to work on some this weekend.20:24
dobeewell, we thought it would  be a good occasion because we implement a media repository based on zope3 and then we would have enouogh resources for such tasks20:25
J1mMaybe y'all could help me with that on some development branch.20:25
dobeeafter the sprint i mein20:25
*** deo has joined #zope3-dev20:25
srichteryeah, but we cannot attack an issue that we (a) do not understand well and (b) do not have the expertise for20:26
srichterwe really need to lobby the author of ZODB blobs to get the work done20:26
srichterJ1m: good idea20:26
J1mI wonder if there is anything to do along those lines based on tramline.20:27
J1mI should say faassen's tramline. :)20:27
srichterdobee: hearing your thoughts just convinced me to give a development process presentation at the sprint20:27
*** MJ has joined #zope3-dev20:28
dobeeJ1m: i think this is the alternative we'll take then20:28
dobeesrichter: zope3 process?20:28
srichterJ1m: you said that streaming does not work with the current Zoep 3 publisher, right?20:29
J1mThat would need a well though out proposal.20:29
srichterthat could be something to think about; but then someone would have to do some serious homework before the sprint20:29
dobeewe, had a lot of discussions today about tasks, that could also make the plone people come a little bit closer to zope320:29
J1mBTW, given the venue, I'd love to see some Five work.20:30
srichterdobee: are there Five developers there?20:30
J1mA few weeks ago, I mentioned some tasks that I'd like people to work on for merging z3 tech into z2.20:30
srichteryeah, that's more down my alley :-)20:31
dobeei think so, dunno if core developers, but new stuff will be pretty much based on five on the sprint20:31
J1mI don't know if Andreas needs help with his ZPT or unicode work, for example.20:31
srichterwe could attack the security system stuff20:31
J1mYes, although I would need to write up some proposals to give guidance.20:31
srichterthat would be great! :-)20:31
J1mUnfortunately, that's some pretty heavy stuff.20:31
J1mAnother area is integration of the Z3 publisher.20:32
dobeeto heavy for me20:32
srichterI think I understand the Zope 3 implementation pretty ok, or at least could read through it again to understand it20:32
J1mSindnei made some noise about working on that, but I don't know if he did or is doing anything.20:33
srichterJ1m: I have not seen a brnach being created yet20:33
J1msrichter, to do the security work requires deep understanding of both z2 and z3 security.20:33
J1mThat includes acquisition.20:33
J1mI'll note that sane local component management is needed for both z2 and z3.20:34
RockyBurtthere's gonna be a high emphasis on Five/z3 work at the snow sprint20:34
* RockyBurt will be attending20:34
J1mSo I think that could still be a good topic.20:34
srichteryeah, I think so too20:34
RockyBurtand out of the people i know that are going, non are real zodb hackers20:34
srichterJ1m: what about exploring eggs?20:34
RockyBurteggs has been brought up as a potential topic at the sprint20:35
RockyBurtin fact whit and i have been discussing it20:35
* RockyBurt has worked with Basket quite a bit lately, adding support for CMF products, etc20:35
J1mZ2 is ahead of Z3 here.20:35
RockyBurtthat was what i figured20:35
J1mRockyBurt, have y'all tried integrating easy_install or such yet?20:36
RockyBurtJ1m: not yet... not sure what mcdonc has planned for that20:36
RockyBurtwell, easy_install is required for building the eggs if i remember20:36
RockyBurtbut i assume you mean for pulling down dependencies at runtime, etc20:36
J1mNo, just setuptools.20:36
RockyBurtoh yeah20:37
RockyBurtit is just setuptools20:37
J1mand even just installing eggs.20:37
RockyBurtyou're right20:37
J1mbenji, keeps telling me to look at Paste Deploy, so I need to look at that soon.20:37
RockyBurtwould be nice if easy_install could be used to create package slugs for eggs upon install too ;)20:37
J1mI expect eggs will have a greater impact than that.20:38
J1mI expect we want something more like baskets.20:38
RockyBurtyeah i've been hearing good things about Paste lately20:38
J1mI'm also hoping that eggs can help manage checkouts (buildouts) and releases, thus the interest in easy_install, Paste Deplot, etc.20:39
RockyBurtah yes -- i would imagine easy_install would be required for something like that20:39
J1measy_install or something like it.20:40
* srichter notes he has to get back to work ... see ya20:40
* RockyBurt wavews20:40
* RockyBurt reads through the website and decides he's going to have to deploy/develop using something using paste in order to understand it20:41
*** tarek has quit IRC20:43
RockyBurtheh, paste requires easy_install20:44
RockyBurtat least installatino of paste deploy does20:44
J1mOne issue I'm wrestling with is how to get easy_install to work with a directory under my control.20:44
J1mLike one internal to a Zope install or a Zope instance.20:45
RockyBurtyou mean so that it downloads and installs eggs into your zope instance ?20:45
J1mNone of the alternatives at
J1mseem workable20:46
J1mI think we need to propose something else.20:47
J1mIMO, '"Traditional" PYTHONPATH-based Installation' is closest to what I want.20:47
J1mBut I need to play with it some more.20:47
RockyBurtyeah i was just about to say20:48
RockyBurtwhy wouldn't that traditional approach just work?20:48
J1mI played with it last weekend.20:49
J1mIt wants to read info from a configuration file and it's hard to manage that.20:49
J1mIt's not an option for us to require a special config file in a users home, nor to modify the one in the Python distutils directory.20:50
RockyBurtof course20:50
J1mIt will use a config file in the current working directory, but that's messy.20:50
RockyBurtbut perhaps a fresh zope instance would include a $INSTANCE_HOME/bin/zope_easy_install that reads its configuration from somewhere inside the $INSTANCE_HOME which by default makes everything go into $INSTANCE_HOME/lib/python and $INSTANCE_HOME/bin20:50
J1mWhat I want is for the easy install bootstrapping process to create a custom easy_install that remembers the configuration information I have at bootstrapping time.20:51
J1mRight, the bootstrapping process creates a custom easy install.20:51
J1m(well, I'm not sure how custom)20:51
J1mI was able to get it to make an easy install script in my buildouts bin directory.20:52
J1mBut that easy install didn't remember any of the configuration options used to create it.20:52
RockyBurtbootstrapping time would be during instance creation? or?20:52
J1mIf it did and used that information, then I think it could work pretty well.20:52
J1mAlso, I want the scripts that that easy install installs to reflect the system paths used to create the easy install.20:53
RockyBurtcouldn't you simply generate a zope_easy_install file that does "cd $INSTANCE_HOME/easy_install" first and reads a setup.cfg from there to get config from that was generated upon instance creation time ?20:53
J1mNote that cabeat in '"Traditional" PYTHONPATH-based Installation' about having to munch the path.20:53
J1mI could, but I'd rather come up with something a bit more elegent. :)20:54
J1mI consider cd to be kinda evil.20:54
J1mI'm willing to do it if forced.20:54
* RockyBurt wonders if easy_install couldn't simply be invoked from a python script using the easy_install python api and somehow feed it a good config that way20:55
J1mI'm hoping that I can come up with another (5th) option for handling custom install locations.20:55
J1mI know that Phillip wants to handle "applications".20:56
J1mI don't think easy_install does yet.20:56
J1mAnyway, hopefully, we can come up with a proposal good enough to be incorporated in easy_install.20:56
J1mThen we don't have to maintain it (only) ourselves.20:57
RockyBurti should mention that developers of a lot of the major linux distributions haven't expressed resentment regarding eggs20:57
RockyBurtno package-based distribution wants another system installed that does its own package management20:57
RockyBurtparticularly the debian mailing lists have went wild on these discussions20:58
RockyBurtthat original "haven't" was supposed to be "have"20:58
*** natea has joined #zope3-dev20:59
J1mYeah, I understand that.20:59
J1mI think they have different needs and don't always appreciate that.21:00
*** RockyBurt has quit IRC21:11
*** slinkP has joined #zope3-dev21:12
*** RockyBurt has joined #zope3-dev21:13
*** faassen has quit IRC21:14
*** dman13 has joined #zope3-dev21:24
*** RockyBurt has quit IRC21:42
*** RockyBurt has joined #zope3-dev21:42
*** ruda_porto is now known as ruda_porto-out21:51
*** danfairs has quit IRC21:54
*** Aiste has quit IRC22:03
*** MrTopf has quit IRC22:05
*** mkerrin has quit IRC22:05
*** tav has joined #zope3-dev22:05
*** ignas has quit IRC22:17
*** xenru has quit IRC22:27
*** WebMaven has quit IRC22:29
*** _anguenot has quit IRC22:31
*** romanofski has quit IRC22:36
*** slinkP has quit IRC22:40
*** Aiste has joined #zope3-dev22:41
*** RockyBurt is now known as RockyBurt|hockey22:46
*** mgedmin has quit IRC22:54
*** zagy has quit IRC23:05
*** ChanServ sets mode: +o hazmat23:10
*** alga has quit IRC23:12
*** zbir has quit IRC23:30
*** dobee has quit IRC23:37
*** dunny has joined #zope3-dev23:37
*** tarek has joined #zope3-dev23:51
*** dman13 has quit IRC23:54
*** zbir has joined #zope3-dev23:55

Generated by 2.15.1 by Marius Gedminas - find it at!